3 ms·

If that's true, then we're following the timeline of https://ai-2027.com/ https://ai-2027.com/

by haellsigh 6mo ago

matltc 6mo ago
Care to expound on that? Maybe a reference to the relevant section?
- 9991 6mo ago
  You should just read the thing, whether or not you believe it, to have an informed opinion on the ongoing debate.
  - matltc 6mo ago
    I did read it a while back. Was curious what parent was referring to specifically
    - 9991 6mo ago
      March 2027 -> Neuralese recurrence and memory > For example, perhaps models will be trained to think in artificial languages that are more efficient than natural language but difficult for humans to interpret.
- ACCount37 6mo ago
  Ctrl-F "neuralese" on that page.
9991 6mo ago
That's not supposed to happen til 2027. Ruh roh.
literalAardvark 6mo ago
Only if you ignore context and just ctrl-f in the timeline. What are you, Haiku? But yeah, in many ways we're at least a year ahead on that timeline.
butlike 6mo ago
Hilariously, I clicked back a bunch and got a client side error. We have a long way to go. I wouldn't worry about it.
magicalist 6mo ago
> If that's true, then we're following the timeline Literally just a citation of Meta's Coconut paper[1]. Notice the 2027 folk's contribution to the prediction is that this will have been implemented by "thousands of Agent-2 automated researchers...making major algorithmic advances". So, considering that the discussion of latent space reasoning dates back to 2022[2] through CoT unfaithfulness, looped transformers, using diffusion for refining latent space thoughts, etc, etc, all published before ai 2027, it seems like to be "following the timeline of ai-2027" we'd actually need to verify that not only was this happening, but that it was implemented by major algorithmic advances made by thousands of automated researchers, otherwise they don't seem to have made a contribution here. [1] https://ai-2027.com/#:~:text=Figure%20from%20Hao%20et%20al.%2C%20a%202024%20paper%20from%20Meta%20implementing%20this%20idea. https://ai-2027.com/#:~:text=Figure%20from%20Hao%20et%20al.%... [2] https://arxiv.org/html/2412.06769v3#S2 https://arxiv.org/html/2412.06769v3#S2

If that's true, then we're following the timeline of https://ai-2027.com/ https://ai-2027.com/

matltc 6mo ago

Care to expound on that? Maybe a reference to the relevant section?

9991 6mo ago
You should just read the thing, whether or not you believe it, to have an informed opinion on the ongoing debate.
- matltc 6mo ago
  I did read it a while back. Was curious what parent was referring to specifically
  - 9991 6mo ago
    March 2027 -> Neuralese recurrence and memory > For example, perhaps models will be trained to think in artificial languages that are more efficient than natural language but difficult for humans to interpret.
ACCount37 6mo ago
Ctrl-F "neuralese" on that page.

9991 6mo ago

That's not supposed to happen til 2027. Ruh roh.

literalAardvark 6mo ago

Only if you ignore context and just ctrl-f in the timeline. What are you, Haiku? But yeah, in many ways we're at least a year ahead on that timeline.

butlike 6mo ago

Hilariously, I clicked back a bunch and got a client side error. We have a long way to go. I wouldn't worry about it.

magicalist 6mo ago

> If that's true, then we're following the timeline Literally just a citation of Meta's Coconut paper[1]. Notice the 2027 folk's contribution to the prediction is that this will have been implemented by "thousands of Agent-2 automated researchers...making major algorithmic advances". So, considering that the discussion of latent space reasoning dates back to 2022[2] through CoT unfaithfulness, looped transformers, using diffusion for refining latent space thoughts, etc, etc, all published before ai 2027, it seems like to be "following the timeline of ai-2027" we'd actually need to verify that not only was this happening, but that it was implemented by major algorithmic advances made by thousands of automated researchers, otherwise they don't seem to have made a contribution here. [1] https://ai-2027.com/#:~:text=Figure%20from%20Hao%20et%20al.%2C%20a%202024%20paper%20from%20Meta%20implementing%20this%20idea. https://ai-2027.com/#:~:text=Figure%20from%20Hao%20et%20al.%... [2] https://arxiv.org/html/2412.06769v3#S2 https://arxiv.org/html/2412.06769v3#S2