5 ms·

I really wonder what the next AI winter will be like. Edit: Perhaps OpenAI becomes a major tech player and we just see a cooling off of other AI investments a

I really wonder what the next AI winter will be like. Edit: Perhaps OpenAI becomes a major tech player and we just see a cooling off of other AI investments as LLM becomes a known in terms of its strengths and weaknesses. Its abilities reach a natural limit which is still generally very useful. Or maybe folks realize the degree of lies/mistruths inherent in its content is actually unmanageable and can’t be improved. After the hype wears off what it will be used for gets greatly curtailed and we see a big contraction. And there’s so many interesting side tracks along the way. I’m hoping for another AOL Time Warner style shit show of a merger. That’d be fun and could really happen down any path.

morkalork 3y ago
Warm with a hint of ionizing radiation.
majikaja 3y ago
>we just see a cooling off of other endeavors as LLM becomes a known in terms of its strengths and weaknesses and it’s abilities reach a natural limit which is still general useful. This is my prediction If the training set is just the internet, it will be not much different than someone who spends all their lives in their room in front of their computer.
- blagie 3y ago
  I don't think that's a limit. If you give me $10M and 5 years, I can tell you a half-dozen ways to train models better than GPT4, primarily by having them do richer tasks than text completion, by having them evaluate themselves in richer ways, by having workflows around them, and by having shared models perform multiple types of tasks (e.g. text, image, controls, etc.). It's not that I'm especially smart; many others could tell you ways to do the same thing now that there's an initial proof-of-concept. It's that these things are new enough no one has had those five years yet. GPT4 came out two months ago, ChatGPT maybe seven months ago, and GPT3 three years ago. I can't predict if these things will level off, grow linearly, grow exponentially, Moore's Law style, or explode off into the singularity. I can say GPT4 is nowhere close to the limit.
  - steve_adams_86 3y ago
    There are plenty of efforts at the moment which indicate that you’re right. I think we’ll learn to compose models for more diverse and capable outputs, validate those outputs better, and do all of it more efficiently. The end result seems like it could be far greater than the sum of its parts.
  - Buttons840 3y ago
    This reminds me of reinforcement learning and the Bellman equations. Let a neural network based agent act in an environment. Record its rewards. Update the neural network so that the best actions become a little more likely, and the worse actions become a little less likely. Thus, we have demonstrated a training step that improves the agent. Now just run that training step in a loop forever and you'll improve each step and reach Godlike intelligence, right? Right? Well, no, there is instability in "pulling yourself up by the bootstraps", and while I expect to see more GTP improvements, I believe there may be limitations we don't know about yet.
    - blagie 3y ago
      That sounds like technobabble, which is to say big words like "instability" and "Bellman equations," but with little connecting logic. If there is meaning behind those words, please write something with multiple paragraphs. I don't see any reason why there would be a limit, instability, or otherwise, any more or less than there is on biological evolution. I also don't see any reason why there wouldn't be a limit. We just have no idea at this point. I do think a diversity of tasks is critical. An AGI should be able to not just complete human language but: - Play a diversity of strategic games like chess or Starcraft - Perform machine vision like Stable Diffusion - Predict the output of Python programs (and vice-versa) - Write (correct) mathematical proofs and arguments - Control a robotic arm or airplane ... and so on. We can train all of those independently -- and we could do so in a single network now -- but we're at the very early stages of architecture for how those things would integrate. At some point, the best solution is a very sophisticated intelligence. How close deep learning in current -- and future -- architectures gets us to super-intelligence is an open question.
  - 3y ago
- xbmcuser 3y ago
  I don't get all the negativity unlike most humans LLM can be made to discard bad or incorrect data. Quickly transition to new ways of doing something. A human spending all their life in a room in front of the computer would not can not go through all the millions of scientific papers and journals. They would not look at unlimited amount of code or go through all the past and present law of every country . No human is capable of that. We are entering into the real culmination of the information age where previously information was available but not exactly accessible or usable for a common person. Which like all tools will result in good and bad outcomes.
  - wg0 3y ago
    Ask any LLM about the cheese inventors in far East china. It'll make something up very plausible. Then all you have to do is to dig through libraries and historic records to verify that this all really is true and is really rooted in reality.
    - danielbln 3y ago
      That's why you are better served asking LLMs for factual information when they are connected to tools like search. ChatGPT with search enabled, Bing.chat or https://www.phind.com/ https://www.phind.com/ will give you sourced and much more reliable output.
  - majikaja 3y ago
    I'm not saying it's not useful. I'm saying that once it learns the content of the data sources that are fed to it, that's it. It will spit out some overlooked connections then stop. Maybe the cure for cancer can be found just by reading in between the lines of millions of old scientific papers but I doubt it.
wg0 3y ago
> Or maybe folks realize the degree of lies/mistruths inherent in its content is actually unmanageable and can’t be improved. After the hype wears off what it will be used for gets greatly curtailed and we see a big contraction. I think it's going to be this the future that's ahead us. There's enormous faith being put in next token predictors as the intelligence breakthrough just because their output coincidentally reassembles like something that's derived through a really intelligent process. LLMs do not hallucinate sometimes. They hallucinate all the time, it just is a coincident that sometimes these autocompletion of Tokens aligns with the reality. Just by chance, not by craft.
- maister 3y ago
  > I don't understand the "coincidentally" argument. Nothing is coincidental about those models. They were designed after processes in the brain. They underwent rigorous training to generate a function that probabilistically maps inputs to outputs. Eventually, it exceeded the threshold where most humans consider it to be intelligent. As these models grow larger, they will surpass human intelligence by far. Currently, large language models (LLMs) have fewer weights than human brains, with a difference of a factor in the thousands (based on my superficial research). But what happens when they have an equal or even 100,000 times more weights? These models will be able to model reality in ways humans cannot. Complex concepts like the connection between time and space, which are difficult for humans to grasp, will be easily understood by such models. > LLMs do not hallucinate sometimes. They hallucinate all the time, it just is a coincident that sometimes these autocompletion of Tokens aligns with the reality. Just by chance, not by craft. That is such a weird way to think about them. I'd rather say, they always provide the answer that is most probabilistic according to their internal model. Hallucination simply means, that the internal model is not good enough yet and needs to be improved, which it will.
  - pixl97 3y ago
    Heh, another one I see "LLMs don't create anything new" and "LLMs hallucinate all the time" I want to ask those people which one is the correct sentence as they appear to be in conflict with each other.
two_in_one 3y ago
> I really wonder what the next AI winter will be like Imagine nuclear winter, -50C, gray sky with barely visible sun, and robots everywhere... Good news: it's not going to happen any time soon.
red75prime 3y ago
> I really wonder what the next AI winter will be like. GPT-n writing "AI winter is coming" articles, while RecurrentGPT-n+1 helps with work on ContinualLearningRecurrentGPT-n+2.

morkalork 3y ago

Warm with a hint of ionizing radiation.

majikaja 3y ago

>we just see a cooling off of other endeavors as LLM becomes a known in terms of its strengths and weaknesses and it’s abilities reach a natural limit which is still general useful. This is my prediction If the training set is just the internet, it will be not much different than someone who spends all their lives in their room in front of their computer.

blagie 3y ago
I don't think that's a limit. If you give me $10M and 5 years, I can tell you a half-dozen ways to train models better than GPT4, primarily by having them do richer tasks than text completion, by having them evaluate themselves in richer ways, by having workflows around them, and by having shared models perform multiple types of tasks (e.g. text, image, controls, etc.). It's not that I'm especially smart; many others could tell you ways to do the same thing now that there's an initial proof-of-concept. It's that these things are new enough no one has had those five years yet. GPT4 came out two months ago, ChatGPT maybe seven months ago, and GPT3 three years ago. I can't predict if these things will level off, grow linearly, grow exponentially, Moore's Law style, or explode off into the singularity. I can say GPT4 is nowhere close to the limit.
- steve_adams_86 3y ago
  There are plenty of efforts at the moment which indicate that you’re right. I think we’ll learn to compose models for more diverse and capable outputs, validate those outputs better, and do all of it more efficiently. The end result seems like it could be far greater than the sum of its parts.
- Buttons840 3y ago
  This reminds me of reinforcement learning and the Bellman equations. Let a neural network based agent act in an environment. Record its rewards. Update the neural network so that the best actions become a little more likely, and the worse actions become a little less likely. Thus, we have demonstrated a training step that improves the agent. Now just run that training step in a loop forever and you'll improve each step and reach Godlike intelligence, right? Right? Well, no, there is instability in "pulling yourself up by the bootstraps", and while I expect to see more GTP improvements, I believe there may be limitations we don't know about yet.
  - blagie 3y ago
    That sounds like technobabble, which is to say big words like "instability" and "Bellman equations," but with little connecting logic. If there is meaning behind those words, please write something with multiple paragraphs. I don't see any reason why there would be a limit, instability, or otherwise, any more or less than there is on biological evolution. I also don't see any reason why there wouldn't be a limit. We just have no idea at this point. I do think a diversity of tasks is critical. An AGI should be able to not just complete human language but: - Play a diversity of strategic games like chess or Starcraft - Perform machine vision like Stable Diffusion - Predict the output of Python programs (and vice-versa) - Write (correct) mathematical proofs and arguments - Control a robotic arm or airplane ... and so on. We can train all of those independently -- and we could do so in a single network now -- but we're at the very early stages of architecture for how those things would integrate. At some point, the best solution is a very sophisticated intelligence. How close deep learning in current -- and future -- architectures gets us to super-intelligence is an open question.
- 3y ago
xbmcuser 3y ago
I don't get all the negativity unlike most humans LLM can be made to discard bad or incorrect data. Quickly transition to new ways of doing something. A human spending all their life in a room in front of the computer would not can not go through all the millions of scientific papers and journals. They would not look at unlimited amount of code or go through all the past and present law of every country . No human is capable of that. We are entering into the real culmination of the information age where previously information was available but not exactly accessible or usable for a common person. Which like all tools will result in good and bad outcomes.
- wg0 3y ago
  Ask any LLM about the cheese inventors in far East china. It'll make something up very plausible. Then all you have to do is to dig through libraries and historic records to verify that this all really is true and is really rooted in reality.
  - danielbln 3y ago
    That's why you are better served asking LLMs for factual information when they are connected to tools like search. ChatGPT with search enabled, Bing.chat or https://www.phind.com/ https://www.phind.com/ will give you sourced and much more reliable output.
- majikaja 3y ago
  I'm not saying it's not useful. I'm saying that once it learns the content of the data sources that are fed to it, that's it. It will spit out some overlooked connections then stop. Maybe the cure for cancer can be found just by reading in between the lines of millions of old scientific papers but I doubt it.

wg0 3y ago

> Or maybe folks realize the degree of lies/mistruths inherent in its content is actually unmanageable and can’t be improved. After the hype wears off what it will be used for gets greatly curtailed and we see a big contraction. I think it's going to be this the future that's ahead us. There's enormous faith being put in next token predictors as the intelligence breakthrough just because their output coincidentally reassembles like something that's derived through a really intelligent process. LLMs do not hallucinate sometimes. They hallucinate all the time, it just is a coincident that sometimes these autocompletion of Tokens aligns with the reality. Just by chance, not by craft.

maister 3y ago
> I don't understand the "coincidentally" argument. Nothing is coincidental about those models. They were designed after processes in the brain. They underwent rigorous training to generate a function that probabilistically maps inputs to outputs. Eventually, it exceeded the threshold where most humans consider it to be intelligent. As these models grow larger, they will surpass human intelligence by far. Currently, large language models (LLMs) have fewer weights than human brains, with a difference of a factor in the thousands (based on my superficial research). But what happens when they have an equal or even 100,000 times more weights? These models will be able to model reality in ways humans cannot. Complex concepts like the connection between time and space, which are difficult for humans to grasp, will be easily understood by such models. > LLMs do not hallucinate sometimes. They hallucinate all the time, it just is a coincident that sometimes these autocompletion of Tokens aligns with the reality. Just by chance, not by craft. That is such a weird way to think about them. I'd rather say, they always provide the answer that is most probabilistic according to their internal model. Hallucination simply means, that the internal model is not good enough yet and needs to be improved, which it will.
- pixl97 3y ago
  Heh, another one I see "LLMs don't create anything new" and "LLMs hallucinate all the time" I want to ask those people which one is the correct sentence as they appear to be in conflict with each other.

two_in_one 3y ago

> I really wonder what the next AI winter will be like Imagine nuclear winter, -50C, gray sky with barely visible sun, and robots everywhere... Good news: it's not going to happen any time soon.

red75prime 3y ago

> I really wonder what the next AI winter will be like. GPT-n writing "AI winter is coming" articles, while RecurrentGPT-n+1 helps with work on ContinualLearningRecurrentGPT-n+2.