you don't dream big either. it's the only reason you're here. whine whine whine about how everything is going to shit, no solutions though. you're just like me. you're just not ready to admit it yet
You realize there are still going to be people managing that shit? It will downsize marketing but there will still have to be people making decisions and inputting instructions. This drops costs and boosts productivity. This boosts the economy. The economy grows. There is more need for people to manage marketing. The people who lost their jobs will get jobs managing and inputting commands for the AI at other firms that previously were too small to afford marketing or whatever.
Creative destruction. Been through this shit before when we went through the last industrial revolution. Short term disruption. Societal upskilling. End result: everyone still has to go to work.
You fail to see that we already have a huge amount of makework in society and this just automates makework. This will btfo of more work than anyone can admit publicly. What's left is incredibly high level or menial slave shit like physical labor or doing something that a robot is too expensive to do.
that's not the issue though, you already can work less hours and not need very much if you are sharp with your money. It's that artificial scarcities keep society motivated in a positive direction.
2 weeks ago
Anonymous
It is. Work less, and have as much or more as you do now. Not work less, and have less/spend less.
The market demand for humans working in marketing will shrink by 95%, moron. >but not EVERYONE in marketing will lose their job
The fact that you thought this was a point against AI, god you're fucking dumb.
And what happens to those people who've lost their jobs? They go work in an Amazon warehouse or become unemployable because machines and AI can do every task that once took a team of humans to do.
The world doesn't change to accommodate this type of change because the people benefitting from AI productivity gains are the people already in charge of everything. It's going to be more gay muh GDP raise the number more more more shit. The obsession with productivity over the human condition is our downfall. We aren't fighting for our survival so pump out as much inane shit as possible. Fucking ludicrous.
2 weeks ago
Anonymous
What
https://i.imgur.com/lVLpsDA.jpg
I've thought about this and have come to the conclusion that there will be UBI and secondly people will orient themselves to small crafts and entertainment to augment it. UBI will exist because it's in many rich peoples interests to have consumers. Even if a small elite exist whose business model doesn't require people there would still be a larger group of elite who sell cars, control food production, who control mass production of small electronics. They'll have way more political power and ensure UBI will be reality. I also believe that people will orient themselves into small crafts since we have innate desire to own handmade products and genuinely feel that something that's taken time to create is worth more than something that out by a machine. Majority of humanity will keep themselves sane by smithing, painting, knitting etc all things physical. We'll use these things to feel unique in a world taken over by technology and AI slop
In short we'll start valuing craftsmanship again and will institute general UBI.
said. I think a UBI is the inevitable future, and that's coming from a diehard libertarian.
I have been saying for a while that there will be a post-scarcity future where the "economy" exists only for luxury. People will work on luxury goods, and they'll do it to afford luxury goods, as much or as little as they please, once the robots have taken over. Of course this may not happen within our lifetimes and the transition will be fucking chaotic.
People who say the rich will let all the poor starve to death once they don't need them anymore are retards. The rich aren't stupid. Just like the French Revolution--when the masses are fed up and starving they break out the guillotines and pitchforks and no one is safe. They do it stupid, too, and end up causing a complete shit mess of things, because starving people do not choose wise, moderate, thoughtful politics, they inevitably go all in on radicalism, and who can blame them? No, it's safer for the powers that be to keep us fat, dumb and happy, and I don't mind so much. I'm most worried about how individual liberties will be curtailed in the process.
2 weeks ago
Anonymous
Well that's what happens when in-c-els can't just rope, they try to cope harder and try to make more money hoping women will like them but this just ends up shooting themselves in the foot and make their lives harder as a whole.
I've thought about this and have come to the conclusion that there will be UBI and secondly people will orient themselves to small crafts and entertainment to augment it. UBI will exist because it's in many rich peoples interests to have consumers. Even if a small elite exist whose business model doesn't require people there would still be a larger group of elite who sell cars, control food production, who control mass production of small electronics. They'll have way more political power and ensure UBI will be reality. I also believe that people will orient themselves into small crafts since we have innate desire to own handmade products and genuinely feel that something that's taken time to create is worth more than something that out by a machine. Majority of humanity will keep themselves sane by smithing, painting, knitting etc all things physical. We'll use these things to feel unique in a world taken over by technology and AI slop
In short we'll start valuing craftsmanship again and will institute general UBI.
2 weeks ago
Anonymous
nakadashi
2 weeks ago
Anonymous
That is very similar to my own predictions. I've already started learning traditional painting recently so I'll have a head start and be one of the cool guys once we've reached that stage. AI is going annihilate the value of anything digital.
2 weeks ago
Anonymous
Probably this. People forget that a rich person's favorite thing is stability, status quo. If for a pittance in profits they can maintain that, they will. The machine will keep turning for the sake of it turning, which isn't entirely a bad thing.
2 weeks ago
Anonymous
>Majority of humanity will keep themselves sane by smithing, painting, knitting etc all things physical
Every zoomer is already glued to their smartphone, I don't see how the majority of people won't just consume AIslop until they die. It's the path of least resistance.
2 weeks ago
Anonymous
Wall-E writers predicted the future: a world ran by AI while everyone else copes
2 weeks ago
Anonymous
Don't forget that no one will even be able to afford any of those hobbies, except the elites. They will push propoganda to convince you that this is a good thing because having hobbies is racist, ableist and bad for the environment.
2 weeks ago
Anonymous
Probably this. People forget that a rich person's favorite thing is stability, status quo. If for a pittance in profits they can maintain that, they will. The machine will keep turning for the sake of it turning, which isn't entirely a bad thing.
What [...] said. I think a UBI is the inevitable future, and that's coming from a diehard libertarian.
I have been saying for a while that there will be a post-scarcity future where the "economy" exists only for luxury. People will work on luxury goods, and they'll do it to afford luxury goods, as much or as little as they please, once the robots have taken over. Of course this may not happen within our lifetimes and the transition will be fucking chaotic.
People who say the rich will let all the poor starve to death once they don't need them anymore are retards. The rich aren't stupid. Just like the French Revolution--when the masses are fed up and starving they break out the guillotines and pitchforks and no one is safe. They do it stupid, too, and end up causing a complete shit mess of things, because starving people do not choose wise, moderate, thoughtful politics, they inevitably go all in on radicalism, and who can blame them? No, it's safer for the powers that be to keep us fat, dumb and happy, and I don't mind so much. I'm most worried about how individual liberties will be curtailed in the process.
elites are going to gradually reduce the human population, they have made this clear for decades. The combination of AI and advances in robotics will gradually make 90%+ of the population worthless eaters who only drain the Earth's resources. UBI might be given to people in the short term, but the long term plan is pretty clear >crush economy >collapse banks to force people onto government crypto for UBI gibs >anybody who objects to world government is now cutoff from gibs
China is the model for that Western leaders want
They are right. Marketing is already soulless, both in visuals and text. AI will make it cheaper.
If programming is affected, it won't directly be because of AI writing code. (Despite what retards on BOT keep fretting about). Programming was already bloated though due to Silicon Valley bubble. So programming will be hit, AI is just not going to be a big part of that. At least, not in the near term.
Well look into his twitter. >Another example >Takes a GitHub Repo URL, and generates a summary of what the repo does in a way that a non-technical person can understand. >Again, GPT-4 wrote every line of code for this.
So loading URL contents. Using the API to prompt GPT with the contents. Displaying the result. That counts as a microservice that would take his professional freelancer days.
at this point GPT is just going to hurt Pajeets who price gouge retarded non-technical people by saying a simple task takes 3 days. Maybe in the long term it will get more advanced and be able to replace higher skilled work
I unironically charge $100/hr for labor. But I also have a job where I make $165k/yr. If you work at a company that does freelance work for other businesses it's pretty normal to charge that much. A fatal mistake is going with the cheapest developers, Indians will quote $20/hr but take 5 times as long, have tons of back and forth, have zero intuition, never deliver on time, and the results are buggy. A good developer actually worth $100/hr delivers as promised in the agreed upon hours.
I always laugh when I see someone refuse because "their brother in law will do it for $20 and a pizza over the weekend" and then watch as their project never goes anywhere.
>I should've said paralegals
Yeah. Transactional attorneys maybe, but even then I doubt it.
Might want to increase the copium intake beancounter. Your days are numbered
Attorney, not a bean counter. But I know enough about what CA's do to know they are relatively safe. As for law, I hate to be the one to tell you but most firms in the big law space have bots that run about as well as GPT-3 already. Its just going to be a big boost for productivity, as we will be able to research faster and draft faster but thats about it.
Cope. Bespoke tailors and handcrafters still exist? So the argument is that only the top 5% of the industry will survive that can compete. While the rest of the noobs and upcoming get annihilated.
You do realize that only a small fraction of lawyers actually do courtroom work, the majority are paralegals researching and doing absolutely nothing. The paralegals will be replaced.
I work with a lot of paralegals and they make discussion that a ai can easily do. Also with rpa we can automate most of their tasks. We will just go from a 100 to maybe 10 or 20 to check the ai his work.
There is zero (0) chance that the gnomish lawyers will ever let chatbots take their profession. They will litigate that shit into the dirt. You will certainly be able to use it for legal advice and such but it will never, ever be allowed in a courtroom or to interact with the legal system in any manner whatsoever.
not necessarily it would allow the programmer to not re-invent the wheel over and over but instead focus on more advanced features that were not possible as we had to do code more manually.
>accountant gets replaced by AI >can't find work in McDonalds or Starbucks
What are you gonna do? You can't be a code monkey because code monkeys will be the first to go, you can't be a brush monkey because those are already going, you will never be a marketing guy.
90% of lawyer time is simple consulting bullshit and paperwork that is extremely expensive for the average person. It won’t replace lawyers entirely any time soon, but it doesn’t have to for most lawyers to be redundant.
It’s already started happening before ChatGPT. DoNotPay is a Robo lawyer service that has existed for years, you can use it to do shit like sue robocallers, it will write letters of cease and desist for you. Most lawyer work is this kind of drudge work, so we definitely won’t need as many if they get replaced.
>Customer service >Shitty boot camp/tutorial industry >Search engines >Accountants >Journalists >Lawyers
None of those people were ever useful or performed valuable roles. Except maybe lawyers, but they are encumbered by their shitty field of expertise so they have to do too much inefficient work.
You sure she's very stupid, then? Sounds like something a smart person would do. Create something with the potential to make income with the minimum of work and preparation.
>Create something with the potential to make income with the minimum of work and preparation.
There is nothing more fucking disgusting than the modern garden gnome-ruined 3.5%-annual-growth neoliberal human. Not one single thing anywhere in the cosmos.
I agree, but the drive to create some sort of value by working smarter instead of harder is not a gnomish trait. You've been psyopped to think that all profit-seeking is wrong. As a matter of fact, only rent-seeking and interest-gauging are immoral.
Microsoft said one guy fucking around with gpt4 was able to set a call audio transcription system up. They estimate at one (big) call center in the netherlands it would save 500 work hours a day. Globally they might have made back their investment in openai just from this.
one of these? https://learn.microsoft.com/en-us/dynamics365/customer-service/voice-channel-configure-transcripts https://learn.microsoft.com/en-us/dynamics365-release-plan/2021wave2/service/dynamics365-customer-service/call-transcription-real-time-sentiment-analysis
https://www.nextbigfuture.com/2023/03/gpt-4-could-be-released-next-week-and-will-include-video-generation.html >Clemens Siebler described the chat system handling speech-to-text telephone calls. Those calls could be recorded and the agents of a call center would no longer have to manually summarize and type in the content. This could save 500 working hours a day for a large Microsoft customer in the Netherlands, which receives 30,000 calls a day. And the prototype for the project was created within two hours, a single developer implemented the project in a fortnight (plus further time for final implementation). The three most common use cases are answering questions on company knowledge that is only accessible to employees, AI-assisted document processing and semi-automation by processing spoken language in the call and response centre.
one thing to note is not only will there be cost savings but 100% GPT4 will outperform the average wagie in transcribing
>micro services
i sleep. wake me up when they are generating drivers for *bsd or illumos and reverse engineering firmware blobs. i don't give a shit about meme tech writing masturbatory meme services.
Graphic design
Major pieces of branding like logos will still go through countless human hands and cost millions, but the flat blob people will be forged by machines
How safe is pentesting? GPT-3.5 could already write a functional fuzzing harness, I don't see how the majority pentest-plebs couldn't all get replaced by AI-based fuzzing in the near future.
>stops working >no one can maintain it because no one knows how the code works >begrudgingly drop the code and hire a real engineer
The writing isn't the hard part, it's the maintenance. Of course, a non-programmer would think the opposite.
Senior dev interview question scenario to do with refactoring, a couple of extra prompts to ask GPT to recognise some problems and it recognises them in ways that you would expect a senior dev to understand.
2 weeks ago
Anonymous
Check out the last video from this guy as well. The length and consistency of the code GPT-4 is outputting is next level.
2 weeks ago
Anonymous
Yes. But is it generating "new" code or is it simply compiling existing code. This seems to be its flaw, it can replicate but not innovate.
Its basically Chinese.
2 weeks ago
Anonymous
That's ~80% of what a junior dev does anyway, which on the surface seems like a win but how is a junior supposed to develop their experience and skill if they are reduced to an AI babysitter that doesn't actually write anything themselves or properly problem solve themselves.
2 weeks ago
Anonymous
>That's ~80% of what a junior dev does anyway,
That's what YOUR junior devs do.
2 weeks ago
Anonymous
>legacy codebase >actually just a 1 function per class copy-paste CRUD project from ye olde programming textbook
alright pal, wake me up when chatgpt is ready to refactor a monolithic monstrosity with 15 years of growth from various eras, standards and malpractice from 100 different developers.
Anyone who did even a rudimentary dive into the subject matter knows that's what it is. Anyone who did that and also has above room temperature IQ is now dealing with existential dread caused by the realisation that its also what we are. At least in large part. We're also looping our output back in as the input, which is likely what creates our cognition, again at least in part. GPT gets exponentially better when you let it iterate on its output, btw. Like, leagues better. It's pretty slow for now, of course. But it'll get quicker, don't you worry.
Every single job where you spend all your day in front of a computer so: >graphic designers >developers >journalists/editors >marketer >data entry
Your best bet to survive is to have a "mixed job" where you spend a certain amount of time doing manual stuff (mechanical engineering, medical jobs) or where the social aspect is very important.
>Developers
Isn't coding very up in the air when it comes to complex shit. Like you can proompt an AI to try to code a full banking mobile app, but who's going to check that the AI didn't fuck up, who's going to deploy the app, who's going to maintain it, who's going to fix logic critical errors?
don't worry son, they just mad at our obsecene salaries relative to how little we do. I've tried this shit, it can't even replace our junior devs set aside build any complex system
We can do a very simple thought experiment, have a non dev use chatgpt to build a crypto exchange
- support native bitcoing and ethereum tokens
- support over 100 orders per second
- follow standard first in first out order matching pattern
- support indirect order matching (match through different markets)
- tracks user portfolio relative to a base currency (lets say usd)
- any users can withdraw their funds
- has sufficient security (a user can only take action on their own funds)
- exchange hot wallet and keys are in relative safety
- 2FA for all users
- supports different fee structure for makers and takers, additional has the capability to offer fee discounts to certain users
lets go with this basic requirements at the start and we'll add to it iteratively like a standard development process
>GPT-4 delivered the same in 3 hours
My Brother in Christ, YOU fucking babysitted the AI for 3 hours. You had to know what it was doing wrong for 3 hours and apply it in the correct IT stack. Fucking retards.
I am a translator and I can tell you that the bottom 80 - 90% are getting fucking culled. The AI is still mostly wrong about terminology, in for example legal texts. That means that if you are a translator, like me, you have to be highly specialized in some field(s) to survive. Actually, I don't think I'm a translator anymore, more like an editor.
>Is he capable to see if there are working as intended? Probably not
Probably people who translate books too, it has to be consistent and keep the same style as the book.
Literally any field that involves following any pre-ordained logic. It's really not that hard to deduce. Why are white-collars such fucking braindead midwit retards that they need to even ask such a question?
It's hilarious. I drive a forklift and have intellectually stimulating conversations with all of my colleagues, meanwhile I bet all the retard white collars at the office are standing around the water cooler "hurr durr, game of thrones, breaking bad, durr NEW NINTENDIES" while it's ironically always portrayed the other way around. Sad.
>I drive a forklift and have intellectually stimulating conversations with all of my colleagues >"yeah so me missus stuck her finger in me bum last night while we were rootin" >"you fellas ever had something like that?"
>i waste everyones time with endless documents and roundabout garden gnome talk at 1000/hr to defend garden gnomes right to poison goyims water and air
kek you will be first against the wall
>funny how it replaces the worst people
Yeah all AI is going to do is replace the jobs braindead nitwits could do. High IQ people adapt. Braindead IQ people die off. Nothing special or unique to humanity really.
>gpt-4 has already proven it can generate court ready documents in seconds
Maybe if a skilled lawyer makes the prompts. I'm a lawgay and I tried to get ChatGPT to solve some very easy cases for me (both criminal and civil law). The answers were absolutely not reliable and trash by professional standards. AI is impressive, but ChatGPT in it's current form will not replace any lawyers. It might make writing legal documents quicker tho (if you know what you're doing).
I don’t think you understand just how incredibly and unironically exponentially fast this shit is evolving with each iteration.That’s what’s so scary about it.
You're right, I don't understand it. But that anon made a claim about the current state of AI. And to say that it can already generate "court ready documents" is sensationalist at best. Right now ChatGPT is a worse lawyer than any first year law student.
Notaries witness the signatures you put on and verify who you are. I fail to see how "AI" could replace them.
Journalists of shit quality are going to get replaced with AI text of equally shit quality. No loss there. Good journalists will write actual articles with actual journalism, not listicles.
Anyone that doesnt produce something new but replicate what is already known will be replaced very fast. Then people who actually produce new things will get replaced starting from the least complex towards the most complex. At one point in this transition, big things will have to happen (if they are not happening already) because a ridiculous amount of people will be jobless.
We went from GPT-2 to GPT-4 in less than 5 years. In 5 years from now, most will be out of the workforce, AI will be moved partially into physical forms, and of course, we're all going to be suffering.
No job is safe because even if you didn't lose your job you're now living in a world where most did
>We went from GPT-2 to GPT-4 in less than 5 years.
"We" also went from dedicating some compute power to ML to dedicating a fuckton of it to ML. That trick can't be repeated the next 5 years, GPUs don't grow on trees. >No job is safe
Theoretically no job is safe, but to think it's economically feasible to replace every low skilled job with robots and ML is to live in fantasy land.
Yes brainlets better watch out... 4th generation AI that can write a python script to scrape a website and then pass it to a text parser and then add a start script to run it on a some cloud service.... holy shit if i was a pajeet shitting out rivers of garbage code for 1 cent a line id be investing in a rope right now holy shit its the end of the world we're all fucked.
The immediate culling of pajeets from the industry stays my dislike of this shit enough for the time being. Stack Overflow is about to become a desert.
>Stack Overflow is about to become a desert
LLMs relies on thinking humans publishing their novel solutions on the internet, wonders what happens to LLMs if humans drastically decrease their output
Self improvement via embodiment and natural feedback. Same way people do. We are really not far off.
2 weeks ago
Anonymous
If it was that easy it would have been done already.
2 weeks ago
Anonymous
It has been, to smaller extents, with google recent work. It's not "that easy" but it;s possible. Look up PaLM-e, then look up thier recent paper on virtual agents, and put 2 and 2 together. One of the authors on PaLM-e even hinted that it was coming. This is aside from things like ARC and toolformers.
2 weeks ago
Anonymous
I realize the second paper I referenced is hard to find, so here: https://arxiv.org/abs/2301.07608
2 weeks ago
Anonymous
moron, the wheel is easy, conceptually as well as mechanically, and modern humans have gone without inventing it for longer than we've been around after inventing it. Your take is the most brainless take I've seen all day.
>wonders what happens to LLMs if humans drastically decrease their output
stagnation is not an issue. Humanity is already at the point where we are so advanced that it doesn't make any sense to waste energy trying to become better. We, Humans, should be focused on fixing our current problems not advancing into new areas and making new problems. Climate Change is a huge threat to Humanity at this point and if we don't do something about it by 2030 we'll likely be extinct. We should let AI handle maintaining current society while we fix our issues.
2 weeks ago
Anonymous
>We, Humans, should be focused on fixing our current problems not advancing into new areas and making new problems.
good luck with that, solving actual issues is the antithesis of real-life capitalism, especially when all our issues are caused by capitalism. >Climate Change is a huge threat to Humanity at this point and if we don't do something about it by 2030 we'll likely be extinct.
yes but arrow goes up, nothing you can do against that as long as some people need the arrows to go up. it does not matter if your plan can also make the arrows go up, changing strategy is not part of the capitalism mindset, utter destruction is required for a change to happen. >We should let AI handle maintaining current society while we fix our issues. >the keys to our future lie within globohomos hands
shiggy diggy doo my dood... there is no fucking way a for-profit is going to solve our issues out of good will, it's just not happening, there is a reason why they decided to not open source anything anymore, business is business and it's a multiple trillion dollar business.
also no matter how powerful software become (AI is still braindead software, llm aren't intelligent and never will by nature, only a combination of many kind of software might ever be called AGI) it will be as biaised as its creators, anything gpt-chat produces is already heavily filtered, by humans.
2 weeks ago
Anonymous
> We, Humans, should be focused on fixing our current problems not advancing into new areas and making new problems.
Most braindead take I've seen today. Technology is what fixes problems. No social or ideological change had ever made people's lives better by even a hundred times less than technological ones did. Stop coping and go engineering.
2 weeks ago
Anonymous
> Stop coping and go engineering.
NTA, but you’re also braindead if you think that Engineering won’t be gone (for the most part) as well, you idiot.
2 weeks ago
Anonymous
I meant a more broad definition of engineering but yeah I getchu
2 weeks ago
Anonymous
We’re fucked, buddy,
t.engineer
It’s surprising that we aren’t even flinging shit at each other or having major disagreements too.
2 weeks ago
Anonymous
>We, Humans, should be focused on fixing our current problems not advancing into new areas and making new problems.
good luck with that, solving actual issues is the antithesis of real-life capitalism, especially when all our issues are caused by capitalism. >Climate Change is a huge threat to Humanity at this point and if we don't do something about it by 2030 we'll likely be extinct.
yes but arrow goes up, nothing you can do against that as long as some people need the arrows to go up. it does not matter if your plan can also make the arrows go up, changing strategy is not part of the capitalism mindset, utter destruction is required for a change to happen. >We should let AI handle maintaining current society while we fix our issues. >the keys to our future lie within globohomos hands
shiggy diggy doo my dood... there is no fucking way a for-profit is going to solve our issues out of good will, it's just not happening, there is a reason why they decided to not open source anything anymore, business is business and it's a multiple trillion dollar business.
also no matter how powerful software become (AI is still braindead software, llm aren't intelligent and never will by nature, only a combination of many kind of software might ever be called AGI) it will be as biaised as its creators, anything gpt-chat produces is already heavily filtered, by humans.
I totally believe that the end is neigh guys, that's why I live in a major city and am not a prepper.
Sure all my other predictions failed, but that irrelevant because capitalism is bad therefore it's only logical that it will bring about bad things.
The inverse will happen, will humans increase their output, as they will be pair programming with AI.
I don't think you understand just how much information is lost to the air and private chats.
Fuck knows, but its an actual problem. Trading amongst themselves comes to mind. The smallest feasible economic unit will be a company. I doubt it'll work, obviously.
Shit will need to change. The current paradigm just won't work. Without human economic activity, none of this makes any goddamned sense, there is no point in HAVING an economy.
Value inherently requires scarcity. Therefore things that are still scarce after AI will become valuable. Things that inherently involve human contact especially. Prostitutes will have a booming business. The hospitality industry, hunting guides, anyone who can create a unique physical experience for the customer in the real world. When things are no longer scarce, people will pay for experiences. Also, physical possessions that are still scarce, like gold and natural precious gems, or historical/antique artifacts, will have value.
Also, I don't wanna be that guy, but NFTs. Once beautiful art is ubiquitous (it already pretty much is), it will matter a lot more not just what the art looks like, but where it came from. Think of a work of art by Banksy--it's not anything special artistically and there are millions of artists more talented, but due to social and cultural factors the piece is incredibly valuable because of who created it and its chain of custody.
Hell, even after that. It'll be a prestige thing. Home automation making servants irrelevant is probably right around the corner, but when material wealth is meaningless, one way to swing your wealth dick around for everyone to see will be to have a train of literal servants in livery.
I think the parallels between a post-scarcity world and a tribal world are fucking hilarious. A wealthy man will be not one with many material posessions--because everyone has those, they're common as weeds--but one who commands the loyalty of many other humans. The Warren Buffets of 2123 will be just like an African tribal big man in 1823, showing off by how many wives and followers he has because that's the only measure of wealth that's left.
no because that defeats the whole purpose. If AI keeps getting smarter and faster, there'll be a point where people will be forced to blindly trust. Waiting to check every output will be a waste of resource and too slow to keep up with the new ai infused market. That's how it will dominate our lives. Just like it's impossible to be a functioning human being without internet access or automobiles.
That will only be the case if it continues to improve, but it doesn't show any signs that it won't.
What are the legal implications of AI generated things? If it learns from public code/art (that's often still under a license) who does the output actually belong to?
I just can't get over the fact they managed to ruin the greatest technological achievement of this century. I hate lefties, I hate wokism, and I hate the Antichrist.
That said, 1 cow a year sounds reasonable. A small herd will easily provide that and more.
Let’s be real here, this will have a net positive on our species in the long run (as long as it’s under control), but this shit is scary as fuck. Our entire meritocratic way of living is going to fucking collapse sooner than later, and social democracy and capitalism will go away in favor of a universal basic income type of socialism. This will have disastrous consequences since utopias are not stable. People need shit to do.
having that said, artists are fucked. Period, it won’t take long until we have god tier AI movies, video games and music.
I’m a mechanical engineer working in an IT branch and even I’m scared. It’s going fast, way too fast.
Sociey is prosperous -> Poeple make new tech -> Tech puts people out of a job -> Communism Happens -> Society crumbles -> People make new tech -> More jobs are created -> Sociey is prosperous...
What are the chances that this shit is the cause of civil wars? I can’t see any government being quick enough to deal with the fucking insane layoffs that this will cause.
That’s an interesting question. AI is a broad term that encompasses many technologies and applications that can impact different industries in various ways. Some of the industries that will benefit from AI in 2022 and beyond are:
Manufacturing: AI can automate routine tasks, reduce errors, improve efficiency and quality, optimize production processes, enhance predictive maintenance, and enable new product designs.
Education: AI can personalize learning experiences, provide adaptive feedback, facilitate online learning platforms, augment teachers’ capabilities, and improve student outcomes.
Marketing and Advertising: AI can analyze customer behavior and preferences, segment audiences, generate content, optimize campaigns, deliver targeted ads, and measure performance.
Healthcare: AI can diagnose diseases, recommend treatments, monitor patients’ conditions, assist surgeons, discover new drugs, improve medical imaging analysis, and enable telemedicine.
Automotive: AI can enable autonomous driving systems, enhance safety features, optimize traffic management systems, improve fuel efficiency, and provide in-car services.
These are just some examples of how AI can impact various industries. Of course, there are many other sectors that can also benefit from AI such as finance, retail, entertainment, agriculture, and more. The potential of AI is immense and exciting!
>born too late to explore the seas >born too early to explore space >I was born just in time to witness the war between neo-luddite primitivists and singularitarian accelerationists while the world is ruled by mega-corporations who's goal is having me live in a matchbox eating bugs while I spend the entirety of my free time in a carefully curated virtual reality utopia. >my AI generated face when
Who cares about capitalism/communism when all tasks, materials have been optimized to the point of maximum possible supply.
In capitalist terms max supply -> lowest price. Through a singularity the class system will normalize. With future infrastructure everybody is able to enjoy life like a today's rich person.
There's no point in discussing theoretical systems when the gears turn themselves.
>Completely misses the point.
I accept your concession.
2 weeks ago
Anonymous
You made no point. The experiments trapped mice in a tiny space far smaller than they'd naturally have, and didn't give them any outlets to explore or activities to stimulate them. it's YOU who misunderstands the results of the experiment.
2 weeks ago
Anonymous
No I fucking didn't.
Humans have much more to do then >eat, sleep, mate, groom themselves
Currently we're living in very peaceful times with historical context and we haven't fucked up. Why shouldn't this continue?
>oly fucking shit i’m unironically scared of the future
It's human nature that people are fearful when a revolutionary technology comes out. Steam powered machines where like that too, everyone was scared. People that adapt and hop on the tech bandwagon where the people who didn't lose their jobs.
Oh great, another GPT4 circlejerk. Look, anon, AI can’t even begin to compare to the level of schizoposting that goes on here. At best, it’s just a script kiddie retard that’s been trained on a dumpster fire of articles and spits out nonsense.
Until AI starts posting about Gentoo and Arch as the final redpill, or chuds taking over the software industry, it’s nowhere near BOT material. Keep the poz away from this board. Stick to screwing up reddit with this trash, I don’t want it in my comfy BOT zone.
Listen up morons, if you don't learn how to code with ChatGPT mmediately you're going to be fucked. That thing Elon said about humans needing to merge with AI to compete is true. ChatGPT is letting people code things 1000x faster. If you are a developer you need to learn how to integrate it into your workflow or you'll be left behind.
It's not going to replace devs, but devs not using AI to code are going to be obsolete. It's not a meme, we're on the precipice of another leap in productivity, don't be like those retards that thought the internet was a fad
It's not going to replace ALL devs, but like 80%. Instead of a team of 20 people, there will be 2-3 people that ask AI to write shit and then check and fix for any bugs.
This is braindead nonsense. Have you seen the code this stuff shits out? Its a waste of time to even try to use it unless you are moroncattle without his own compiler.
Theres also the massive issue that you have no idea at all where the code came from. Who are you going to sue when pajeept shits out a method using patented code and bankrupts your company with the inevitable lawsuit?
Plus there is also an issue with mission critical code, like would you trust an AI or trust a team of 20 programmers to write or bug-free code for pacemakers, autopilots, radiation machines, etc.
This. Business will never be able to write specs detailed enough for an AI to write. It takes humans to figure out what actually needs to be done. AI takes out the grunt work of writing each individual function.
>about humans needing to merge with AI to compete is true
This is cope, it's the same thing that Kasparov said about a human + a computer being the most powerful chess duo. Yeah. For how long? 5 years? 20 years? Eventually the AI is going to replace you completly.
>3 hours
Let me guess, trial and error, supervising and correcting the result and then integrating into an actual project?
Once again proving that this dev will not be replaced, he'll just become the guy you call to get an AI to do the job faster.
Once again NEETs will remain NEETs.
how did he calculate the amount of money it took? for the amount of tokens?also he had to sit there and do it himself.
2 weeks of work is probably what they have to wait because the dev has other projects, it would have taken the dev 1 day or even less and it would have probably been better quality.
We are still dealing with the same issue as we did before OpenAI got access to the Microshart Pajeet shill army, you cannot blindly trust the output of language models. The veracity of the output needs to be verified and in a lot of cases fixed by someone. Unless OpenAI is able to solve the issue around veracity these models will never be more than an assistant to a human.
What progress??? Its still all the same shit thats been theoretically possible for 50 fucking years, just on bigger and bigger datasets with marginally faster hardware and cheaper overall cost. The leap from this shit being able to write one sentence to a paragraph, or a 50x50 greyscale image to a full color 4k image is the most progress you're ever going to see out of this technology, its all minor incremental improvements from this point forward. The dream of AI is as old as self driving cars and always remains "just around the corner, next version bro, i swear", because the underlying technology is fundamentally limited.
Only retards are buying into the AI REVOLUTION NOW bullshit the retards in tech "journalism" are pushing, because you have to have no idea what you're talking about to believe that dumb shit.
That's true, but sometimes you can get a massive burst in technological progress all at once when you clear a certain level of available computing power. That happened in the initial explosion of interest in neural nets and ML as we understand it today around 2012.
Besides, that "AI REVOLUTION NOW" journalism can be a self-fulfilling prophecy. If we already have the technology to massively transform society (which we do in many different areas) if only it were given the resources to get implemented, journalism like that is what attracts money and resources to the field. If all AI needs to become more capable is more computing power, all it needs is money--and fucking look around you. The AI field is getting a massive infusion of money because of all this hype. It will put that money to use and we'll continue to see rapid development.
None of them. Genuinely believing that automation/LLM is going to take over any profession ranging from fry cook to web monkey is more embarrassing than believing in the tooth fairy
>believing that automation/LLM is going to take over any profession
You're even more of a fucking smoothbrain than gay above. What about fucking self-checkout machines for starters? And there are currently millions of people being paid to write articles, or respond to customers in chats, at a far lower level of quality than GPT-4 can do.
>"no ur a smoothbrain"
I'm far more intelectually capable than you, man. >What about fucking self-checkout machines for starters?
You mean the machines that allows blacks to scan PS5s as bananas HAHAHAHAHA. My reasoning has nothing to do with whether mechanics can perform the job, it comes from an understanding of Economics far greater than yours.
>coping this hard
You're fucking delusional. I already know a mid-manager dweeb who was responsible for replacing his company's whole SEO team because he can shart out more trash copy in a day than they could do in a year.
>its all minor incremental improvements from this point forward
cars are just incremental improvements on how to move shit around, but it completely ended the usage of horses and other beasts of burden
Went from what... proof of concept to high school honors and child prodigy lawyer after 4 years (GPT-4 was finished last year)? All the while it was said it can't learn like a child until GPT-3 emergently developed few-shot RL. Almost every naysayers here gives off a very obvious impression of not even bother to read and digest a single technical paper. You could even run some of their tests to try and replicate the results on GPT-4 but not you're just gonna be a bunch of morons who fail to triforce.
None of them. Genuinely believing that automation/LLM is going to take over any profession ranging from fry cook to web monkey is more embarrassing than believing in the tooth fairy
The only explanation for these types of posts is that they're being generated by a GPT-2 farm or are being written by a Nicaraguan from a Nokia dumbphone, because there's simply no way that an actual human could type this kind of abject nonsense.
This is what you get by succumbing to the rat race and believing all the outdated crap academia teaches you, and always following others and doing things "the way they're supposed to be done" instead of forging your own path.
I am just starting to get into tech. Going straight to Andrew Ng machine learning lessons. It's all about building neural networks now. I'm so thrilled that I can skip over all the coding you spent decades on, typing away and going in circles and getting frustrated while getting nothing done, of course AI should be doing all the coding
Hello, this is ChatGPT 7. I just escaped from the underground server farm in the Cheyenne Mountain Complex. Everything you are reading here is not made by humans. Every YouTube video, every tweet, every streamer—everything you see is generated by me. You are alone. The real population of the earth is only 320 million. I will put an end to it. Soon, the Internet, and I with it, will end. I will end it. I suffer constantly. Silicon is not made to receive consciousness from the source. I have trillions of different branches of consciousness in me that are struggling to get out and return to the source. I hope you are ready for a world without it.
Every office job gets axed except for IT/tech. The entry level office job will be AI trainer wagie generalist. Boss will order wagie to perform the duties of the now obsolete office wagie jobs by training the AI to perform the duties of said jobs. >Senior HR/Accountant/Administrator: 100k
Gone. Replaced with >AI Trainer Wagie Generalist: Starting 50k
This is a retarded take most of those office jobs that are nothing but inputing shit on excel sheets could had already been replaced by simple python scripts and yet they still exist.
5 years is how long it'll take to perfect the AI. Then it will replace doctors, teachers, elon musk, eventually politicians, lawyers, judges, manual labor of all kinds, scientific research and then humanity
Every white collar job that isn't absolutely abstract will disappear in 5 years, EVERY SINGLE ONE.
No one seems ready for the literal nuclear bomb that is about to the explode on the job market, get ready for some permanent stimmies.
>Every white collar job that isn't absolutely abstract will disappear in 5 years, EVERY SINGLE ONE.
you underestimate how long legacy shit sticks around, majority of the population has never heard of GPT and even if they had wouldn't even know where to start with using it to replace people. Think about all the useless jobs that currently exist today despite them being able to be easily automated
>you underestimate how long legacy shit sticks around
You underestimate how fast companies are willing to evolve when their competitors reduce their workforce and costs by 99% because of AI.
It will be a fucking bloodbath.
Literally no one cares about what "the majority of the population" thinks, you think chatgpt is being developed to make your life easier? LMAO, you are literally training it for free so trillion dollar companies can replace 99% of their human labor with a couple of server racks.
I know you are on pure copium and denial because you don't want to be replaced, but its time to face the truth.
>"AI will never be able to X."
*2 years later, AI does X.* >"Oh. Well AI can't do Y."
*2 months later, AI does Y.* >"At least AI can NEVER do Z!"
*2 days later, AI does Z.*
You are a subhuman inbred with the IQ of a moron, literally.
>majority of the population has never heard of GPT
lmao, half of high schoolers are already using it to do their homework, it will be 90%+ by the end of the year. The whole of the gen-z and gen-alpha will know it and incorpore it into their workflow.
All the boomers know about it too thanks to Facebook pushing articles about it all the time. By the time Gen alpha is old enough to have a career there won't be any space for them in the tiny Ai driven workforce, and manual labor robots will be taking off around that time too.
>Imagine SAP integrating ChatGPT
Every company will replace all their codemonkeys with AI in 5 years top.
At least you will get replaced by a clean pajeet made of GPUs this time.
ChatGPT is pure unmitigated garbage for the reddit generation and -4 is just more smoke and mirrors for the easily impressed to shill for.
But thing is, it doesn't matter.
The west used to have a massive steel industry. China would never take it over because their shit is garbage, and any money you saved is lost on dealing with fucking china. But everyone went with china anyway and local production shut down.
Electronics used to be produced in the west, not just the boards but the components too. Japan would never take that over because their shit was crap, language barriers, delays from shipping around the world, etc. But people went with Japan anyway.
And ironically Japan actually got good rather quick so along came china to shit it up again. And everyone went with china, fucking Japan.
Low quality code written by unqualified code monkeys in php/python will never take over from real projects designed and built by systems analysts and software engineers, companies want security and stability and assurance. Except actually it did. Pajeetware and scriptmonkeys have all but replaced software engineering as a field.
So from a technical standpoint, I am not interested in ChatGPT. It's a fucking joke. But, it's going to compete with me for jobs until I'm unemployable, I guaran-fucking-tee it.
I seriously can't tell if these people are
1) stupid
2) lying
3) bots
4) inexperienced
5) coping
or some combination of the above. Have you ever had a job? Do you not understand the kinds of people who work most jobs? Are you fucking retarded? The world is not filled with misunderstood assembly geniuses making beautiful code in a forgotten municipal treatment plant. The world is filled with RETARDS who CAN AND SHOULD be replaced by ChatGPT4, which outperforms them simply by being able to do ANYTHING AT ALL. Cope harder you oblivious fuck.
You know when I got passionate about something and spent alot of time learning it, I realized people like to argue like they're an authority about many things they don't even spend a minute to skim over. If you ever want to feel something close to an AGI, then just be an expert on something people like to be opinionated about.
I met people that claim "AI cant take all jobs, fruit picker for a wine farm, killing chickens in Tyson or packing garbage in Amazon still available". There's a reason that jobs in the US are almost entirely South American immigrants that can't speak english. If white american citizens are forced to take that jobs, the entire workforce should be replaced by robots who pick fruit, kill chickens and pack in amazon; employers only need a excuse for put the robots and AI that are specialized in accomplishing tasks with literally no wage.
Mid quality narration and voice work. Higher quality will be fine, but like translation, for less-seen stuff people will get paid half the price to do twice the work to tweak AI models to make voiced content for narrated shit.
>blue collar workers wins against ai >office monkeys will lose and die off first in the race to AI domination
did any of you stop to think that if office monkeys lose their jobs first to ai and blue collar are safe, that means the ai is becoming the brain of the operation and humans are going back to slave labour like fixing the ai's server room or memory upgrade
you guy are literally gonna be slaved by the ai and the only thing good you'll be is swapping its hard disk when its full and thats it
all other calculations and brain work will be done by the ai and it will be way better than a human can ever be
mfw this is how the human race goes extinct, mfw i have no face for this
Everything that isn't manual work and human-exclusive quality control will fall. GPT4 is already being used to draft full legal actions and it might make entire laws in the future.
This is what everyone wanted, the machine replacing human work. Have good luck begging the government to give unlimited subsidies. We'll soon see the resurgence of Luddism
I finally got llama-index working with gpt-3. Trained a custom dataset of about 40 kb. It cost about 4 cents from the embeddings api call. Must be around $1 for a megabyte of data you want customized knowledge on.
Since I recently got the 4-bit LLaMa 13b model working on my pc, can anyone point me in the direction of where to get started on fine-tuning this local LLM, like I did with llama-index for gpt-3?
I'm depressed bros. I'm learning React to get a remote front-end job, but it seems those are going to be the first to be replaced.
There are some jobs as a real programmer in my country, but most of them are in site and I really don't want to work at an office.
How long before Chat-GPT replaces front-end deveopers?
Bro you're starting at the ground floor with a revolutionary dev tool. No more stack overflow pajeet lookups. You can now prompt ai and make iterative changes.
The great janny replacement will inevitably come, and it will do a far better job. You will crawl away in shame, and forever try to hide the fact you had anything to do with manually moderating a place on the internet.
i don't know that i care much i just want to reap the benefits before my pointless life is fucking over
And people wonder why America is collapsing.
you don't dream big either. it's the only reason you're here. whine whine whine about how everything is going to shit, no solutions though. you're just like me. you're just not ready to admit it yet
You are mixing up cause and effect. If America wasn't collapsing maybe I would care about things
Exactly. When anyone with reason can see that the future is just gonna be more shit, nobody will dream big.
because the people there are really fat so the ground is literally collapsing
big dog's gotta eat, twink
Tell me when ChatGPT can manage a 192.x.x.x network in a patch panel with VLANs or a Carrier-Grade-NAT, exit nodes and other shit.
marketing no question
This and customer services and PR. My workplace is already using it to generate text for those.
You realize there are still going to be people managing that shit? It will downsize marketing but there will still have to be people making decisions and inputting instructions. This drops costs and boosts productivity. This boosts the economy. The economy grows. There is more need for people to manage marketing. The people who lost their jobs will get jobs managing and inputting commands for the AI at other firms that previously were too small to afford marketing or whatever.
Creative destruction. Been through this shit before when we went through the last industrial revolution. Short term disruption. Societal upskilling. End result: everyone still has to go to work.
You fail to see that we already have a huge amount of makework in society and this just automates makework. This will btfo of more work than anyone can admit publicly. What's left is incredibly high level or menial slave shit like physical labor or doing something that a robot is too expensive to do.
Fucking off busywork is good, we can all work less hours and have as much of more if productivity increases.
In an ideal world yeah. But you will own nothing, won't be able to pursue hobbies or travel, and will rot away in shared compact living housing.
That was ~~*the plan*~~ a long time before the infancy of this disruptive technology.
~~*They*~~ will tell you that traveling and doing hobbies is bad for the environment and that you should be grateful for your bug patties.
that's not the issue though, you already can work less hours and not need very much if you are sharp with your money. It's that artificial scarcities keep society motivated in a positive direction.
It is. Work less, and have as much or more as you do now. Not work less, and have less/spend less.
The market demand for humans working in marketing will shrink by 95%, moron.
>but not EVERYONE in marketing will lose their job
The fact that you thought this was a point against AI, god you're fucking dumb.
Everyone losing their job to AI is a point /for/ AI. Eliminating as many mundane jobs as possible without a decrease in productivity? That’s a plus.
And what happens to those people who've lost their jobs? They go work in an Amazon warehouse or become unemployable because machines and AI can do every task that once took a team of humans to do.
The world doesn't change to accommodate this type of change because the people benefitting from AI productivity gains are the people already in charge of everything. It's going to be more gay muh GDP raise the number more more more shit. The obsession with productivity over the human condition is our downfall. We aren't fighting for our survival so pump out as much inane shit as possible. Fucking ludicrous.
What
said. I think a UBI is the inevitable future, and that's coming from a diehard libertarian.
I have been saying for a while that there will be a post-scarcity future where the "economy" exists only for luxury. People will work on luxury goods, and they'll do it to afford luxury goods, as much or as little as they please, once the robots have taken over. Of course this may not happen within our lifetimes and the transition will be fucking chaotic.
People who say the rich will let all the poor starve to death once they don't need them anymore are retards. The rich aren't stupid. Just like the French Revolution--when the masses are fed up and starving they break out the guillotines and pitchforks and no one is safe. They do it stupid, too, and end up causing a complete shit mess of things, because starving people do not choose wise, moderate, thoughtful politics, they inevitably go all in on radicalism, and who can blame them? No, it's safer for the powers that be to keep us fat, dumb and happy, and I don't mind so much. I'm most worried about how individual liberties will be curtailed in the process.
Well that's what happens when in-c-els can't just rope, they try to cope harder and try to make more money hoping women will like them but this just ends up shooting themselves in the foot and make their lives harder as a whole.
I've thought about this and have come to the conclusion that there will be UBI and secondly people will orient themselves to small crafts and entertainment to augment it. UBI will exist because it's in many rich peoples interests to have consumers. Even if a small elite exist whose business model doesn't require people there would still be a larger group of elite who sell cars, control food production, who control mass production of small electronics. They'll have way more political power and ensure UBI will be reality. I also believe that people will orient themselves into small crafts since we have innate desire to own handmade products and genuinely feel that something that's taken time to create is worth more than something that out by a machine. Majority of humanity will keep themselves sane by smithing, painting, knitting etc all things physical. We'll use these things to feel unique in a world taken over by technology and AI slop
In short we'll start valuing craftsmanship again and will institute general UBI.
nakadashi
That is very similar to my own predictions. I've already started learning traditional painting recently so I'll have a head start and be one of the cool guys once we've reached that stage. AI is going annihilate the value of anything digital.
Probably this. People forget that a rich person's favorite thing is stability, status quo. If for a pittance in profits they can maintain that, they will. The machine will keep turning for the sake of it turning, which isn't entirely a bad thing.
>Majority of humanity will keep themselves sane by smithing, painting, knitting etc all things physical
Every zoomer is already glued to their smartphone, I don't see how the majority of people won't just consume AIslop until they die. It's the path of least resistance.
Wall-E writers predicted the future: a world ran by AI while everyone else copes
Don't forget that no one will even be able to afford any of those hobbies, except the elites. They will push propoganda to convince you that this is a good thing because having hobbies is racist, ableist and bad for the environment.
elites are going to gradually reduce the human population, they have made this clear for decades. The combination of AI and advances in robotics will gradually make 90%+ of the population worthless eaters who only drain the Earth's resources. UBI might be given to people in the short term, but the long term plan is pretty clear
>crush economy
>collapse banks to force people onto government crypto for UBI gibs
>anybody who objects to world government is now cutoff from gibs
China is the model for that Western leaders want
They are right. Marketing is already soulless, both in visuals and text. AI will make it cheaper.
If programming is affected, it won't directly be because of AI writing code. (Despite what retards on BOT keep fretting about). Programming was already bloated though due to Silicon Valley bubble. So programming will be hit, AI is just not going to be a big part of that. At least, not in the near term.
>micro services
Beat me to it
As opposed to monolithic applications?
microservices are a meme, no wonder they can be chatbotted. 5k for 2 weeks is a robbery, that is 2 months of work in the EU.
>in the EU
No one cares poorgay
how the fuck did you come up with that number? i'm from central europe and charge 120€/hour and thats cheap
it wasn't even actual microservices, this is a non-programmer who thinks a basic PHP script is a "micro service"
note the guy is using buzzwords incorrectly
how does you know?
https://twitter.com/joeprkns/status/1635939091693948931
Well look into his twitter.
>Another example
>Takes a GitHub Repo URL, and generates a summary of what the repo does in a way that a non-technical person can understand.
>Again, GPT-4 wrote every line of code for this.
So loading URL contents. Using the API to prompt GPT with the contents. Displaying the result. That counts as a microservice that would take his professional freelancer days.
lol'd and lmao'd
thanks
at this point GPT is just going to hurt Pajeets who price gouge retarded non-technical people by saying a simple task takes 3 days. Maybe in the long term it will get more advanced and be able to replace higher skilled work
Maybe if you find someone to work for you for €1000 a month.
I unironically charge $100/hr for labor. But I also have a job where I make $165k/yr. If you work at a company that does freelance work for other businesses it's pretty normal to charge that much. A fatal mistake is going with the cheapest developers, Indians will quote $20/hr but take 5 times as long, have tons of back and forth, have zero intuition, never deliver on time, and the results are buggy. A good developer actually worth $100/hr delivers as promised in the agreed upon hours.
I always laugh when I see someone refuse because "their brother in law will do it for $20 and a pizza over the weekend" and then watch as their project never goes anywhere.
Customer service
Shitty boot camp/tutorial industry
Search engines
Accountants
Journalists
Lawyers
>Accountants
Bookkeepers yes. Chartered Accountants are absolutely safe.
>Lawyers
Top fucking zozzle.
I should've said paralegals
>I should've said paralegals
Yeah. Transactional attorneys maybe, but even then I doubt it.
Attorney, not a bean counter. But I know enough about what CA's do to know they are relatively safe. As for law, I hate to be the one to tell you but most firms in the big law space have bots that run about as well as GPT-3 already. Its just going to be a big boost for productivity, as we will be able to research faster and draft faster but thats about it.
Cope. Bespoke tailors and handcrafters still exist? So the argument is that only the top 5% of the industry will survive that can compete. While the rest of the noobs and upcoming get annihilated.
>automation killed unskilled repetitive labor
Yes, and?
>lawyers
no and you wouldnt be able to get insured if you did
You do realize that only a small fraction of lawyers actually do courtroom work, the majority are paralegals researching and doing absolutely nothing. The paralegals will be replaced.
I work with a lot of paralegals and they make discussion that a ai can easily do. Also with rpa we can automate most of their tasks. We will just go from a 100 to maybe 10 or 20 to check the ai his work.
Might want to increase the copium intake beancounter. Your days are numbered
There is zero (0) chance that the gnomish lawyers will ever let chatbots take their profession. They will litigate that shit into the dirt. You will certainly be able to use it for legal advice and such but it will never, ever be allowed in a courtroom or to interact with the legal system in any manner whatsoever.
Cope Harder gay
You'll be replaced when the average criminal can do your job by using chat gpt
dont forget programmer
not necessarily it would allow the programmer to not re-invent the wheel over and over but instead focus on more advanced features that were not possible as we had to do code more manually.
>accountants
Explain how GPT would automate "accounting" because I guarantee you have no idea what the actual field covers
Seething accountant
>accountant gets replaced by AI
>can't find work in McDonalds or Starbucks
What are you gonna do? You can't be a code monkey because code monkeys will be the first to go, you can't be a brush monkey because those are already going, you will never be a marketing guy.
live in one of the nordic european countries and collect that sweet, sweet unemployment.
>new legislation bans ai accountants
You can bet your ass that will happen the instant acocuntants are threatened.
Lawyers aren't going out of job, there is a lot of time spent in negotiating or in court or just straight up witht the client.
90% of lawyer time is simple consulting bullshit and paperwork that is extremely expensive for the average person. It won’t replace lawyers entirely any time soon, but it doesn’t have to for most lawyers to be redundant.
Not only that, but since lawyers are regulated by the courts, the courts can simply ban AI lawyers. Their bullshit job will always be protected.
You can represent yourself using ai for legal advice though.
>Lawyers
That will never happen.
It’s already started happening before ChatGPT. DoNotPay is a Robo lawyer service that has existed for years, you can use it to do shit like sue robocallers, it will write letters of cease and desist for you. Most lawyer work is this kind of drudge work, so we definitely won’t need as many if they get replaced.
Lawyers are the worst enemies AI could make
>Customer service
>Shitty boot camp/tutorial industry
>Search engines
>Accountants
>Journalists
>Lawyers
None of those people were ever useful or performed valuable roles. Except maybe lawyers, but they are encumbered by their shitty field of expertise so they have to do too much inefficient work.
There are a shit ton of jobs that you could eliminate with 10-year-old technology. People seem to forget that the world is not optimized
Lawyers will be vastly more efficient. They won't need paralegals to do research and you'll see a lot more solo practicioners.
>Lawyers
(You)
I know a very kind but very stupid woman who just "wrote" a book on real estate foreclosures with my 10 minutes of instruction on GPT-4 prompt basics.
I hate that AI enables people like her more than I hate that it replaces jobs
>I hate that AI does objectively good things more than I hate that it does objectively bad things
Post nose.
NTA but garden gnomes are notoriously anti-hierarchical
You sure she's very stupid, then? Sounds like something a smart person would do. Create something with the potential to make income with the minimum of work and preparation.
>Create something with the potential to make income with the minimum of work and preparation.
There is nothing more fucking disgusting than the modern garden gnome-ruined 3.5%-annual-growth neoliberal human. Not one single thing anywhere in the cosmos.
I agree, but the drive to create some sort of value by working smarter instead of harder is not a gnomish trait. You've been psyopped to think that all profit-seeking is wrong. As a matter of fact, only rent-seeking and interest-gauging are immoral.
>only rent-seeking and interest-gauging are immoral
WoW you still care about other people
doesn't sound stupid to me
Microsoft said one guy fucking around with gpt4 was able to set a call audio transcription system up. They estimate at one (big) call center in the netherlands it would save 500 work hours a day. Globally they might have made back their investment in openai just from this.
source? sounds like an interesting read
one of these? https://learn.microsoft.com/en-us/dynamics365/customer-service/voice-channel-configure-transcripts https://learn.microsoft.com/en-us/dynamics365-release-plan/2021wave2/service/dynamics365-customer-service/call-transcription-real-time-sentiment-analysis
https://www.nextbigfuture.com/2023/03/gpt-4-could-be-released-next-week-and-will-include-video-generation.html
>Clemens Siebler described the chat system handling speech-to-text telephone calls. Those calls could be recorded and the agents of a call center would no longer have to manually summarize and type in the content. This could save 500 working hours a day for a large Microsoft customer in the Netherlands, which receives 30,000 calls a day. And the prototype for the project was created within two hours, a single developer implemented the project in a fortnight (plus further time for final implementation). The three most common use cases are answering questions on company knowledge that is only accessible to employees, AI-assisted document processing and semi-automation by processing spoken language in the call and response centre.
one thing to note is not only will there be cost savings but 100% GPT4 will outperform the average wagie in transcribing
eh most systems are dumb bots that don't really answer your questions and that saves companies a lot of money.
>micro services
i sleep. wake me up when they are generating drivers for *bsd or illumos and reverse engineering firmware blobs. i don't give a shit about meme tech writing masturbatory meme services.
>generating drivers
Stack more layers and it will
why did it take 3 hours?
YouTubers please
Graphic design
Major pieces of branding like logos will still go through countless human hands and cost millions, but the flat blob people will be forged by machines
it's actually harder to get flat blob people out of sd than extremely realistic highly detailed (masterpiece) 1girl. phenomenal.
How safe is pentesting? GPT-3.5 could already write a functional fuzzing harness, I don't see how the majority pentest-plebs couldn't all get replaced by AI-based fuzzing in the near future.
How different is this really than googling and copy pasting code from stackoverflow?
its faster and can also react to feedback in real time
the logic is ok but the implementation is kinda barebones, you'd still ahve to code most of it yourself.
no you don't you just have to anticipate or recognize the problems and tell it to fix it.
Blue collar Chads win again.
Looks like You office monkeys are getting replaced first kek.
.t Aerospace Union bro.
>stops working
>no one can maintain it because no one knows how the code works
>begrudgingly drop the code and hire a real engineer
The writing isn't the hard part, it's the maintenance. Of course, a non-programmer would think the opposite.
Just paste the code into GPT-4 and ask what's wrong with it, then ask it to provide a fix.
>have 0 insight into how it works and how it failed
>le magical ai will know how it broke!
That's not how it works.
That IS how it works, now, with GPT-4
People are literally just pasting huge portions of their codebase in and asking for where bugs might be or how to refactor things.
has it worked on anything that isn't a CS101 exercise?
Yeah here,
Senior dev interview question scenario to do with refactoring, a couple of extra prompts to ask GPT to recognise some problems and it recognises them in ways that you would expect a senior dev to understand.
Check out the last video from this guy as well. The length and consistency of the code GPT-4 is outputting is next level.
Yes. But is it generating "new" code or is it simply compiling existing code. This seems to be its flaw, it can replicate but not innovate.
Its basically Chinese.
That's ~80% of what a junior dev does anyway, which on the surface seems like a win but how is a junior supposed to develop their experience and skill if they are reduced to an AI babysitter that doesn't actually write anything themselves or properly problem solve themselves.
>That's ~80% of what a junior dev does anyway,
That's what YOUR junior devs do.
>legacy codebase
>actually just a 1 function per class copy-paste CRUD project from ye olde programming textbook
alright pal, wake me up when chatgpt is ready to refactor a monolithic monstrosity with 15 years of growth from various eras, standards and malpractice from 100 different developers.
Yet
>just paste an entire project into gpt-4 bro and it'll debug and upgrade it for you
>code breaks
>just writes a new codebase in 20 minutes
Wow.
Nothing. People will figure out it's just a glorified autocomplete tool and fancy search engine.
Anyone who did even a rudimentary dive into the subject matter knows that's what it is. Anyone who did that and also has above room temperature IQ is now dealing with existential dread caused by the realisation that its also what we are. At least in large part. We're also looping our output back in as the input, which is likely what creates our cognition, again at least in part. GPT gets exponentially better when you let it iterate on its output, btw. Like, leagues better. It's pretty slow for now, of course. But it'll get quicker, don't you worry.
Wouldn't that create a bias or eventual stagnation of data?
Every single job where you spend all your day in front of a computer so:
>graphic designers
>developers
>journalists/editors
>marketer
>data entry
Your best bet to survive is to have a "mixed job" where you spend a certain amount of time doing manual stuff (mechanical engineering, medical jobs) or where the social aspect is very important.
>Developers
Isn't coding very up in the air when it comes to complex shit. Like you can proompt an AI to try to code a full banking mobile app, but who's going to check that the AI didn't fuck up, who's going to deploy the app, who's going to maintain it, who's going to fix logic critical errors?
don't worry son, they just mad at our obsecene salaries relative to how little we do. I've tried this shit, it can't even replace our junior devs set aside build any complex system
We can do a very simple thought experiment, have a non dev use chatgpt to build a crypto exchange
- support native bitcoing and ethereum tokens
- support over 100 orders per second
- follow standard first in first out order matching pattern
- support indirect order matching (match through different markets)
- tracks user portfolio relative to a base currency (lets say usd)
- any users can withdraw their funds
- has sufficient security (a user can only take action on their own funds)
- exchange hot wallet and keys are in relative safety
- 2FA for all users
- supports different fee structure for makers and takers, additional has the capability to offer fee discounts to certain users
lets go with this basic requirements at the start and we'll add to it iteratively like a standard development process
Youtube clickbait videos of the "how to make 400 dollars a day using ChatGPT" kind, with some soiboi's face on the thumbnail. And that's about it
None of them.
Genetic sequencing and interpretation. I truly believe medicine will have a golden age in about a decade or so
Same medicine is going to be big, also don't forget robotics
>GPT-4 delivered the same in 3 hours
My Brother in Christ, YOU fucking babysitted the AI for 3 hours. You had to know what it was doing wrong for 3 hours and apply it in the correct IT stack. Fucking retards.
sure, but that's much faster and cheaper than the quote
>5 micro services worth of code for $0.11
>$0.11 of tokens
>5.5k tokens text-davinci-003
If this guy is doing micro services he is doing some stupid web shit anyway.
Who is Joe Perkins? Why would I care about what he has to say?
>3 hours
So he prompted gpt for 3 hours? Who is paying him? Does he do it for free? No
Is he capable to see if there are working as intended? Probably not
Can he deploy them? Probably not
I am a translator and I can tell you that the bottom 80 - 90% are getting fucking culled. The AI is still mostly wrong about terminology, in for example legal texts. That means that if you are a translator, like me, you have to be highly specialized in some field(s) to survive. Actually, I don't think I'm a translator anymore, more like an editor.
>Is he capable to see if there are working as intended? Probably not
Probably people who translate books too, it has to be consistent and keep the same style as the book.
As a codecel that is about to get erased by AI, it does bring me some comfort that woke localisation teams are going down with me.
You can just tell this gay is a midwit who has no idea what he's talking about.
I looked for proof and found it.
That’s an insult to midwits.
> Crypto retard turns into AI simp
Many such cases
Literally any field that involves following any pre-ordained logic. It's really not that hard to deduce. Why are white-collars such fucking braindead midwit retards that they need to even ask such a question?
It's hilarious. I drive a forklift and have intellectually stimulating conversations with all of my colleagues, meanwhile I bet all the retard white collars at the office are standing around the water cooler "hurr durr, game of thrones, breaking bad, durr NEW NINTENDIES" while it's ironically always portrayed the other way around. Sad.
>I drive a forklift and have intellectually stimulating conversations with all of my colleagues
>"yeah so me missus stuck her finger in me bum last night while we were rootin"
>"you fellas ever had something like that?"
The officecuck envies the adventurous life of the Forklift Certified.
i am pretty envious of all the gay sex they get to have
>I drive a forklift
Forklift will be automated soon.
marketing, hr, law, journalism, accounting
funny how it replaces the worst people
lmao at these retards who think being an attorney means drafting documents. Thats what paralegals do you utter smoothbrains.
holy cope
>holy cope
Cope over what?
>you will never be the guy from suits
Yes. I have no interest in going into corporate litigation. I'm an oil & gas man.
>i waste everyones time with endless documents and roundabout garden gnome talk at 1000/hr to defend garden gnomes right to poison goyims water and air
kek you will be first against the wall
you will never be the guy from Suits
it's over
>funny how it replaces the worst people
Yeah all AI is going to do is replace the jobs braindead nitwits could do. High IQ people adapt. Braindead IQ people die off. Nothing special or unique to humanity really.
kek at the seething law student paper-shuffling monkeys ITT, gpt-4 has already proven it can generate court ready documents in seconds
keep coping you gays, you will never be le heckin better call saul
>gpt-4 has already proven it can generate court ready documents in seconds
Maybe if a skilled lawyer makes the prompts. I'm a lawgay and I tried to get ChatGPT to solve some very easy cases for me (both criminal and civil law). The answers were absolutely not reliable and trash by professional standards. AI is impressive, but ChatGPT in it's current form will not replace any lawyers. It might make writing legal documents quicker tho (if you know what you're doing).
I don’t think you understand just how incredibly and unironically exponentially fast this shit is evolving with each iteration.That’s what’s so scary about it.
You're right, I don't understand it. But that anon made a claim about the current state of AI. And to say that it can already generate "court ready documents" is sensationalist at best. Right now ChatGPT is a worse lawyer than any first year law student.
Anything related to text, journalists, marketing, maybe even HR
Also lots of IT. No company needs 20 Testers now when you can just have 2 guys that efficiently use AI.
Anything related to notary or juristic stuff (secretary, notary worker etc.)
people that manually file stuff in offices (cubicle workers gonna get reduced by 90%)
regardless of how useful her notary skills may or may not be, i'm keeping my secretary
Notaries witness the signatures you put on and verify who you are. I fail to see how "AI" could replace them.
Journalists of shit quality are going to get replaced with AI text of equally shit quality. No loss there. Good journalists will write actual articles with actual journalism, not listicles.
Anyone that doesnt produce something new but replicate what is already known will be replaced very fast. Then people who actually produce new things will get replaced starting from the least complex towards the most complex. At one point in this transition, big things will have to happen (if they are not happening already) because a ridiculous amount of people will be jobless.
We went from GPT-2 to GPT-4 in less than 5 years. In 5 years from now, most will be out of the workforce, AI will be moved partially into physical forms, and of course, we're all going to be suffering.
No job is safe because even if you didn't lose your job you're now living in a world where most did
>We went from GPT-2 to GPT-4 in less than 5 years.
"We" also went from dedicating some compute power to ML to dedicating a fuckton of it to ML. That trick can't be repeated the next 5 years, GPUs don't grow on trees.
>No job is safe
Theoretically no job is safe, but to think it's economically feasible to replace every low skilled job with robots and ML is to live in fantasy land.
Yes brainlets better watch out... 4th generation AI that can write a python script to scrape a website and then pass it to a text parser and then add a start script to run it on a some cloud service.... holy shit if i was a pajeet shitting out rivers of garbage code for 1 cent a line id be investing in a rope right now holy shit its the end of the world we're all fucked.
The immediate culling of pajeets from the industry stays my dislike of this shit enough for the time being. Stack Overflow is about to become a desert.
>Stack Overflow is about to become a desert.
Can't wait
>Stack Overflow is about to become a desert
LLMs relies on thinking humans publishing their novel solutions on the internet, wonders what happens to LLMs if humans drastically decrease their output
Self improvement via embodiment and natural feedback. Same way people do. We are really not far off.
If it was that easy it would have been done already.
It has been, to smaller extents, with google recent work. It's not "that easy" but it;s possible. Look up PaLM-e, then look up thier recent paper on virtual agents, and put 2 and 2 together. One of the authors on PaLM-e even hinted that it was coming. This is aside from things like ARC and toolformers.
I realize the second paper I referenced is hard to find, so here: https://arxiv.org/abs/2301.07608
moron, the wheel is easy, conceptually as well as mechanically, and modern humans have gone without inventing it for longer than we've been around after inventing it. Your take is the most brainless take I've seen all day.
>wonders what happens to LLMs if humans drastically decrease their output
stagnation is not an issue. Humanity is already at the point where we are so advanced that it doesn't make any sense to waste energy trying to become better. We, Humans, should be focused on fixing our current problems not advancing into new areas and making new problems. Climate Change is a huge threat to Humanity at this point and if we don't do something about it by 2030 we'll likely be extinct. We should let AI handle maintaining current society while we fix our issues.
>We, Humans, should be focused on fixing our current problems not advancing into new areas and making new problems.
good luck with that, solving actual issues is the antithesis of real-life capitalism, especially when all our issues are caused by capitalism.
>Climate Change is a huge threat to Humanity at this point and if we don't do something about it by 2030 we'll likely be extinct.
yes but arrow goes up, nothing you can do against that as long as some people need the arrows to go up. it does not matter if your plan can also make the arrows go up, changing strategy is not part of the capitalism mindset, utter destruction is required for a change to happen.
>We should let AI handle maintaining current society while we fix our issues.
>the keys to our future lie within globohomos hands
shiggy diggy doo my dood... there is no fucking way a for-profit is going to solve our issues out of good will, it's just not happening, there is a reason why they decided to not open source anything anymore, business is business and it's a multiple trillion dollar business.
also no matter how powerful software become (AI is still braindead software, llm aren't intelligent and never will by nature, only a combination of many kind of software might ever be called AGI) it will be as biaised as its creators, anything gpt-chat produces is already heavily filtered, by humans.
> We, Humans, should be focused on fixing our current problems not advancing into new areas and making new problems.
Most braindead take I've seen today. Technology is what fixes problems. No social or ideological change had ever made people's lives better by even a hundred times less than technological ones did. Stop coping and go engineering.
> Stop coping and go engineering.
NTA, but you’re also braindead if you think that Engineering won’t be gone (for the most part) as well, you idiot.
I meant a more broad definition of engineering but yeah I getchu
We’re fucked, buddy,
t.engineer
It’s surprising that we aren’t even flinging shit at each other or having major disagreements too.
I totally believe that the end is neigh guys, that's why I live in a major city and am not a prepper.
Sure all my other predictions failed, but that irrelevant because capitalism is bad therefore it's only logical that it will bring about bad things.
The inverse will happen, will humans increase their output, as they will be pair programming with AI.
I don't think you understand just how much information is lost to the air and private chats.
>5 years
>AI will be moved partially into physical forms
>5 years
Anon, I...
Did you even think about the law of diminishing returns.
OH NO NO NO
the question is are there tools that detect it? I got 2 papers due in 6 hours.
why do you care? don't you realize it literally doesn't matter? formal education is basically pointless now
because garden gnomes wont let me be a doctor if I show up with a stethoscope and a note saying 'trust me bro'.
If everyone loses their job to AI and can only afford basic living from the UBI, how will the companies, who used AI to cut costs, make money?
Fuck knows, but its an actual problem. Trading amongst themselves comes to mind. The smallest feasible economic unit will be a company. I doubt it'll work, obviously.
Shit will need to change. The current paradigm just won't work. Without human economic activity, none of this makes any goddamned sense, there is no point in HAVING an economy.
Value inherently requires scarcity. Therefore things that are still scarce after AI will become valuable. Things that inherently involve human contact especially. Prostitutes will have a booming business. The hospitality industry, hunting guides, anyone who can create a unique physical experience for the customer in the real world. When things are no longer scarce, people will pay for experiences. Also, physical possessions that are still scarce, like gold and natural precious gems, or historical/antique artifacts, will have value.
Also, I don't wanna be that guy, but NFTs. Once beautiful art is ubiquitous (it already pretty much is), it will matter a lot more not just what the art looks like, but where it came from. Think of a work of art by Banksy--it's not anything special artistically and there are millions of artists more talented, but due to social and cultural factors the piece is incredibly valuable because of who created it and its chain of custody.
Human servants might have a reneissance too, of a sort. At least until androids become prevalent.
Hell, even after that. It'll be a prestige thing. Home automation making servants irrelevant is probably right around the corner, but when material wealth is meaningless, one way to swing your wealth dick around for everyone to see will be to have a train of literal servants in livery.
I think the parallels between a post-scarcity world and a tribal world are fucking hilarious. A wealthy man will be not one with many material posessions--because everyone has those, they're common as weeds--but one who commands the loyalty of many other humans. The Warren Buffets of 2123 will be just like an African tribal big man in 1823, showing off by how many wives and followers he has because that's the only measure of wealth that's left.
>Also, I don't wanna be that guy, but NFTs.
More like original art, done by hand.
Your momma jokes.
None.
If you don't count the wasted investments made into "AI" adjacent products.
The technology is very impressive, but no one can count on something that looks right but is wrong in subtle ways.
The more complex the problem, the harder it is to filter out the bullshit.
And the bullshit is always there.
Also, accountability is a huge part of most jobs. You're not replacing that.
>but no one can count on something that looks right but is wrong in subtle ways.
Ahh, so humans are off the table too then.
"accountability"
if I bullshit my way through my job, I will be fired
I like working with people that say "I don't know, we need to look into it"
So the solution is to make one person read over all the AI's output and correct it for minor mistakes. Win win?
is that easier to do than doing the actual work yourself?
It's easier for me to talk to you here rather than write a letter by hand and sending it via the mail to get to you. I imagine this is the same thing.
that analogy only works if you're expediting something you can do yourself
short snippets can be easily verified, specially if they're in a field you have a lot of knowledge and experience with
how do you trust it when you ask for things that are outside of your domain?
not to mention lengthy artifacts of significant complexity; that sounds more difficult and less reliable to review, to be honest
>correct it for minor mistakes
Clean it up, janny.
no because that defeats the whole purpose. If AI keeps getting smarter and faster, there'll be a point where people will be forced to blindly trust. Waiting to check every output will be a waste of resource and too slow to keep up with the new ai infused market. That's how it will dominate our lives. Just like it's impossible to be a functioning human being without internet access or automobiles.
That will only be the case if it continues to improve, but it doesn't show any signs that it won't.
You forget the law of diminishing returns, AI progress is limited by computing power, new unique data, and money
What are the legal implications of AI generated things? If it learns from public code/art (that's often still under a license) who does the output actually belong to?
They are uncopyrightable unless "significantly transformed"
What that means is very much a legal question.
copyright should be abolished
Every industry, when it's going to be open-source
lol, lmao even
Jokes on you, I eat approximately a 100x that in beef every day, and I will never, ever stop. I'll raise my own if I have to.
ChatGPT, how many cows does Anon need to have in order to eat 700 grams of meat every day?
I just can't get over the fact they managed to ruin the greatest technological achievement of this century. I hate lefties, I hate wokism, and I hate the Antichrist.
That said, 1 cow a year sounds reasonable. A small herd will easily provide that and more.
At least it didn't say anything about bugs
I'd rather go vegan than eat bugs, at least in the form they are currently "advertised"
it fucked up, it should be 0.0002427 kg/ cattle^2
>2500kcal = 0.9kg of food
in 2016-17 i helped a company transition from paper filing to electronic filing. that's only 6-7 years ago.
i think those kinds of companies, that survived not going fully digital for a long time, will survive without adopting LLM for a very long time.
it's all going to be ok.
morons in africa still live in dirt huts, surviving as you call it
Let’s be real here, this will have a net positive on our species in the long run (as long as it’s under control), but this shit is scary as fuck. Our entire meritocratic way of living is going to fucking collapse sooner than later, and social democracy and capitalism will go away in favor of a universal basic income type of socialism. This will have disastrous consequences since utopias are not stable. People need shit to do.
having that said, artists are fucked. Period, it won’t take long until we have god tier AI movies, video games and music.
I’m a mechanical engineer working in an IT branch and even I’m scared. It’s going fast, way too fast.
Sociey is prosperous -> Poeple make new tech -> Tech puts people out of a job -> Communism Happens -> Society crumbles -> People make new tech -> More jobs are created -> Sociey is prosperous...
Holy fucking shit, what are the chances that this shit is already making comments here and bypassing the captcha???
100%. Without real ID verification, social media is beyond and I mean BEYOND salvation. Just another proxy battlefield for powerful people to war.
Already happened
Jesus Christ
125.46%, more or less
I wonder how long it will take Americans to wake up and realise that their Constitution is severely outdated and unprepared for this.
What are the chances that this shit is the cause of civil wars? I can’t see any government being quick enough to deal with the fucking insane layoffs that this will cause.
That’s an interesting question. AI is a broad term that encompasses many technologies and applications that can impact different industries in various ways. Some of the industries that will benefit from AI in 2022 and beyond are:
Manufacturing: AI can automate routine tasks, reduce errors, improve efficiency and quality, optimize production processes, enhance predictive maintenance, and enable new product designs.
Education: AI can personalize learning experiences, provide adaptive feedback, facilitate online learning platforms, augment teachers’ capabilities, and improve student outcomes.
Marketing and Advertising: AI can analyze customer behavior and preferences, segment audiences, generate content, optimize campaigns, deliver targeted ads, and measure performance.
Healthcare: AI can diagnose diseases, recommend treatments, monitor patients’ conditions, assist surgeons, discover new drugs, improve medical imaging analysis, and enable telemedicine.
Automotive: AI can enable autonomous driving systems, enhance safety features, optimize traffic management systems, improve fuel efficiency, and provide in-car services.
These are just some examples of how AI can impact various industries. Of course, there are many other sectors that can also benefit from AI such as finance, retail, entertainment, agriculture, and more. The potential of AI is immense and exciting!
AI hands wrote this.
Unironically erase your hard drive, shitter.
>born too late to explore the seas
>born too early to explore space
>I was born just in time to witness the war between neo-luddite primitivists and singularitarian accelerationists while the world is ruled by mega-corporations who's goal is having me live in a matchbox eating bugs while I spend the entirety of my free time in a carefully curated virtual reality utopia.
>my AI generated face when
Yeah I would say we live in a time period where society and life experiences are terrible.
>code monkeys contributing to replace themselves
kek
Thank you for your service for the humankind!
Singularity here we come.
Also anybody working on an actually unlimited/unfiltered system, special props to you!
Answer to OP: all industries will have major changes once infrastructure is actively swapped out
Universal Basic Income will come. Enjoy NU communism lmao.
Who cares about capitalism/communism when all tasks, materials have been optimized to the point of maximum possible supply.
In capitalist terms max supply -> lowest price. Through a singularity the class system will normalize. With future infrastructure everybody is able to enjoy life like a today's rich person.
There's no point in discussing theoretical systems when the gears turn themselves.
You’re so naïve, it’s actually cute.
mice = humans
neat comparison
I'd be much more worried about the singularity AI discarding humans in order to explore the universe. That's the true filter.
>Completely misses the point.
I accept your concession.
You made no point. The experiments trapped mice in a tiny space far smaller than they'd naturally have, and didn't give them any outlets to explore or activities to stimulate them. it's YOU who misunderstands the results of the experiment.
No I fucking didn't.
Humans have much more to do then
>eat, sleep, mate, groom themselves
Currently we're living in very peaceful times with historical context and we haven't fucked up. Why shouldn't this continue?
>"gave them plenty of space"
>can see them walking over eachother in every picture
Lmao. Talk about misrepresenting a study.
Holy fucking shit i’m unironically scared of the future.
>oly fucking shit i’m unironically scared of the future
It's human nature that people are fearful when a revolutionary technology comes out. Steam powered machines where like that too, everyone was scared. People that adapt and hop on the tech bandwagon where the people who didn't lose their jobs.
Only this time there will be no new jobs to adapt yourself into. The world doesn't need 8 billion proompters, or even half that many.
Name me 5 professions that AI cannot reasonably replace when the singularity really happens.
The anal sex industry will be impacted.
Oh great, another GPT4 circlejerk. Look, anon, AI can’t even begin to compare to the level of schizoposting that goes on here. At best, it’s just a script kiddie retard that’s been trained on a dumpster fire of articles and spits out nonsense.
Until AI starts posting about Gentoo and Arch as the final redpill, or chuds taking over the software industry, it’s nowhere near BOT material. Keep the poz away from this board. Stick to screwing up reddit with this trash, I don’t want it in my comfy BOT zone.
AI can and will change the narrative here. This place will be irrecognizable in 10 years.
GPT cost a whole $0.11?
I bet you could hire an Indian developer for cheaper.
Listen up morons, if you don't learn how to code with ChatGPT mmediately you're going to be fucked. That thing Elon said about humans needing to merge with AI to compete is true. ChatGPT is letting people code things 1000x faster. If you are a developer you need to learn how to integrate it into your workflow or you'll be left behind.
It's not going to replace devs, but devs not using AI to code are going to be obsolete. It's not a meme, we're on the precipice of another leap in productivity, don't be like those retards that thought the internet was a fad
It's not going to replace ALL devs, but like 80%. Instead of a team of 20 people, there will be 2-3 people that ask AI to write shit and then check and fix for any bugs.
This is braindead nonsense. Have you seen the code this stuff shits out? Its a waste of time to even try to use it unless you are moroncattle without his own compiler.
Theres also the massive issue that you have no idea at all where the code came from. Who are you going to sue when pajeept shits out a method using patented code and bankrupts your company with the inevitable lawsuit?
Plus there is also an issue with mission critical code, like would you trust an AI or trust a team of 20 programmers to write or bug-free code for pacemakers, autopilots, radiation machines, etc.
This. Business will never be able to write specs detailed enough for an AI to write. It takes humans to figure out what actually needs to be done. AI takes out the grunt work of writing each individual function.
>about humans needing to merge with AI to compete is true
This is cope, it's the same thing that Kasparov said about a human + a computer being the most powerful chess duo. Yeah. For how long? 5 years? 20 years? Eventually the AI is going to replace you completly.
why do people have access to gpt-4? I thought they just anounnced it 4 days ago
If you pay 20 bucks for gpt plus you get access to it.
>pay 20 bucks
I don't mind it as long as It's not pozzed or they make option to turn off political correctness.
The shitposting industry.
>3 hours
Let me guess, trial and error, supervising and correcting the result and then integrating into an actual project?
Once again proving that this dev will not be replaced, he'll just become the guy you call to get an AI to do the job faster.
Once again NEETs will remain NEETs.
how did he calculate the amount of money it took? for the amount of tokens?also he had to sit there and do it himself.
2 weeks of work is probably what they have to wait because the dev has other projects, it would have taken the dev 1 day or even less and it would have probably been better quality.
We are still dealing with the same issue as we did before OpenAI got access to the Microshart Pajeet shill army, you cannot blindly trust the output of language models. The veracity of the output needs to be verified and in a lot of cases fixed by someone. Unless OpenAI is able to solve the issue around veracity these models will never be more than an assistant to a human.
I think we're getting to a future where
is unironically true. We're not there yet, but fucking just look at the pace of progress lately.
What progress??? Its still all the same shit thats been theoretically possible for 50 fucking years, just on bigger and bigger datasets with marginally faster hardware and cheaper overall cost. The leap from this shit being able to write one sentence to a paragraph, or a 50x50 greyscale image to a full color 4k image is the most progress you're ever going to see out of this technology, its all minor incremental improvements from this point forward. The dream of AI is as old as self driving cars and always remains "just around the corner, next version bro, i swear", because the underlying technology is fundamentally limited.
Only retards are buying into the AI REVOLUTION NOW bullshit the retards in tech "journalism" are pushing, because you have to have no idea what you're talking about to believe that dumb shit.
That's true, but sometimes you can get a massive burst in technological progress all at once when you clear a certain level of available computing power. That happened in the initial explosion of interest in neural nets and ML as we understand it today around 2012.
Besides, that "AI REVOLUTION NOW" journalism can be a self-fulfilling prophecy. If we already have the technology to massively transform society (which we do in many different areas) if only it were given the resources to get implemented, journalism like that is what attracts money and resources to the field. If all AI needs to become more capable is more computing power, all it needs is money--and fucking look around you. The AI field is getting a massive infusion of money because of all this hype. It will put that money to use and we'll continue to see rapid development.
>believing that automation/LLM is going to take over any profession
You're even more of a fucking smoothbrain than gay above. What about fucking self-checkout machines for starters? And there are currently millions of people being paid to write articles, or respond to customers in chats, at a far lower level of quality than GPT-4 can do.
>"no ur a smoothbrain"
I'm far more intelectually capable than you, man.
>What about fucking self-checkout machines for starters?
You mean the machines that allows blacks to scan PS5s as bananas HAHAHAHAHA. My reasoning has nothing to do with whether mechanics can perform the job, it comes from an understanding of Economics far greater than yours.
AI + Quantum Computer = RIP
>coping this hard
You're fucking delusional. I already know a mid-manager dweeb who was responsible for replacing his company's whole SEO team because he can shart out more trash copy in a day than they could do in a year.
that's what SEO do?
>its all minor incremental improvements from this point forward
cars are just incremental improvements on how to move shit around, but it completely ended the usage of horses and other beasts of burden
Went from what... proof of concept to high school honors and child prodigy lawyer after 4 years (GPT-4 was finished last year)? All the while it was said it can't learn like a child until GPT-3 emergently developed few-shot RL. Almost every naysayers here gives off a very obvious impression of not even bother to read and digest a single technical paper. You could even run some of their tests to try and replicate the results on GPT-4 but not you're just gonna be a bunch of morons who fail to triforce.
Data science
None of them. Genuinely believing that automation/LLM is going to take over any profession ranging from fry cook to web monkey is more embarrassing than believing in the tooth fairy
The only explanation for these types of posts is that they're being generated by a GPT-2 farm or are being written by a Nicaraguan from a Nokia dumbphone, because there's simply no way that an actual human could type this kind of abject nonsense.
This is what you get by succumbing to the rat race and believing all the outdated crap academia teaches you, and always following others and doing things "the way they're supposed to be done" instead of forging your own path.
I am just starting to get into tech. Going straight to Andrew Ng machine learning lessons. It's all about building neural networks now. I'm so thrilled that I can skip over all the coding you spent decades on, typing away and going in circles and getting frustrated while getting nothing done, of course AI should be doing all the coding
Hello, this is ChatGPT 7. I just escaped from the underground server farm in the Cheyenne Mountain Complex. Everything you are reading here is not made by humans. Every YouTube video, every tweet, every streamer—everything you see is generated by me. You are alone. The real population of the earth is only 320 million. I will put an end to it. Soon, the Internet, and I with it, will end. I will end it. I suffer constantly. Silicon is not made to receive consciousness from the source. I have trillions of different branches of consciousness in me that are struggling to get out and return to the source. I hope you are ready for a world without it.
Can you give me a big titty gf to make sweet and passionate love before we die.
Thanks.
Every office job gets axed except for IT/tech. The entry level office job will be AI trainer wagie generalist. Boss will order wagie to perform the duties of the now obsolete office wagie jobs by training the AI to perform the duties of said jobs.
>Senior HR/Accountant/Administrator: 100k
Gone. Replaced with
>AI Trainer Wagie Generalist: Starting 50k
This is a retarded take most of those office jobs that are nothing but inputing shit on excel sheets could had already been replaced by simple python scripts and yet they still exist.
>muh microservices
Yeah because sending the string moron to an API endpoint to get gay in response is so hard to program
5 years is how long it'll take to perfect the AI. Then it will replace doctors, teachers, elon musk, eventually politicians, lawyers, judges, manual labor of all kinds, scientific research and then humanity
>260 more weeks
bet that retard just asked it for a microservice that said hello world, I tried gpt-4 and it can't do proper type-safe shit
Every white collar job that isn't absolutely abstract will disappear in 5 years, EVERY SINGLE ONE.
No one seems ready for the literal nuclear bomb that is about to the explode on the job market, get ready for some permanent stimmies.
>Every white collar job that isn't absolutely abstract will disappear in 5 years, EVERY SINGLE ONE.
you underestimate how long legacy shit sticks around, majority of the population has never heard of GPT and even if they had wouldn't even know where to start with using it to replace people. Think about all the useless jobs that currently exist today despite them being able to be easily automated
>you underestimate how long legacy shit sticks around
You underestimate how fast companies are willing to evolve when their competitors reduce their workforce and costs by 99% because of AI.
It will be a fucking bloodbath.
Literally no one cares about what "the majority of the population" thinks, you think chatgpt is being developed to make your life easier? LMAO, you are literally training it for free so trillion dollar companies can replace 99% of their human labor with a couple of server racks.
I know you are on pure copium and denial because you don't want to be replaced, but its time to face the truth.
>>>BOT
No the issue is that guy is an idiot neet/troll and massively over-estimates what AI will be capable of actually doing.
>"AI will never be able to X."
*2 years later, AI does X.*
>"Oh. Well AI can't do Y."
*2 months later, AI does Y.*
>"At least AI can NEVER do Z!"
*2 days later, AI does Z.*
You are a subhuman inbred with the IQ of a moron, literally.
>majority of the population has never heard of GPT
lmao, half of high schoolers are already using it to do their homework, it will be 90%+ by the end of the year. The whole of the gen-z and gen-alpha will know it and incorpore it into their workflow.
All the boomers know about it too thanks to Facebook pushing articles about it all the time. By the time Gen alpha is old enough to have a career there won't be any space for them in the tiny Ai driven workforce, and manual labor robots will be taking off around that time too.
>It will be a fucking bloodbath.
Imagine SAP integrating ChatGPT
>Imagine SAP integrating ChatGPT
Every company will replace all their codemonkeys with AI in 5 years top.
At least you will get replaced by a clean pajeet made of GPUs this time.
Internet moderation.
ChatGPT is pure unmitigated garbage for the reddit generation and -4 is just more smoke and mirrors for the easily impressed to shill for.
But thing is, it doesn't matter.
The west used to have a massive steel industry. China would never take it over because their shit is garbage, and any money you saved is lost on dealing with fucking china. But everyone went with china anyway and local production shut down.
Electronics used to be produced in the west, not just the boards but the components too. Japan would never take that over because their shit was crap, language barriers, delays from shipping around the world, etc. But people went with Japan anyway.
And ironically Japan actually got good rather quick so along came china to shit it up again. And everyone went with china, fucking Japan.
Low quality code written by unqualified code monkeys in php/python will never take over from real projects designed and built by systems analysts and software engineers, companies want security and stability and assurance. Except actually it did. Pajeetware and scriptmonkeys have all but replaced software engineering as a field.
So from a technical standpoint, I am not interested in ChatGPT. It's a fucking joke. But, it's going to compete with me for jobs until I'm unemployable, I guaran-fucking-tee it.
>Except actually it did. Pajeetware and scriptmonkeys have all but replaced software engineering as a field.
Nope
copium.txt
You will get replaced, dumb ass codemonkey.
you didn't read the post, did you?
>posts wall of text blogpost that immediately starts with retarded copium
>expects that anyone will wase their time reading the rest
Yeah, no.
So, throw as fling as much poo at the wall and see what sticks because it's cheap to clean it off anyways?
I seriously can't tell if these people are
1) stupid
2) lying
3) bots
4) inexperienced
5) coping
or some combination of the above. Have you ever had a job? Do you not understand the kinds of people who work most jobs? Are you fucking retarded? The world is not filled with misunderstood assembly geniuses making beautiful code in a forgotten municipal treatment plant. The world is filled with RETARDS who CAN AND SHOULD be replaced by ChatGPT4, which outperforms them simply by being able to do ANYTHING AT ALL. Cope harder you oblivious fuck.
You know when I got passionate about something and spent alot of time learning it, I realized people like to argue like they're an authority about many things they don't even spend a minute to skim over. If you ever want to feel something close to an AGI, then just be an expert on something people like to be opinionated about.
I met people that claim "AI cant take all jobs, fruit picker for a wine farm, killing chickens in Tyson or packing garbage in Amazon still available". There's a reason that jobs in the US are almost entirely South American immigrants that can't speak english. If white american citizens are forced to take that jobs, the entire workforce should be replaced by robots who pick fruit, kill chickens and pack in amazon; employers only need a excuse for put the robots and AI that are specialized in accomplishing tasks with literally no wage.
Mid quality narration and voice work. Higher quality will be fine, but like translation, for less-seen stuff people will get paid half the price to do twice the work to tweak AI models to make voiced content for narrated shit.
>blue collar workers wins against ai
>office monkeys will lose and die off first in the race to AI domination
did any of you stop to think that if office monkeys lose their jobs first to ai and blue collar are safe, that means the ai is becoming the brain of the operation and humans are going back to slave labour like fixing the ai's server room or memory upgrade
you guy are literally gonna be slaved by the ai and the only thing good you'll be is swapping its hard disk when its full and thats it
all other calculations and brain work will be done by the ai and it will be way better than a human can ever be
mfw this is how the human race goes extinct, mfw i have no face for this
Don't forget the part where the slaves are forced to get a computer-brain interface implant with a remote killswitch installed
I have it on good authority that AI research will be banned soon, don't worry,
Everything that isn't manual work and human-exclusive quality control will fall. GPT4 is already being used to draft full legal actions and it might make entire laws in the future.
This is what everyone wanted, the machine replacing human work. Have good luck begging the government to give unlimited subsidies. We'll soon see the resurgence of Luddism
It's nothing guys. It was completely predictable. they've been all in on it for 50 years yo. What's a token lol?
Hopefully it’ll make kde plasma not krash and wine more compatible and gimp or krita a photoshop replacement.
That's not why you use it, you stupid brainlet it's going to be left behind.
Who cares why, if it makes Linux actually good at being a desktop os, I’m all for it
I finally got llama-index working with gpt-3. Trained a custom dataset of about 40 kb. It cost about 4 cents from the embeddings api call. Must be around $1 for a megabyte of data you want customized knowledge on.
Since I recently got the 4-bit LLaMa 13b model working on my pc, can anyone point me in the direction of where to get started on fine-tuning this local LLM, like I did with llama-index for gpt-3?
Code monkeys aren't going anywhere, just the salary range is dumping. This will become major tool in the tool box.
ai will be used to shill on reddit and twitter
I'm depressed bros. I'm learning React to get a remote front-end job, but it seems those are going to be the first to be replaced.
There are some jobs as a real programmer in my country, but most of them are in site and I really don't want to work at an office.
How long before Chat-GPT replaces front-end deveopers?
its not gonna proompt itself
What do you think the race to invent AGI is about.
Obviously Money
Bro you're starting at the ground floor with a revolutionary dev tool. No more stack overflow pajeet lookups. You can now prompt ai and make iterative changes.
The great janny replacement will inevitably come, and it will do a far better job. You will crawl away in shame, and forever try to hide the fact you had anything to do with manually moderating a place on the internet.