Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Prompting Claude Opus 5.5 (claude.com)
    —discuss
  2. Owed a billion dollars in Nvidia stock (colo.to)
    259comments
  3. Thinking fast and slow in AI: The role of metacognition (2021) (arxiv.org)
    22comments
  4. Ember-1 (fireworks.ai)
    206comments
  5. When did Google get so weird? (sancho.bearblog.dev)
    670comments
  6. Malleable software: Restoring user agency in a world of locked-down apps (2025) (inkandswitch.com)
    28comments
  7. Functional Mechanical Sympathy [video] (youtube.com)
    —discuss
  8. Guitar amp and effects pedal built on the Waveshare ESP32-S3-Touch-AMOLED-2.06 (github.com/dashersw)
    25comments
  9. Made by Mechanical Means (felixrieseberg.com)
    1comments
  10. Alan Kay's answer to “Did the ENIAC have a BIOS”? (quora.com)
    39comments
  11. Self-Hosting on the Dark Web (alvarezrosa.com)
    54comments
  12. Was silent reading unusual during Augustine's time? (historyofinformation.com)
    22comments
  13. There is more to code review than (automatable) detection (adaptivecapacitylabs.com)
    73comments
  14. Don't couple your Go code to GitHub (iain.rocks)
    104comments
  15. The state of SIMD in Rust in 2026 (shnatsel.github.io)
    29comments
  16. Deterministic Concurrency [video] (youtube.com)
    2comments
  17. Show HN: Lofi Cities – Pixel-art city nights with browser-generated lofi (loficities.com)
    108comments
  18. Maybe don't let Muse run your Facebook Marketplace account (threads.com)
    —discuss
  19. Footguns with Postgres "at time zone 'UTC'" (bookofrevenue.com)
    —discuss
  20. What I did at Recurse Center (thill.me)
    32comments
  21. Reading’s Bayeux Tapestry (diamondgeezer.blogspot.com)
    13comments
  22. Imp is a full port of DSPy to the BEAM (github.com/deepfates)
    7comments
  23. In an $80 motel room, a discovery to shed light on the origins of life (nytimes.com)
    92comments
  24. Replacing the old battery on rechargeable bike lights (jvns.ca)
    90comments
  25. An Antidote to Roko's Basilisk
    11comments
  26. Previously unheard recordings of John Coltrane, captured by Frank Tiberi (jazzwise.com)
    33comments
  27. Writing Efficient C++ Code (2013) (asawicki.info)
    116comments
  28. Oral history of John Chowning, inventor of FM synthesis [video] (youtube.com)
    21comments
  29. Packing Binary Is Fun (hereticpleb.vercel.app)
    3comments
  30. A New Experiment Meta-Strategy (chillphysicsenjoyer.substack.com)
    1comments

Thinking fast and slow in AI: The role of metacognition (2021)

83 pointsby 5h agoarxiv.org
22 comments
4h agoHN ↗

submitted Oct 5 2021

(In case people miss that before discussion)

3h agoHN ↗

This is still a great paper, but it's missing the second axis of the quadric -- if the only two options are thinking fast or thinking about thinking, that leaves no room for thinking slow yet deliberately, AKA selfconsciousness. See https://www.gutenberg.org/cache/epub/4280/pg4280-images.html for details

I do wonder if any of these folks ever got a chance to try this at one of the big labs, tho...

2h agoHN ↗

It looks like a lot like how data bases query optimizers work, with the exception that in the paper there is also a learning/memory component that conditions the evaluation of the answer provided by the first model.

2h agoHN ↗

It reminds me of "At the time we drew boxes labeled 'perception', 'cognition' with arrows between them." An imprecise quote that I can't place.

I guess my box labelled 'subconsciousness' is trying to say that low-level mechanisms that give rise to the observed cognitive phenomena might have nothing to do with neat boxes.

17m agoHN ↗

McDermott's A Critique of Pure Reason pretty much captured all of the misgivings I had about "Good Old-Fashioned Artificial Intelligence", which was slightly unfortunate as I was trying to complete a PhD in that very area at the time (around 1990 or so...)

2h agoHN ↗

2021. Please remember the rule of HN to add the year if it’s not actual.

2h agoHN ↗

This has already been solved by GPT 5 Adaptive reasoning. A single model that knows when to reason or not based on a thinking parameter we provide (like xhigh). What’s the relevancy to post it today?

edit: why is this downvoted?

57m agoHN ↗

This has already been solved by GPT 5 Adaptive reasoning. A single model that knows when to reason or not based on a thinking parameter we provide (like xhigh). What’s the relevancy to post it today?

Tell me you didn't read Daniel Khaneman's book without telling me you didn't read Daniel Khaneman's book.

52m agoHN ↗

Asking earnestly, I don’t know what you mean by this reply. I know what system 1 and 2 is. But this has already been solved using same model.

33m agoHN ↗

I know what system 1 and 2 is. But this has already been solved using same model.

No, it hasn't. Maybe you have a different definition of System 1 and System 2. I last read the book well over a decade ago (2011, maybe? 2012?), but System 1 and System 2 are different systems. IOW, System 2 is not a more computational version of System 1.

The argument you made implies that System 2 is just a more capable System 1, which is not what the book (nor this paper, AIUI) proposes.

In computery terms, System 1 runs in O(1) time, System 2 runs in O(log n) (or maybe just O(n)) time.

This means that any System 1 will run the input once through the heuristics, using the same computational power and taking the same time whether the input is 100 tokens or 1 million tokens, for quick but perhaps wrong decision (not "answer"). We don't have LLMs that do that. We have System 2 - run in O(log n) time and produce an answer.

System 1 is completely bereft of thought.

27m agoHN ↗

The argument you made implies that System 2 is just a more capable System 1, which is not what the book (nor this paper, AIUI) proposes.

No, system 2 is the emergent capability to reason and increase the space of places to find the answer. Forget the paper's proposal, and look at the problem it is trying to solve. Ability to give quick answers, ability to give thought out answers, and the ability to know when to choose what. Adaptive reasoning does all three.

This means that any System 1 will run the input once through the heuristics, using the same computational power and taking the same time whether the input is 100 tokens or 1 million tokens, for quick but perhaps wrong decision (not "answer").

No, I don't think we humans use o(1) to for understanding 1000 tokens or 2 tokens. I simply don't think that's the case. There's a new model called "Jev" and it is literally named System 1 (from the book) and even it is billed per input token.

35m agoHN ↗

It's being downvoted, I think, for a few reasons:

* The person who posted it likely posted it not as an out-of-date paper but as an interesting idea. Your comment ignores the idea and focuses on what you're calling its out-of-dateness.

* You say "this has been solved" without defining what "this" is.

* Your description of the solution -- different effort levels -- seems to indicate that you misunderstand the idea that the paper is proposing. If I understand their proposal, it's that the system itself decides how to reason based on the nature of the problem it faces, given the model's world model and past experience. "Effort" isn't so much the issue as types of effort using different systems, modeled specifically after Kahneman's idea of fast and slow thinking.

* The title is an allusion to a book by Daniel Kahneman. The brisk dismissal without acknowledging the idea or the history doesn't leave a good impression, even if I'm mistaken and you're right.

In short, Hacker News readers tend to reward depth and detail (the FAQ specifically encourages thoughtful contributions and explicitly discourages dismissal). Your comment doesn't provide them, and it appears to make a mistake that further undermines its value as a contribution to discussion.

34m agoHN ↗

it's that the system itself decides how to reason based on the nature of the problem it faces, given the model's world model.

do you even know how adaptive reasoning works?

27m agoHN ↗

Godspeed to you in your efforts to contribute productively to Hacker News threads.

24m agoHN ↗

For the first time, GPT‑5.1 Instant can use adaptive reasoning to decide when to think before responding to more challenging questions, resulting in more thorough and accurate answers, while still responding quickly. This is reflected in significant improvements on math and coding evaluations like AIME 2025 and Codeforces.

https://openai.com/index/gpt-5-1/

It says literally the thing you wanted from system 2. Its almost exactly that.

This is what you said btw:

"it's that the system itself decides how to reason based on the nature of the problem it faces"

59m agoHN ↗

How relevant is this fast/slow thinking thing with regards to current frontier models?

I know a large organization who's built their AI framework completely around this concept, and I feel that it's not really meaningful concept with the capabilities of current models.

33m agoHN ↗

You can ask a model for output directly and stop, or you can recursively ask it to keep refining the output.

That seems to fit the fast vs slow model of human thought reasonably well.

26m agoHN ↗

Trying to get LLMs to 'think about their thinking' is my daily struggle. This paper nails why it's so critical.