Why is the progression with AI so slow? I feel like smartphones and other technologies evolved a lot faster.
Why is the progression with AI so slow? I feel like smartphones and other technologies evolved a lot faster.
Why is the progression with AI so slow? I feel like smartphones and other technologies evolved a lot faster.
>waifubots
nothing short of AGI will get us there. check back in 60 years anon.
this. by the time waifubots exist your penis will no longer work
Nah we'll be able to have waifubots from purely statistical methods.
Which is why I want the progression to be faster.
We've had a foundation since early 2010s.
>We've had a foundation since early 2010s.
Lmao, try 1950.
Because we're mostly applying more computing power to already existing methods, and the increase of performance in computers has mostly tanked.
I want my scandinavian 10/10 indistinguishable from reality sex bot and I WANT it NOW!!1!
nordmoron women are subzero/10
The development of AI is a complex process that involves many different factors. One reason that the progression of AI might seem slow is that it requires a lot of time and resources to develop and test new AI technologies. Additionally, AI involves many different subfields, such as machine learning, natural language processing, and computer vision, each of which requires its own specialized knowledge and expertise. Finally, the development of AI is also limited by the availability of data and computing power, both of which are necessary for training and evaluating AI models.
Its amazing how easy it is to spot ChatGPT. I wonder if its inherent to LLMs or whether they just gimped it somehow.
It's just lazy shitposting for the most part. The first sentence always sounds like the start of a high school essay. I'm sure with some clever proompting (like specifying "as a BOT green text" or something) they'll be nearly impossible to distinguish
I think its going to speed up exponentially from here now that they have a good foundation to build on.
because most of the big progress comes from data, and we're already using a fucking shitload of data
It's very different making something that can "think" then do vs something that ultimately just facilitates what the user wants to do.
At this point, computer performances are the bottleneck. You can't just go and train a model like dall-e or gpt's models on your computer, you need many millions of dollars (close to a billion) worth hardware.
>computer performances are the bottleneck
Why can't we optimize the models instead of waiting for hardware to catch up?
that would require actual work.
The industry just started doing this because they could just throw money at it previously.
I don't need ai personality sexbots.
I just want them to be able to perform commands and be self cleansing. They can be mute and blindfolded to prevent uncanny Valley for all I care.
Nice fud, if you were around during char-ai you'd be aware they are already here
Hand held Radios came out in the 50s
Touch screens came out in the 70s, and at that time they were resistive and you could see the wires on the screen.
Tablets came out in the 80s
PDAs came out in the 90s
SmartPhones finally came out in the 2000s.
Imagine some soldier in the Korean War with a radio complaining that the tech is so primitive and automobile tech seemed to come out faster.
>Hand held Radios came out in the 50s
When do you think the first AI came out?