4 ms·

There is some magical thinking at play, that once you have what appears to be an intelligence of some kind, simply allowing it to recursively self-critique will

There is some magical thinking at play, that once you have what appears to be an intelligence of some kind, simply allowing it to recursively self-critique will lead to singularity, but the other possibility is the signal just degrades, like deep-fried jpgs.

corobo 4y ago
My testing would definitely lean towards deep fried jpgs aye. Every time it seemed to hit a bump somewhere then got caught up in a loop of googling whatever problem it was having and not finding an answer it was looking for, then going on to look into that problem, etc, etc. I guess you could put a step in to allow the human to nudge the AI back on track but then it's not really autonomous. Also it's really slow at doing anything, even before it goes off into the weeds. I understand it's new and wont be polished but man. Really slow. At least it's a decent motivator! "Oooh this is a task AutoGPT/BabyAGI could do... ehh it'd be quicker to just do it myself" I'll keep an open mind, still eager to see a successful attempt at something. That'd at least open up a chance I'm doing it wrong rather than it not doing anything useful.
Buttons840 4y ago
This has been reinforcement learning for decades. People say "OMG! The agent is acting and learning and getting better all on its own!" And they're right, in the beginning. Eventually the agent plateaus or collapses. Supervised learning is much more stable. GTP is supervised learning. Once you start letting the agent choose or modify its own training data, then you're moving towards the much less stable realms of reinforcement learning.
- AnimalMuppet 4y ago
  Well... humans who pick their own "training data" can wind up in echo chambers and, shall we say, diverging from reality. I don't think we're on the threshold of AGI, but it's interesting that the wanna-be AIs are running into human issues...
  - Buttons840 4y ago
    True. I've read a lot about reinforcement learning and the exploration vs exploitation tradeoff is both technically interesting and philosophically interesting. I've also seen algorithms I've implemented fall to all sorts of human-ish flaws, such as learned helplessness, etc.
  - qualudeheart 4y ago
    Humans who pick their own training data can make greats results by chosing which courses at college or grad school to attend.
- dontwearitout 4y ago
  GPT is not supervised learning, it is self-supervised learning (a form of unsupervised learning) since the learning objective is "fill in the missing training data".

corobo 4y ago

My testing would definitely lean towards deep fried jpgs aye. Every time it seemed to hit a bump somewhere then got caught up in a loop of googling whatever problem it was having and not finding an answer it was looking for, then going on to look into that problem, etc, etc. I guess you could put a step in to allow the human to nudge the AI back on track but then it's not really autonomous. Also it's really slow at doing anything, even before it goes off into the weeds. I understand it's new and wont be polished but man. Really slow. At least it's a decent motivator! "Oooh this is a task AutoGPT/BabyAGI could do... ehh it'd be quicker to just do it myself" I'll keep an open mind, still eager to see a successful attempt at something. That'd at least open up a chance I'm doing it wrong rather than it not doing anything useful.

Buttons840 4y ago

This has been reinforcement learning for decades. People say "OMG! The agent is acting and learning and getting better all on its own!" And they're right, in the beginning. Eventually the agent plateaus or collapses. Supervised learning is much more stable. GTP is supervised learning. Once you start letting the agent choose or modify its own training data, then you're moving towards the much less stable realms of reinforcement learning.

AnimalMuppet 4y ago
Well... humans who pick their own "training data" can wind up in echo chambers and, shall we say, diverging from reality. I don't think we're on the threshold of AGI, but it's interesting that the wanna-be AIs are running into human issues...
- Buttons840 4y ago
  True. I've read a lot about reinforcement learning and the exploration vs exploitation tradeoff is both technically interesting and philosophically interesting. I've also seen algorithms I've implemented fall to all sorts of human-ish flaws, such as learned helplessness, etc.
- qualudeheart 4y ago
  Humans who pick their own training data can make greats results by chosing which courses at college or grad school to attend.
dontwearitout 4y ago
GPT is not supervised learning, it is self-supervised learning (a form of unsupervised learning) since the learning objective is "fill in the missing training data".