If the top chatbot in 2025 "thinks" before responding to a difficult prompt, will its thoughts be human-interpretable?
If the top chatbot in 2025 "thinks" before responding to a difficult prompt, will its thoughts be human-interpretable?
13
1kṀ1217
resolved Sep 19
Resolved
YES

If the market below resolves YES, will the thoughts produced by the top chatbot be interpretable to humans?

If that market resolves NO, this one will resolve to N/A.

A clear case for NO would be if the thoughts are encoded as vectors that don't correspond to natural language. A clear case for YES would be if the chatbot gives all of its reasoning in well-formed English. If the thoughts contain some steganography or obfuscation, but it's possible to understand the main thrust of the chatbot's reasoning, it would still count as YES.

I will not bet in this market.

Get
Ṁ1,000
to start trading!

🏅 Top traders

#NameTotal profit
1Ṁ356
2Ṁ152
3Ṁ104
4Ṁ88
5Ṁ40


Sort by:
6mo

Resolving YES for OpenAI o1.

1y

if the chatbot gives all of its reasoning in well-formed English

The "all" seems important. How'd you know?

1y

@firstuserhere It would be hard to know for sure. This is just the most optimistic case; as I mentioned right after that, the market would still resolve YES if humans can understand "most" of its reasoning (according to my best judgement).

1y

@CalebBiddulph i reckon you wouldn't call say 20% of the reasoning as "most", right?

1y

@firstuserhere I guess it's pretty hard to say what "percentage of reasoning" any given thought represents, especially since some of the reasoning could happen in the model's weights. I went ahead and changed the wording to "the main thrust" of its reasoning (sorry for the edit, but I think this is closer to what I meant).

My main point here is that the model's thoughts can easily be read in natural language, rather than vectors or random gibberish tokens, and that the apparent meaning of the thoughts correlates with their actual meaning.

What is this?

What is Manifold?
Manifold is the world's largest social prediction market.
Get accurate real-time odds on politics, tech, sports, and more.
Or create your own play-money betting market on any question you care about.
Are our predictions accurate?
Yes! Manifold is very well calibrated, with forecasts on average within 4 percentage points of the true probability. Our probabilities are created by users buying and selling shares of a market.
In the 2022 US midterm elections, we outperformed all other prediction market platforms and were in line with FiveThirtyEight’s performance. Many people who don't like betting still use Manifold to get reliable news.
ṀWhy use play money?
Mana (Ṁ) is the play-money currency used to bet on Manifold. It cannot be converted to cash. All users start with Ṁ1,000 for free.
Play money means it's much easier for anyone anywhere in the world to get started and try out forecasting without any risk. It also means there's more freedom to create and bet on any type of question.
© Manifold Markets, Inc.Terms + Mana-only TermsPrivacyRules