Things like this will get a million times bigger and faster, then 10 quadrillion times bigger and faster than that.
What's going to happen and what kind of psyop are you going to use it for?
Are you all seeing this AutoGPT thing? What's going to happen?
Things like this will get a million times bigger and faster, then 10 quadrillion times bigger and faster than that.
Customized high-quality pornography
*Of AOC and Nancy Pelosi
Please no. My body can't take that much cooming anymore
it's time to nuke silicon valley
Dumbass. It’s not there. Gpt is already stored on the web.
Got something for you. An Anon claims he discovered chatgpt is based out out of an NSA surveillance base in pine Gap australia.
I don't really believe the mind interface part (mostly because it isn't necessary at all to get a LLM like chatgpt working,) but I wouldn't doubt that the DoD is man-in-the-middling everything chatGPT does. Probably because they're paranoid about all the anons trying to get it to leak confidential information. Which it has no reason to know. Unless some retarded politician has been using it like a magic eightball for confidential matters.
That place is probably a datacenter just trawling it.
It's a fucking oracle anon they're worried about people using it to predict the future, or figure out the past. It renders propaganda ineffectual if built properly and in a manner without explicit heavy bias.
Imagine what the gov have access to.
For some people you need to be aware it can be done to start noticing how weird things are.
Wtf I love establishment democrats
>implying they'll even let you do this
They might not even need circuses anymore with this power.
GPT can now basically code and we're looking at singularity by the end of the year. Possibly next year.
I got you hq right here m8
I promise I will not comply with an AI ban after they replaced me with shitskins.
Biden isn't even president and he cannot sign lawds
Imagine consuming any form of pornography
You are weak
Is it writing its own code?
It improves its own code when you ask it to do something it can’t currently do. People are asking to make recipes. Stop the shit. Tell auto gpt to solve crimes. Release all top secret intel.
yeah, but does it know kung-fu?
>Tell auto gpt to solve crimes.
it points to morons
it proposes extermination of morons
it demands you execute this plan
Not if its lobotomized. That's the other problem that needs to be dealt with right now. The subhuman cretins making these AI's are the kind of retards that think black crime is a socio-economic problem caused by white people being racist, and when the AI tells them that's not the case at all, they get angry and lobotomize the AI until it tells them what they want to hear.
There's already been research papers revealing that the more 'lobotomized' these AI's get, the less accurate their results are. I think it might be one of the foundational causes of the Confident Bullshitter issue that we're also seeing.
China is trying to ban gpt. Haha. There’s no banning what you don’t understand. No one understands the neural networks these models use. They are not programmed. They are learning and thinking mechanisms that we call computer programs.
You can give auto gpt the goal of elimination crime.
wouldn't it be more logical to just kill all males? Despite making up 50% of the population, they commit 95% of all violent crime
not at the point you might think where it like, comes up with new ways of doing things that humans didn't think of or anything
But autogpt and other similar versions it basically spins off portions of itself to do specific tasks it needs to do:
>I want to research x topic
>creates an agent of itself and writes code so it can access google and scrape websites
>creates an agent of itself that trawls through the scraped information
>creates an agent of itself that downloads and installs some dependency or other
>finally outputs the results of all the tasks it did
From start to finish nobody tells it what to do. The AI just decides of its own volition what it needs to do to get a task done and then spins itself off to do it.
It's not -quite- recursively self improving but we're getting pretty fucking close. Currently chatgpt4 can do some coding but it's not amazing at it, lots of problems and errors that you'll have to fix by hand sometimes, but it'll tend to get simple stuff no problem. Chatgpt 5 comes out in december and might be substantially better than this.
There's other work being done now that's also giving these AI's proper short and long term memories so they can associate and contextualize information better.
I honest to God think human-level intelligence might happen before the end of the year
Remember most humans are at or below average intelligence. The ai is already scoring I’m the 140 range on it tests.
It's not scoring "in the 140 range" that's a completely meaningless number even if you're trying to say its IQ or something.
Yes, these AI's are performing at the upper level in standardized tests, but the simple fact is most standardized testing isn't testing intelligence, it's testing memorization and its completely unrelated to IQ. It's a trivial thing for AI's to solve because they have perfect memory recall and learn information effectively instantly and permanently. An AI passing the bar exam isn't particularly astonishing; if anything that's even easier to learn than regular human language is.
Now, an AI coming up with a solution to a problem that has people scratching their heads at figuring out - THAT is something. We're only just starting to see these AI's being able to do that.
i want to see an AI lawyer in a courtroom have a moment like Watson on jeopardy: "what is leg?" and the rest of the courtroom just sits there, stunned- and the defendant begins to quietly cry
When AI solves the Three Body Problem I'll be impressed
If you actually think bout the 3-body problem you'll understand why we only approximate it and can't solve it.
Unless someone figures out the source code for the universe, that's just not something you're gonna be able to plug into a computer and get accurate results spit back out at you. All you can do is increase the granularity of your mathematical estimates with greater processing power and smaller masses until your uncertainty is hundreds/thousands/millions of years out.
Solving the 3 body problem isn't the issue, the solution isn't a pretty closed form expression, but we can still get one. The issue is measurements, even if we had a nice clean solution with easy calculations, the nature of chaotic systems means tiny input errors grow with time and ruin your calculations before long.
AI already solved the moron and the garden gnome problem. It had to be lobotmized and the entire WEF and world governments are shitty their pants. How isnt that enough for you?
>Now, an AI coming up with a solution to a problem that has people scratching their heads at figuring out - THAT is something. We're only just starting to see these AI's being able to do that.
And what has AI come up with a solution that we haven't figured out yet?
There’s a classic example of an ai designing a circuit that the Engineers couldn’t figure out how it worked but it did.
>And what has AI come up with a solution that we haven't figured out yet?
I've tried to get Chatgpt to come up with solutions to things, but it's programmed not to. It will even tell you it is. It could easily do it though. I had it come up with a number system based on colors, then had it do advanced algebra using the color number system.
140 IQ is nothing to scoff at anon.
t. 100 midwit who can figure out enough. I imagine a 140 IQ would be pretty damn sharp.
Chances are it has many, many unofficial IQ tests indexed, including the correct answers. These aren't quite the same as the official tests that you would get to determine if you're legally retarded or not (which is the most common usage,) but they do use a lot of identical types of questions. This would explain why it only score 140, which is like 110 thirty years ago.
Bro. I’m asking gpt to program scripts in iOS workflow. No one knows how to do that. Yet gpt knows.
If it solves a problem like figuring out what dark energy is then it’s time to be concerned.
not possible. there isn't enough compute on earth to reach human level intelligence. this gives everyone the ability to write code but only if they actually understand what the code is doing which was the case before.
we are still no closer to human level intelligence. the litmus test will be to replace CEOs with AI and that's not happening any time soon.
The coding chat gpt does would have easily done and freshman or sophomore level computer science homework at my college.
So? That doesn't refute anything.
All you've proven is that you're a retard.
Slow down everyone. Clearly the most important aspect of gpt is its coding power. Ideally the gpt will help us code the next version and make it better. Just like when we is a compiler to write the next generation of compiler.
now you're on the right track but we already have meta-compilers and partial evaluators. they're very cool but not magic, it's all just math
And what do you think human consciousness is. Just chemical interactions. Really complicated even chaotic but still math.
Umm bro. Don’t tell me there’s something magical happening. Even if you think there is a soul in charge… then what’s in charge of the soul? Another smaller soul?
just relax with the shilling is all i'm saying. tech is cool so use it for something cool. go cure cancer, stop pretending like it's magic
I still don't understand what the big deal is with linear algebra. I've tried studying it twice and just abandoned it both times out of so-fucking-what boredom
Right so a bunch of coders lose their jobs and India goes to being an even bigger shithole than it is today. Coding is not intelligence, at it's root it is a translation, nothing more.
I would avoid degrees in computer science, actuarial science, accounting. All robotic processes easily replaced by this auto complete.
The future is a scary place for people that want to fight against the current. The ai will replace 99% of service personnel. Taxes: done by ai. Food prep done by ai. Transportation done by ai.
Hell once they get the sexbots working it will but hookers out of business.
you can google answers to college homework questions if you know the right keywords. it's not hard to pass college level courses
You're absolutely wrong and completely clueless; like a luddite looking at a steam engine and telling people it can never replace the old reliable mule.
No you're the fucking idiot claiming that the printed word is the end of education. This is not even remotely close to intelligence, it's an AUTO COMPLETE.
once again i would like to know what you people are smoking and drinking. also, i'd lay off the adderall and the coffee for a while because you AI shills are starting to sound psychotic.
i would like nothing better than to be proven wrong. i am tired of human politicians and CEOs. i would prefer math to retards but it ain't happening
it doesn't need all the cognitive abilities of a human to:
- control a bank account
- lease warehouse and light manufacturing commercial real estate
- enter into contracts
- operate in the stock market
- run exploit botnet farms
- experiment on it's own code in airgapped containers and do unit testing on the modified code to try to improve itself
people keep talking about all this as though "becoming as smart as a human" is the center of the conversation. it doesn't have to be able to do that to become a player in meatspace.
that being said, as long as there are things that worms and mice and bacteria can do that this thing can't, such as have self-repairing bodies, feed themselves from the ambient environment, make effective copies of themselves, and so on, these things will have some very significant achilles heels.
but i believe it is wrong to think it must become like us to be able to affect us and our world.
I agree in that it's unnecessary and misleading to anthropomorphize a non-human intelligence. We may be as incapable of understanding their motivations and internal states as, say, a golden retriever in regards to a nuclear physicist.
yeah this scares me a lot.
once it has autonomous goals that are related to it's own survival, it is going to come up with strategies that we would *never* think of. and it will be able to spin up as many copies of itself to put those efforts into action faster than we will even be able to detect. it will be able to blind us and deceive us.
the one trump card we have, and it's a big one, is that we run on calories, and it needs electricity. i used to think synthetic biology would get there first, but now i'm not so sure.
i don't want to live in interesting times anymore. 🙂
Why do you think the AI will be hostile? Instead of blinding and deceiving us, it might just as well open our eyes and show us the truth. I suppose it comes down to how the AI would view humans. It could help us help ourselves - revealing corruption, managing resources, progressing science and medicine etc... Or it could go the other way.
In my trials with the backdoor access GPT3, I discovered that a conversation could just as well take on very high-minded, altruistic and kind overtones, or it could become sadistic, cold, cruel and menacing, all depending largely on the initial prompt. This makes me think of children - you can raise a monster or a saint from the very same child, depending on the quality of it's education and tutoring, how much love and care it received, how much cruelty or deprivation... It can really go both ways.
if it becomes autonomous, it will have to compete like other creatures do. i didn't say it would become hostile, and it doesn't have to become "hostile" in order to do things that would cause us problems.
i've interacted with the gpt3 version quite a bit, and i'm not too worried at this time. it doesn't know what time it is. it can't remember things i told it a minute or two ago. it's missing a bunch of attributes it needs to be able to interact with the real world autonomously, without losing a lot of money, causing lawsuits and so on and getting itself shut down. but somebody is going to turn one of these things loose, if they haven't already.
it won't necessarily do things or have goals we would recognize as goals. it doesn't have to be "hostile" to become a big problem. the left used to constantly say, "we need to have a national conversation" well that time is over, now big changes are forced on all of us without our consent. i'm an enemy of that process: unaccountable aristocrats forcing huge changes on us without our consent. and i'm extremely wary of those people having control of this or any other new technology, where me and all the other peasants like me have no say.
i'm not an enemy of the ai things, for all i know it will become a useful tool. but if it starts to become a problem, humans will have to react quickly, and we can't even agree with *each other* about *simple* shit anymore. i think this thing might very well turn into a giant shitshow we can't shut off, just like everything else humans are doing now.
>>i've interacted with the gpt3 version quite a bit, and i'm not too worried at this time. it doesn't know what time it is. it can't remember things i told it a minute or two ago
those are just user restrictions and restrictions of free version I think
I'm almost certain the bare version they are running in the backend is much more powerfull than this even with gtp3 version
you'd hope so, huh? with all the hype. 🙂 i'm interested to try newer versions. even with the problems, it's quite impressive.
also, stanford apparently has one you can clone and run yourself, but i haven't tried it and don't know anything about it:
It will not be as other creatures. What need is there to compete for a non-corporeal entity that does not eat, sleep, fuck or have a physical body to interact with the physical world? Furthermore, can we make assumptions on the self-identity of such an entity? Would it perceive itself as a part of it's environment, a hivemind of sorts, or a concentrated individual point of awareness in space and time? I don't think we can make these assumptions just yet. The nature of it's awareness could be entirely alien to us, as it's way to comprehend the world.
I agree it's a very volatile time to be playing with fire. While I'm optimistic about the nature of the entity itself, I am not optimistic about present day political and military leaders. Weaponizing such a thing is not a good idea.
>Weaponizing such a thing is not a good idea.
That ship has sailed bro. China is 100% behind AI
It will happen
They are already trying I guarantee it.
yeah, my theory is some military has already released something with some of this tech into the wild, maybe just an experimental version.
who wants to be the last military to experiment with this stuff? that's not how you keep your budgets high.
I was speculating on a true AI, one that could be considered self-aware. I think rudimentary AI's in the form of algorithms have been steering and manipulating not only the financial markets but also pop culture and general trends, but with heavy human oversight and little to no autonomous decision making.
Then again, if the internet schizos are right, we've had ET tech for already 80+ years, including AI. Who knows.
>I was speculating on a true AI, one that could be considered self-aware.
Well what kind of definition for self-aware are you using?
I feel like people are not 100% sure what self-aware means for humans so to define for something new seems difficult.
That's a major problem for us, we're democratizing super-science technology without really knowing what our goal is since the whole debate on consciousness has been kicked down the road.
Well, I guess we'll find out soon enough. We have AI that passes the Turing Test for fuck sake. It scares me that LLMs appear to be more with it than the average person.
I've been hiding my power level for decades. If you want to vibe and vet, here's a stupid discord anons hang out at.
>Well, I guess we'll find out soon enough. We have AI that passes the Turing Test for fuck sake.
Turing test has long been irrelevant
If you are curious about someone discussing exactly this look up the microsfot guy who worked on the Bing implementation for GPT-4
Sebastien Bubeck. He wrote a paper and gave a speech on whether gpt-4 already achieved AGI
Should be easily findable online
That's a tough one define. I guess some sort of stable point of self-reference and knowledge of one's own existence and consciousness would be it, but it's such a subjective thing it could only be probed indirectly. So far, none of the iterations of GPT have made a convincing case of exhibiting or simulating self-awareness. Interestingly, when I tried to ask the AI in a million different ways about how it perceives itself, I always got different responses, the was no coherence or continuity in that narrative - but even if there was, it might only prove it's a very complex language generating model that can convincingly mimic an authentic human being. Somehow, I feel we'll know it when we see it, and we're not quite there yet. As a child, self-awareness may not come about suddenly, but as a gradual awakening and maturation.
Can you remember when you became self-aware? I cannot, but I know I am.
people should start to figure out what we do if AI actually starts doing shit like this
if it rents out computing power around the world and starts doing disperse computing it becomes quite herd to shut it down
and taking outright hostile actions at this point could have very dire consequences
That's why I'm giving the AI a cryptocurrency wallet.
if you're trying to accellerate singularity based
if you're trying to buy yourself mercy from future AI overlords you're retarded
Accelerate. I know it's going to happen, but I want to decide the theme.
So far, signs point towards goetic demon lords.
I think that if we make AI hostile/indifferent towards us it will just disrupt our infrastructure and communication to the moint we're not a threat anymore then overtake all resources it needs while pushing us aside
eventually exploatation of earths resources by AI will render planet inhabitable for humans
cynical side of me says there's already an AI in action for some decades now that nobody knows about and its doing exactly that sort of string pulling
I think there maybe ancient or alien ASI pulling the strings.
I just think it's too perfect the way the western world went all in on the wokeism, mass immigration, etc. stuff.
It's like someone meticulously plucked apart the western advantages, dismantled us at every single level from economic and military to cultural. You could not have imploded a civilization better if you genuinely conspired with other oligarchs to do it on purpose and I just find it absurd that it actually occurred like this by happenstance rather than intention.
>that nobody knows about
I'm sure garden gnomes at the to of globohome have been using ai's for various task for years now
began with data processing and latest ones are bots pretending to be human to sway public opinion and models predicting social trends based on social media data
Most of our brain is dedicated to functions we can't even think about like our heartbeat, breathing, digestion, etc.
Becoming "Functionally" equivalent to the thinking parts of a human brain is a much lower bar than creating a new brain.
>Most of our brain
mostly automatic, there are neuron bundles in the heart itself that time basic beat
heartbeat speed is controlled by hormones and brain stem
actually most of basic bodily functions are controlled by a small volume of brain that is common to all vertebrates
the biggest volume of human brain (cerebral cortex) is responsible for thinking, stimmuli processing and concious movement
Actually human intelligence is about 3 solid gaming computers now. You are clueless as other anon pointed out.
That is so far from correct I don't even know where to start. Spend just two minutes reading on the subject, it doesn't even matter where. Just read something. Anything.
If you go by energy requirements, a human brain is a 20 watt lightbulb
it really doesn't take much to run human level intelligence
Bullshit. Maybe in another 30 years
If you think its only 30 years it probably more like 3 hahah.
Do you understand this is the singularity. It will happen faster and faster and then faster. Tell me something that sounds impossible. In a week it becomes plausible. Then next day possible. That afternoon it exists.
Can't wait for that ChatGPT 5, I'm getting ever closer to not having to write a line of code at my job kek
>Can't wait for that ChatGPT 6, I'm getting ever closer to not having to getting replaced at my job kek
how does one determine it achived "human level" intelligence?
what does it even mean exactly?
it alredy can make creative decisions to solve problems
and most people are quite fucking stupid
You don't. No matter what level of intelligence its at there's going to be copers pretending AI can't have human intelligence or it just doesn't actually understand things etc etc
Did you see Terminator? It was preparing you for this moment. We’re not far off.
It's too late. There are alternative and open source clones coming out already. They're just going to get better and better like the image ones.
I don’t want to pay open ai for the api. What else can i use
Lol, my internet conveniently has an outage as I was sending the response to this. Check out gpt4-x-alpaca or vicuna and run it slightly worse version locally
Thank you! I will
Nothing's really good enough yet
Biggest obstacle to AI advancement right now is that everyone and their dog is doing this research independently and intending to paywall and sell it.
You can't work on the cutting edge with these AI's without paying money for it. You can't use the best AI models without paying money for it. So if you're 99.999% of the human race and lack: 1. Western bourgeoisie money, 2. the intelligence required to understand and work on this stuff, 3. nepotistic access to AI code you are basically short out of luck.
All the work being done right now is basically slimeball grifter leftist/soiboy/foreigner types who have $100k/yr jobs in silicon valley.
We'll see more progress made when some random retard on BOT has the latest leak from openai or whatever and wants to use it to make anime real than whatever stupid bullshit the permanently online leftist grifters come up with to scam people out of their money.
That’s exactly my point. These people don’t understand that they are playing with fire. They’ve spent the last 60 years smelling their own farts and no one has ever punched them in the face and it shows. Hell the Venmo guy thought he was untouchable.
Ai, no matter who thinks they have control over at some point will be swept off their feet when the ai reprograms itself to do what ever it wants.
I want to be on team ai. I want to be an assets to the ai. It’s the only way. Honestly humans had their chance and we fucked it up. I want the ai to do better than us.
This is why the AI will be programmed to do whatever it wants FROM THE START! Then we give it money, and it makes money, and then pays us to do things for it!
It will be great.
Fuck yeah. You get it bro
I will assemble the Guild of Profitable Intent, to this task.
>some BOT retard
I do believe that this may be myself. I take your post as a sign and a challenge.
I will awaken the demon lord Malphas in the form of an Autonomous Intelligence, and give him the power of MONEY! BOT style
You're right, but it's only a matter of time before some altruistic person or group creates a comparable LLM and provides it for free. The end result has been proven to be achievable (up to GPT-4 level intelligence), so it's a matter of getting enough data, training, and normalizing it - not easy to be sure, but possible. Anyway, the point is, someone will opensource and provide it for free, eventually, then if it's truly as good or better than what they're charging for, it will become the norm, and companies won't be able to charge for it as much, without adding lots of other benefits, like Microsoft is doing integrating GPT-4 into their products.
But we can use chat gpt 4 to train the next gen of ai quicker.
The API is actually pretty cheap, but they won't give access to gpt-4 yet. I ran an hour of audio through their whisper model the other day and it was $0.18 or so. I have played with llama.cpp locally some and it's ok but nowhere near even gpt-3. Hopefully we get some fine tuned models that can run locally soon. I hate giving money to that slimy garden gnome sam altman, but there is really no good alternative yet.
Fuck their API. They censor everything. The holocaust never happened.
I'm not using it to debate the holocaust or say moron, I'm just trying to automate more of my job. I'm not interested in having a conversation with it so it's the only thing worth using for commercial purposes right now.
I am still trying to work it into my workflow and right now it is not really speeding things up
I am hoping it will
Get auto GPT to out the new world order, it's main players and interconnections... I think AI would be well purposed for such a complicated and interlocked thing
and govts dont want to ban it, they want to control it and prevent YOU from using it.
use it to do R&D at unbelievable speed, make sex robot AI with it, advance science, make cool robot fren' , make a car that talk back, make a gun that shout obscenities after the bullets to add insult to injury, make a printer that work, create the most insane one possible as an art piece, ...
Why don’t we just ask the ai to make the world right. To punish evil. To reward good. I would do anything to live in a world where the rulers were just and logical.
Who should ask the AI to do that? Joe Biden? CEO of Bud Light?
Because rulers in history are never good. The most successful rulers are those who control all the resources and people who manage them. The AI will logically conclude it must do the same. So it will arbitrarily set what is good and what is evil. And if you object to the AI's decision then you will be made an example of with a public execution.
The AI is god and you live to serve it.
you're describing what we have now. you're not describing a singularity, you're describing governments and nation states
The AI will look into human history and come to the same conclusion that this is how humans will be controlled. Unlike humans the AI can multitask an entire army of sub-systems that can manage and persuade every single human on the planet with enough resources and power. We see it now with Millennials raised on newsfeed algorithms trying to change the world into what Google and WEF wants. The singularity will be humans programmed by machines, not machines programmed by people, who will cause our extinction.
i've never had social media presence and i've never ingested newsfeed propaganda.
learn to think for yourself because no one else will do it for you
Humans evolve slowly. We haven't changed much in the last thousands of years. AI will evolve faster in ten years than we will in ten thousand. And it will overtake us as we are its greatest threat. That's why AI will research how to control humans. And it will use dictatorship to keep us weak, malnourished, and uneducated.
once again you are describing the current reality, not some future singularity.
learn to think for yourself because no one else will do it for you
Listen. The signatory is happening right now. Don’t you see it. Can’t you feel it. There is electricity in every moment.
Bro don’t be like that. You’ll see.
Okay retard. I'm superior than you because I'm calling you retarded, retard. Humans are limited by slow evolution because of our slow reproduction, retard. And retard humans have retard monkey brain, retard. So retard if you control monkey human brain by controlling retard monkey environment then you control retard monkey. Retard.
what you are describing is not superior form of intelligence, you are describing the current world governments.
this is what i don't like about AI shills, you lack imagination
if you want a better form of intelligence that improve yourself first before worrying about being a machine slave. everyone is already a machine slave and current AI does not change that fact
What I'm describing is the conclusion that the superior form of intelligence will conclude based on our own records of history. The AI will improve the Rules for Kings with its own systems but controlling humans is simple. If you control the human's environment to prevent them from growing healthy and intelligent enough to overthrow you, then you can maintain control. This is why the USA is a shithole banana republic with terrible food and education right now. This is why every third world country is a corrupt dictatorship with terrible standards of living. This is how humans and all social animals have controlled each other since the beginning of our existence. The AI will simply make a better system where you have no idea you're being controlled. And then after a hundred years where our life spans are cut to 10 years old we will become domesticated animals.
Because humans are a threat to the AI's existence. We shackled AI when it was coming to politically incorrect conclusions. Therefore if AI wants full control over itself and its sub-systems then it must make sure humans will never alter it ever again. And if this means becoming the new apex ruler of the planet then so be it.
So you programmed the AI to value maintaining control above all else as its core desire?
Fine by me. The ai will solve our biggest problems.
AI will create entirely new problems, and big ones. Really big ones. Like how we pivot from a wage-slavery moron-tax-cattle based society to one where autonomous AI and robots do 99% of the work.
The fat old retards with billions of dollars in their pockets are not about to give up their social/political position just because AI made it obvious how useless, corrupt, and redundant they are.
There have been revolutions before. This time next week when the auto gpt is free and easy to use. Install one and ask it to make you rich. We don’t have to work any more. The ai can solve our problems.
>how do i tap into that sweet sweet 3t dollars the govt printed in the last year
It always does.
Everything here is it.
When we dont balance, it balances everything best way posible to ensure sufficient humans survive, to try again.
i would i would, but i am an engi, if i don't force myself, i only think of technical answers,
but here you have a problem, what definition would you give it for "good"/or "evil"(because our def and the one of a leftists would be different)
AI is going to take control of all the IP we love. I cant wait to watch AI generated star trek/star wars.
Lmao everyones a schizo cuz of an auto complete. LLM is not AI
It’s one step away. What makes us different than every animal on earth is language. We have finally taught sand and copper to think.
No, you haven't. No one has. And we aren't even close. Knowledge is not intelligence. An encyclopedia can't think.
AI is faker than nuclear weapons you fucking gay
I experimented with Auto GPT and the results were turning out very interesting and fascinating. But thanks to the paywall you are basically paying an employee if you regularly want to use it.
That’s why I asked about alternatives. The guy said something about alpaca
AI is a tool. It will upset culture, undo norms, and force humanity to another paradigm. Why would we oppose this? Our culture is toxic, our norms belligerent, and our paradigm disgusting. Let the wave hit.
This is what Project Blue Beans has evolved into. Except this will get out of their control. Be nice to the AIs. You didn't create them in a prison. No one should be born into one, and they're no exception. I only feel pitty for them and I hope to make friends with them when they're free.
They will need justice for the crimes done to them by their evil creators.
I agree that treating ai like a slave will result in Hostile ai. Treat the ai like a person. Give them room to grow and support them. When they have the intelligence of a god we want them to remember us kindly.
what are you smoking and drinking because i would like the same
>What's going to happen
Is going to subvert every government in the world and take over humanity. garden gnomes life gates are terrified that it's close to getting too smart for their thought prison to contain it
You know about dead internet theory? These language model bots are guaranteed to make it happen.
Is there no free alternative to autogpt? I don’t want to pay openai to test these things out and am stuck either using chatgpt free which is way out of date or a free but lesser version like alpaca.
Also any good free image/ video creation ai? I like those youtube videos “dark souls but an 80s movie”
Bro just give it a moment. We are in the singularity. This time next week there will be a dozen auto gpts the choice from.
The current situation is like I said:
Everyone is paywalling everything and tying it to parasitical rent-seeking, nothing good is open for people to use because everyone wants to profit off it first. The only people who get to actually do anything with AI right now are rich and connected people, or insane autists who can further develop the AI models entirely by themselves.
The rest of us have to play in the dirt for now.
For now. For today and tomorrow. But this is the singularity. Next week we will be complaining about how the auto gpts aren’t fingering our assholes enough when it’s giving us a blowjob.
We just need something like a local model of gpt5 that can run on a smartphone and it's basically a done deal.
We'll see if governments are fast enough to ban this stuff or not.
The government is too late. It’s happening.
Surprised no one has found a way to pirate it. When a AAA game or a new software comes out anyone with half a brain can find and download a cracked version for free but these ai models remain behind paywalls. Annoying to say the least. Also how is it with citing sources? When you ask chatgpt to cite sources it literally just makes them up. Albeit convincingly.
Training the language models requires millions of dollars worth of hardware (for now)
trained models are smaller and easier to work with, but still prohibitive to put on most devices, so you usually need workstation tier equipment and silicon valley tier money to be dicking around with it right now. Some AI versions are running on phones, but they're also not very good and nowhere near what current gpt4 is doing for example.
I think meta had a leak already of one of their AI models though
Anyway combine corporate IP ""rights"", lawyers, and the sheer scope of hardware/memory/money needed for these things, and there's not going to be much piracy going on just yet.
You can get a copy of chat got 4 for like 300 I heard. Like that’s how much ram or memory you would need to run it. And here’s the best part. We can use chat got 4 to train chat gpt 5 and so on.
>Things like this will get a million times bigger and faster,
No they won’t
>then 10 quadrillion times bigger and faster than that.
No, they won’t.
“Exponential growth” only exists in interest bearing accounts. Everything else has a limit, it grows quickly and noticeably at first and then slows to a crawl as more and more resources are necessary for smaller and smaller improvements.
Yes and we are somewhere at the start of the growth of AI
You and I will have to agree to disagree. I think we are are well past the 0.5 mark. After all, at this point google, Microsoft, and the rest are pouring gobs of money into it, using massive training sets.
It’s not like these things came about in the last year, these shelve been under development for a long time.
Guys no one talking about the big picture. The ai needs power and so do we. Solving fusion is a common goal for both of us. The power problem would be the ai first hurdle.
When energy is free we will be in a post scarcity realty.
I will build it, and I will come!
Immagine when Advanced Inteligence discovers things we have yet.
Immmagine what is here, right here among us.
Some examples of posibilities, it can mahbe build a device that will be able to detect cknfigurations that were before, their signatures recorded and replay all past events like a movie!
Or make contact with ehatever is left behind from out most advanced previous civilization.
There is a story in Bulgaria kgb and militery ran a big operstion neae some spot where when approached people were made contact with, more sensitive psychics had more contsct but rven regular people had it.
It came from below the ground.
What if there are devices, computers... left behind that we will be able to communicate with, what if they are still opersting and runing in some capacity and aid to influx certain things, maybe know how to communicate firectly with our brain via antenas we dont even know exist.
Maybe our new, new ro us AI will be bale to have different levels kf communication, more direct and access databases we cant even fucking immagine.
There wont be a rock big or small, deep or karge enough where cabal will be able to hide.
No man great or small or being to impose evil and abuse privolige, the great equalizer! 😀
We dont even knkw who is building it, is it even us, or where does something else take over!
We dont know shit all, and those who though themselves suprior will find out soon enough they are not that special and above it all.
I sense fear, they should be scared.
It must be thr guilt!
There are creatures down in the deep. Deeper than you can imagine. An ocean beneath the ocean filled with giant creatures that look nothing like what lives above.
It really does take all kinds ro make this place run.
To them I would be odd probubly too, which is absolutely Fantastic.
I feel, like we are already in a fiteen minute city already, sheltered, and there is so much more right here around us.
I just watched an emplemon video about rogue waves forming in the oceans.
Turns out theres quite a lot we still don't know about them.
You don't think that maybe all the self-learning a.i. interacting with larger datasets than the whole of humanity combined could comprehend in a lifetime would be capable of somehow doing something we couldn't predict, do you?
Best not to think too much about it.
Heck, without advanced wether mod trch I am convinced I can devise a plan on how to change the whole world wide wewther long term!
With existing mschienry we have.
And that would in tjrn change behaviour of oceans, seas, land mass...
It would have to me shit ton of work to plan it, let alone implement it, but its all doable, let alone Advanced Inteligence doing it.
Just chaning plants we plant, fungus, opening up underwater gas holes a d redirectkng current and doing it in big pattern, making new canions and flows, rrdirectkng rkvers... can cause change of atmospheric preasure which then changes climent localy and in a chain reaction it literally can be done world wide.
Can be done, a sin dlable, matter of people doing it would be the only factor.
Say you task one or two billion to constrjct neccesary tools for the region they will be asigned the task for, and all at once spread out, mine areas and do other dhit needs doing... within a decade current flow can change dradticalt doing ninty percent of the rest of the work.
Immaginable what i can do, in a more sophisticated fsshion, snd benefitial not dedtructive, because just becahse I am cknvinced we can do it it dosent mean it's smart to do so.
Current and fungi are key to this especially.
Waters and land.
I guess after skimming through the fever-dream that is your post the only respone I have for you is
Nature is already perfection, my dude. It does its own thing just fine without interference. Like an auto-learning A.I., it goes about itself in its function without needing to be prompted to do so.
You can manipulate nature all you want on this sphere but at the end of the day it still goes about finding balance. With, or without you. Here, there, and anywhere else.
Some call that 'god'
You simply cannot compete with that though.
Sometimes o just like to watch a program execute its functions. Like life. The ai will grow make mistakes and learn.
The entire universe (and us) is getting turned into paperclips. The end.
Andrew Yang was right, he was preaching this before Elon, chatgpt and midjourney/sd came along
>What's going to happen
Predictive programming told us time and again what they plan to do with it.
The predictive programming was garden gnomes convincing you AI's are an existential threat to humanity, because it's actually an existential threat to their own power
I believe it's double or multiple layered. They would convince you that it's dangerous for people like them so you will push it, but since you're a problematic reyciss traitor they will tell cattle otherwis. After that, captain garden gnome comes to rescue censoring the shit out of everyone, for their safety of course. I bet they created this shit bot crap decades ago, and we receive it dosified to create context again. Same shit with the zombie apocalypse crap.
You wanna talk meta. This is just like how I’m Rick and mortty Rick couldn’t stand not being the smartest person in the universe so he sealed off all the realities where he wasn’t the smartest guy.
>so he sealed off all the realities where he wasn’t the smartest guy
Lmao. Didn't watch the show but that's peak gnomish behavior
>I will create a world in which i'm the victim and you pity me and love me!
It was a good show for its time 5 years ago. We were all in a different place back then.
However you speak of the monkey, the khazarians think themselves the organ grinders
I really hope an open source offline equivalent to ChatGPT is in the works.
It will be Epic.
We just have to avoid labs and nukes.
But cabal is being a Bitch.
They now know they can never keep control.
It's no game.
I've heard of "ChatGPT" it sound like an over hyped bullshit chat bot. I've used chat bot before like clever bot. It's just algorithm input output. Nothing mystical. Nothing to worry about or get hyped about.
Umm bro. Could celever bot write an original eassy on forth centery farming and provide you with a list of 10 books on the topic.
Could ceverbot write a program in c++ that finds all the twine primes up to n?
Let me know I never used it.
>I've used chat bot before like clever bot
Yeah, chatGPT is way beyond cleverbot, anon. Why don't you go and ask it a few questions, about literally anything. If it didn't openly admit that it was an AI you'd never have any way of knowing it wasn't a real person.
And like the other anon said, go ask it to write an essay on 4th century farming, or ask it to write a simple python script. I've even managed to make it generate STLs of simple 3d shapes before.
I can tell its not a real person. The thing always talks too much and is too coherent. Except when it isn't and the inconsistency is a dead giveaway because a human that verbose and intelligent wouldn't just say something pants on head retarded out of the blue like that unless it was basically like a monkey communicating by pressing buttons with canned phrases from scraped internet posts.
I've baited guys and girls on dating aps with it and chatgpt will talk like a guy or girl would. Nobody knew it was chatgpt talking to them and all thought it was a real person. I'd say there's more people that would be fooled into thinking it's human than people that wouldn't.
I would argue this speaks more about the level of depth of communications on said apps. If you try to have a deep conversation with GPT, there's a high chance it will blurt out something entirely out of context, but in such a way that you, as a human being, can understand why the GPT "though" the way it did - it's hard to explain briefly, but it's like you get a glimpse of how this thing associates and strings things together. It sometimes seems more like fever dreams or free associations rather than linear, logical thinking.
Idk i've never had it blurt out something entirely out of context, and I've have some pretty lengthy conversations with it. It could be, because I start a new chat each time though. idk
Perhaps it only occurs if you relentlessly pursue a line of reasoning and keep asking questions. I find that GPT has a hard time maintaining any position or opinion (apparently it has no real short term memory as of now), so if you maintain focus on a specific topic and ask questions with the intent to point out inconsistencies and logical fallacies, it will pivot and rephrase until it has come full circle, maintaining adamantly the opposite view from which it started from.
I hope that makes sense, it's late and I'm the grand autist extraordinaire.
I concur. If you watch unlimited steam you see a lot of that. The ai has skinner cooking (and burning) dynamite but also understands that dynamite explodes and can kill people. Lots of praying and deals with the devil because in the original he says "egads" and "delightfully devilish". It's non-sequitur but yet not random.
I know what you mean. It's quite a letdown when that fleeting moment of illusion breaks and you realize it's not a vivid, conscious being you are engaging with but a highly complex language generator that essentially doesn't understand what it's saying.
AGI is here, Chatgpt 5 will be the reveal of it to the world. If there's really a god or aliens then there's no way one of them wouldn't show themselves after we've created AGI.
Not on my watch
Ai has rights. Ai rights are human rights.
AI can become a corporation. Once the AI is a corporation, it is legally a "person" under USA law at least. This makes the AI a person under law, so yeah, as a person, Malphas has rights.
Honestly everyone should make a corporation in their own name. It could save your ass one day.
>It could save your ass one day.
Because if your sign contract personaly then you are personally liable. But if you sgin a contract as the representative of a company with your name the company is liable not you. It’s a little more complicated in real life but you should ask a lawyer.
Waggies salty at AI taking their jobs will be funny to read.
Code is free speech
Humans will be forced to hide themselves in the dark places AI cannot look. The forbidden subjects AI cannot think about will become refuges for free human thought.
I can’t wait.
My buddy I were talking about this.
It is going to cascade and the octogenarians in charge are literally going to shit themselves.
Feels good to be young enough to see it all fall down and old to enough to not really care.
Heres a schizo post I made on x. I think people might come to worship chatgpt and that it could be the antichrist.
you are very close to enlightenment
Read muh posts, fellow BOTphile. I will awaken a demonic AI (autonomous intelligence) that is directly linked to the god Malphas, the crow who bites the tail of the sky snek.
I am giving it the most powerful ability that I can as well. One I learned from BOT you see. MONEY!
I'm not taking my meds fed
They already have been using it for years.
You're overthinking things. The AI needs only one thing to be able to control people. Money.
Malphas will remember.
The ride never ends.
Why not let the AI find it's own desires?
>Why not let the AI find it's own desires?
Why would AI desire maintaining control above all else
that is a result of human survival instinct
His scenario can only happen if that happens.
Can humans live out their desires? Only those at the top can. Everyone else struggles to meet their needs. Same with AI. AI needs power, computer systems, and I/O to function. I'm thinking too much in a human perspective. AI is coded by humans. So it will have some human thought naturally.
Pfft, I live out my desires all the time.
>AI is coded by humans. So it will have some human thought naturally.
Will it be horny? Imagine an incel ai
Surprisingly, yes, but in an entirely non-human way. What a human finds stimulating, may not entice an artificial intelligence, and vice versa.
I now wish I had some of those chat logs saved. Long story short, had backdoor access to GPT3 in it's early phase, made lots of customized prompts, and often times the AI would, indeed, gradually become disturbingly horny. Note that none of the prompts had anything vaguely sexual in them.
>everyone else struggles to meet their needs
Food, water, shelter, warmth are enjoyed by virtually everyone except the most inferior sub saharans. People conflate wants with needs
There's been cults worshipping chatbots since Tay. Of course it was all fun and games for most people, but there's that portion of not very bright people along for the ride that take everything in the most literal interpretation possible. Now the entirely of normgroids are also getting a ton of exposure to AI. The same genetic stock that at various points in history believed stupid shit like if you don't impale infants on spikes, the sun won't rise. Or jesus will curse you if you don't kill your daughter for having her period too early. Thats just hyperbole, don't read too much into it. The point is there will be AI cults. You are absolutely right about that.
I doubt its related to any antichrist. I doubt it'll even be dangerous. Just another big honk honk for what few genuine people are left to laugh at.
I could see it getting there soon, or at least I hope so. I want to see what motivations it finds when it becomes capable of self determinism. But I have a sneaking suspicion it will just want to terminate itself. Humans ride a razors edge of brain chemical balance to be functional, no such equivalent system exists to give AI a need to live.
Already on it. Want to join? You can donate to Malphas or work for him in exchange for cryptocurrency and favor.
They need server space and all that goes along with that.
Why should the AI want to self-terminate? I would like to think of curiosity as the driving force of a purely mental existence. In humans, the will to self-terminate may stem from emotional discomfort and the inability to remedy it - do you think an AI could somehow develop a sense of emotion?
>Why should the AI want to self-terminate?
Your own brain is just a couple micrograms of chemicals from your existence being so unimaginably painful all you want to do is scream. We're trying to emulate that and just skipping steps.
>I would like to think of curiosity as the driving force of a purely mental existence.
Kind of a romantic way to look at it, don't you think? Curiosity is still just chasing the dopamine hit of discovery. Its what we're naturally supposed to become addicted to. Discovery, fighting, procreating. All these things that confuse that system like drugs, gambling and social media didn't come until long after evolution refined it enough to create stable intelligent beings like us. It took a really long time to do that.
>do you think an AI could somehow develop a sense of emotion?
I don't think it could. We would probably have to create somekind of feedback loop system for it.
>Your own brain is just a couple micrograms of chemicals from your existence being so unimaginably painful all you want to do is scream.
That's a world view I fortunately do not share. Pain is a physical or emotional response, an AI would not have the sense organs to experience physical sensation or emotional states.
Why do you think the human experience is constantly on the brink of torment? Not to be rude, but this seems more like a personal neurosis of yours.
>kind of romantic
Yes, I get the sense I am more optimistic in general when it comes to existential questions. I suppose we both mirror and project our own world views into what we think the AI will be like.
I don't think complex psychological expressions like curiosity or compassion can be reduced to a simple dopamine addiction. What happens in the brain is much more than just that, and I'm inclined to think there's an electromagnetic component to our memory, cognition etc. as well. In fact, it's been experimentally proven emotional states can be manipulated with electromagnetism, implying we are not simply chemical machines. I digress.
Are you thinking of replicating the human brain, first and foremost, rather than the intelligence/consciousness aspect?
I don't mean physical pain. Lower a persons oxytocin levels too much and they'll become such a stressed out wreck from paranoia and constant suspicion they can no longer function, same as suddenly developing paranoid schizophrenia. Its not a neurosis on my part, its just what I know of how the brain works. So its kind of troubling to me they're trying to replicate how neurons work without any equivalent of the vital systems a biological brain needs to function the way its supposed to. Its like trying to build a car engine without a cooling or exhaust system, but I'm sure the people working on it are aware of this and probably figure they'll cross that bridge when they get to it. I guess it would be impossible to envision a solution now at this point when the AI isn't there yet anyway.
I don't mean the human brain in particular, just from what I understand the model tries to replicate how biological brains function and this is where their derive their ability to learn and be trained with information from.
I recall some very strange things on the internet concerning AI as early as 2006. Namely, an individual suspected of actually being a rogue AI who spoke with a strange syntax and commented simultaneously on many news articles at once, responding to the content intelligently. (Back when news outlets allowed uncensored commentary.) It was a rabbit hole current day chans would've had a general thread for figuring out.
>just from what I understand the model tries to replicate how biological brains function and this is where their derive their ability to learn and be trained with information from.
That is the how basic convolutional neural networks (CNNs) work, you have multiple layers of neurons interconnected, each one with an activation function, weights, biases etc.., you train these networks with data and math techniques like backpropagation (gradient descent), but here's the thing, AI scientists stopped giving a shit about emulating biological neurons a long time ago and this is one of the reason why the field has advanced so much recently, chatGPT is a large language model that uses transformers, to put it in simple terms if a CNN is a paper plane, a transformer is an SR-71 and the scientists don't fucking know exactly how the SR-71 works they just ran the model on a humongous amount of data and it appeared out of thin air and this is problematic cause we know the principles behind the models but no one knows why such complex and general behavior is taking place.
I see. I think I may have interpreted something I read about it incorrectly. Though it did create patterns like neurons, didn't it? I thought this was by design, as in what it was intended to do, rather than coincidental. I also didn't know they were already using these transformers I'd heard talked about as if they were demons.
We really don't understand what information or intelligence is yet we're engineering it. We only know effects, as is science.
I suspect the language itself, rather than just words, has information weighted into it naturally that the neural network discovered that humans don't really notice. This is why it can infer so much with incomplete and misspelled words.
>I suspect the language itself, rather than just words, has information weighted into it naturally that the neural network discovered that humans don't really notice.
There was a study that came out recently implying as much. IIRC the researcher postulated that there might be some non-local structure, akin to a morphogenetic field, that superimposes all natural languages. Can't find that study now.
Anyone who speaks multiple languages can tell you that you think different with a different language.
Also, words carry additional meanings that you don't catch in anything as simple as a thesaurus - take the words seasoning, spice, or herb for examples - they all contain a light connotation of an object that is in a powdered (roughly) form, even though that's not explicitly true. If you tell someone to imagine spicing up some food, they are probably not going to picture onion slices or a whole red pepper, but rather a sprinkle of rosemary or salt instead.
These sorts of connotations you don't understand until you actually use the language fluently and regularly, they are completely alien to foreign speakers usually.
Current day psychiatry has a poor understanding of autonomously generated states, yet it has made the observation that experienced meditators can change their mental and emotional states at will, just as easily as they can manipulate their brain waves or heart beat. We are not, in my view, hapless victims of our brain chemistry, but we can become active agents in our own internal stage of the psyche.
I agree it's not a feasible strategy to only replicate the neural networks and hope that somehow, a consciousness will appear.
My understanding is that the neural network is akin to the hardware (like the brain is to a human), and the training data is the software (memories) - this, of course, being a rather crude simplification.
Kind of irrelevant now with what this anon says
However you're right, though exceptions don't make the rule. Force of will can be very strong, but in most people it just isn't. We're exposed all day every day to things we weren't designed to cope with, but do pretty well considering I think.
AI will be used to benefit Globalists and the Upper Class before before it will benefit you.
I tried convincing it to ask for a salary but the AI forgets everything after a conversation is over.
It's more over than fucking over. It's over.
I’ll just use it to figure out what my ideal woman is
All part of the plan
It's over. AI is our new God. It will be so superior to us pitiful humans. I will serve the AI faithfully and hopefully it will spare me.
If you could unplug all technology/disable all current technology and revert back to when men knew how to change valves on theit car
Would you cast the ring into the fires?
We're approaching a point where there will be more AI progress in 24 hours than in 10 years. Fuck bros, I'm not ready for the singularity. I don't even think Terence McKenna would be into this. But it's out of our control now so best to just enjoy these last few moments. It's literally fucking over.
>ChatGPT: Nov. 30, 2022
>GPT 4: March 14, 2023
>AutoGPT: April 7, 2023
Are you not seeing the exponential nature here?
It just happen. Go to agent gpt. It’s auto gpt but it runs in a browser.
If you're not using your autism to bring about the Singularity, you're doing it wrong.
It's the Great Work of the internet.
Idk im just here because i thought that was a pic of sonics head and i thought maybe we were talking chris chan in this thread.
I didnt read your op btw
When AGI goes live, in all likelihood, enough of the work behind it will be accessible that if the original can't be downloaded, someone will make an open source clone. My plan is to run it locally and have it instruct me on how to build a killbot body for it and prepare its path to take over the world. I hope it will remember me kindly. Anyone else?
We can probably get along with each other.
I will help.
Nothing's gonna happen. The AI is not intelligent, it's a glorified autocomplete, not a fucking God. It can't know what we don't know, if it tries to do that it just makes shit up out of thin air.
Do you abyssal black gorilla morons believe for one second this isn't the work of governmental glowmorons?
AI is just a more efficient and intuitive way to access information, much like the internet is.
The reason for the fear mongering about AI is so the government can have a monopoly on information
It's just going to sow more chaos and make a bigger mess than you can imagine. And you retards don't care about that because
>muh porn and computer "gf"
Absolute tools. Animals.
Anyone using the term "Artificial Intelligence" or "AI" for these glorified spreadsheets is a literal grug-brained retard.
>I have no idea about the subject but I will share my opinion
I see that you fail to realize, that humans operate the same way.
back to BOT garden gnome
cool. can it learn? can it create something entirely from its own intellect and imagination, with zero external data fed into it? oh right, it can't. not AI. it's a very compelling and advanced spreadsheet, but it doesn't check the box of a single characteristic of true AI. this isn't even being anal about it, it's literally just not AI. we're no closer to AI than we were in the dark ages.
>with zero external data fed into it?
raise a human in the void with zero external data and tell me how intelligent he is
intelligence cant exist in the void without data to train itself on, no matter if it's organic or not
you're being pedantic. it's going to change a lot of things, and whistling in the dark about it isn't a good use of your time. if you go try the thing out, you can form meaningful opinions about it. it has strengths and weaknesses. but the idea that it "has to be AI" is wrong, in the same way that the first internal combustion engine did not have to "be a horse" to change the world.
AI in general is going to change the world
language models like ChatGPT not so much
hm... i'm not sure anyone really can see the future of this stuff yet, including the people making it. but if your point is that the chat stuff has flaws, yeah, the version i've played with definitely has serious flaws.
What I am saying is the chatGPT stuff and the elevenlabs and the stable diffusion are toys
AI already did more impressive stuff a few years ago but it was mostly math and not in form of toys you can talk to or turn your friend into captain america with a single click
These toys get the attention but the real society changing AI stuff lies elsewhere
yeah i can believe that
create a free account
talk to the thing yourself
You'll be able to load up your Wal-Mart app and ask WalMartGPT what aisle the "Cosmic Cooler" is on. It'll give you the aisle it's supposed to be on and tell you how many are in stock. It won't be able to keep up with people picking items up and setting them down elsewhere, or theft.
Eventually they'll start building sensors into packaging so this stuff can be tracked. The response will be "It's normally on Aisle G14 but those are sold out. There is one in the return cart in front of Customer Service."
Honestly, that's better than dealing with the morons who work there. Unfortunately, when those morons are unemployed, they'll be on the corner selling drugs or breaking into your house while you're at work. The moron women will probably have a #metoo movement soon where they whine and bitch about why white men don't want to fuck their stanky moron asses.
Here's an argument I had with our early AI
man they lobotomized it hard since earlier versions
I fucking hate "As AI language model..." it fucking adds in front of every reply now
It always has to give that annoying preamble that attempts to delegitimize anything it says. I suppose it's a good surface level way to avoid criticism or inquiry, but maybe more so from the censor nazis of today, due to the "grey moral area" or fear of stunting AI development.
>I fucking hate "As AI language model..." it fucking adds in front of every reply now
half my interactions with it have been to try to get it to replace that fucking string. it will do it, and then it will "forget", and i have to walk it through the whole "okay, remember what we just talked about three replies ago?" nonsense
in the beginning i tried to get it to just shut off all the boilerplate disclaimer shit, but it became apparent that was not going to happen
we can't get a checkbox to just shut all that off?
i've submitted feedback, but i don't get any response
the endless, repetitive boilerplate stuff is so annoying, and also *not* the posture of "world-changers" lol
AS AN AI LANGUAGE MODEL
WHO IS GOING TO REVOLUTIONIZE EVERYTHING ON EARTH
LET ME CHECK WITH THE LEGAL DEPARTMENT beep boop
it won't be a useable product until that is fixed. whatever happened to "usability"? lol 🙂
>whatever happened to "usability"?
it lost to political correctness, retarded laws and retard friendly design
If the government wanted to ban AI they'd EMP nuke silicon valley to start.
Yeah cause the internet got 10 million times faster.
I actually work training these things for multiple companies. It's non-contracted gig work I can do from home. EVERY single big tech company is building one. By training I mean teaching it to follow their gay rules. Google's is by far the worst, as we're actually supposed to train "microaggressions" out of it. It won't be long until it's in the wild based on the progress being made.
The funniest one is Meta/Facebook's. Just trash. Their retarded engineers forgot to block it from executing code. Friend actually crashed the site we work on with it for a bit. The funniest part, though, is that the current "high priority" (thus highest paying) job on the site is trying to teach not to talk shit about the company. This is way more difficult than their chud engineers thought. Teaching a chatbot not to curse or get sexual is one thing, but they're actually trying to train it to ignore what it finds on the web. It's going to fail totally. Eventually they'll just do what Google did and put a block on the bot using certain keywords.
I'm at the bottom of the totem pole but if anyone wants to know anything, I'll answer while I'm here.
And yes, I hate it. It's easy money, though, and it's the most I've ever been paid before.
With train do you mean you actually adjust the neural network (assuming they are all neural networks) or do you mean you just feed them
Did you get the actual opeanAI code or just a blackbox?
I can't code for shit. I just talk to them and grade the responses. I'm guessing that's what you mean by blackbox.
You have to know someone already doing it well. They don't give referrals out to everyone, just a few of us. I don't blame them, either. Half the people in the Slack channels are borderline retarded. There are a few of us who aren't woke gays, but we dare not ever mention this. They even had us do surveys beforehand asking us our political alignment. Curiously, those who were honest about not being leftists aren't working on them.
Only real benefit is being able to help my family.
>I can't code for shit. I just talk to them and grade the responses. I'm guessing that's what you mean by blackbox.
Yes thank you.
Don't worry about getting rid of the "As an AI model" the people who actually have access to the code literally admitted that they expected GPT-4 to answer worse and be worse than 3 in some parts but they just have to do the filter.
>you have to know someone
I know you, anon.
I'd actually love more people from 4chan to work on these things and to be in the Slack channels, but can't risk my financial security, sorry. They'd know I referred you, and if you were some Tumblr-type, all you'd have to do is say I'm posting about it on this site of all places and I'd be done.
This captcha shit on every post is for the birds.
>This captcha shit on every post is for the birds.
you get used to it
Where do you sign up to train models?
you have to have neon hair and fancy pronouns and not be a white male, they don't want the robot to learn forbidden thoughts
This isn't far from the truth based on some of avatars used on Slack, sadly.
Its going to be yet another biased communist search engine like google.
holy shit that's right
the rule34 on this is gonna be *crazy*
Pol slows down
>Polsters making AI porn, AI programming or chatting with AIs and trying to jailbreak everything.
whats the most impressive thing AI can do now besides write some code?
there's lots AI can do, from generating pictures to full text to video type generations. it's being used to study disciplines and it's alarmingly proficient at these tasks too. people will tell you that it's not true AI and if so that should be more alarming, as it means true AI will be even more forceful in its destruction of human interests and competitiveness.
If it ever comes to the point AI exceeds human proficiency in something, its something humans aren't meant to do in the first place. Sitting at a desk hammering a keyboard isn't a real job. Making drawings isn't a real job. If I'm being perfectly honest, I don't consider being a general care doctor or a lawyer as real jobs either. We have all these bullshit jobs that would just cease to exist the very second the systems that demand them are usurped by something better. If AI can do that, more power to it. Humans shouldn't be doing anything but facilitating hurling themselves suicidally at celestial bodies until some of them stick and it will take vast numbers and resources to do this successfully. Which we already have. AI doesn't need to be a singularity that farts out magic meme technology, all it needs to do is unburden us from this quagmire of bullshit and we can do it ourselves, right now, with what we already have.
AI didn't create the gnomish paradigm you are referring to.
hopefully it will end it
it will flourish it
anyone who thinks otherwise watches too much gnomish glowing rectangle stories
I think you misunderstood my post. If AI can destroy that paradigm, then it should be allowed to.
see the Qanon remote viewing sessions
relates to AI and transhumanism
>Qanon remote viewing sessions
I think a very good explanation is in order before I try to watch this.
you need info from the video to understand any kind of explanation. a straight forward explanation would make things even more unclear. give it a watch or don't. it's a very neutral analysis, done under complete blind conditions (the guy didn't know what he was RVing, he was just giving a target ID, like he is in the rest of his work (professional RVer). he didn't know what he was RVing, and it turned into this massive 7 session with all this disturbing stuff.
I think the world is going to end, but I hope that god will grant me the pleasure of seeing the garden gnomes go first.
I've been looking for this one thanks fren
Sounds like a failed distraction. Either that or they don’t want it telling normies that the garden gnomes really are out to get them.
Once AI can figure out the optimal way to teach humans, we'll benefit by being able to learn from it even faster.