This market predicts whether GPT-4 will be able to solve an "easy" Sudoku puzzle by December 31, 2023. I've tried some prompts, but it mostly pretends to do logical deduction  and guesses, ends up in an inconsistency(or doesn't notice), and eventually gets stuck in a loop making mistakes.

I'll use the Chat Completions API to describe this more precisely. The simplest valid pattern is:

With at most 100 entries(50 turns). The only "dynamic" input here is entry #2 which has the puzzle, and the rest is ChatGPT's responses. So this counts as a "fixed prompt" solution. You're allowed to insert more prompts into the chain after the puzzle, as long as the decision to include them or their contents do not depend on the puzzle. For example, you might have a prompt that causes ChatGPT to expand the puzzle into a set of logical constraints.

Once image support is released, images will count as an entry(a "half-turn") but since GPT-4 cannot generate image outputs you'll still have to use text to track the state. The puzzle can be given as an image, and you can have other fixed images in the pattern(such as examples for reducing sections).

The puzzle must be given with minimal interpretation. The example puzzle in the description qualifies, or minor adjustments like replacing "0" with "." or adding spaces to help tokenization. The operator can't add details like logical constraints.

ChatGPT 

Technical AI Timelines

OpenAI

Technology

GPT-4 speculation

Resolves YES if Donald Trump is arrested before January of 2024

Update: For clarification, the use of the word arrested is the legal term in the jurisdiction in which the event occurs. I'm trying to gather the definitions from each state. If you would like to assist in finding references for each state, that's helpful too. 

In most states, being arrested is being held involuntarily in custody. I suspect that once we know which state indictments occur in, we'll have more detail on the process in scope.

Politics

Donald Trump

Culture War

Eliezer Yudkowsky

A.  Humanity successfully coordinates worldwide to prevent the creation of powerful AGIs for long enough to develop human intelligence augmentation, uploading, or some other pathway into transcending humanity's window of fragility.

B.  Humanity puts forth a tremendous effort, and delays AI for long enough, and puts enough desperate work into alignment, that alignment gets solved first.

C.  Solving prosaic alignment on the first critical try is not as difficult, nor as dangerous, nor taking as much extra time, as Yudkowsky predicts; whatever effort is put forth by the leading coalition works inside of their lead time.

D.  Early powerful AGIs realize that they wouldn't be able to align their own future selves/successors if their intelligence got raised further, and work honestly with humans on solving the problem in a way acceptable to both factions.

E.  Whatever strange motivations end up inside an unalignable AGI, or the internal slice through that AGI which codes its successor, they max out at a universe full of cheerful qualia-bearing life and an okay outcome for existing humans.

F.  Somebody pulls off a hat trick involving blah blah acausal blah blah simulations blah blah, or other amazingly clever idea, which leads an AGI to put the reachable galaxies to good use despite that AGI not being otherwise alignable.

G.  It's impossible/improbable for something sufficiently smarter and more capable than modern humanity to be created, that it can just do whatever without needing humans to cooperate; nor does it successfully cheat/trick us.

H.  Many competing AGIs form an equilibrium whereby no faction is allowed to get too powerful, and humanity is part of this equilibrium and survives and gets a big chunk of cosmic pie.

I.  The tech path to AGI superintelligence is naturally slow enough and gradual enough, that world-destroyingly-critical alignment problems never appear faster than previous discoveries generalize to allow safe further experimentation.

J.  Something 'just works' on the order of eg: train a predictive/imitative/generative AI on a human-generated dataset, and RLHF her to be unfailingly nice, generous to weaker entities, and determined to make the cosmos a lovely place.

K.  Somebody discovers a new AI paradigm that's powerful enough and matures fast enough to beat deep learning to the punch, and the new paradigm is much much more alignable than giant inscrutable matrices of floating-point numbers.

L.  Earth's present civilization crashes before powerful AGI, and the next civilization that rises is wiser and better at ops.  (Exception to 'okay' as defined originally, will be said to count as 'okay' even if many current humans die.)

M.  "We'll make the AI do our AI alignment homework" just works as a plan.  (Eg the helping AI doesn't need to be smart enough to be deadly; the alignment proposals that most impress human judges are honest and truthful and successful.)

N.  A crash project at augmenting human intelligence via neurotech, training mentats via neurofeedback, etc, produces people who can solve alignment before it's too late, despite Earth civ not slowing AI down much.

O.  Early applications of AI/AGI drastically increase human civilization's sanity and coordination ability; enabling humanity to solve alignment, or slow down further descent into AGI, etc.  (Not in principle mutex with all other answers.)

If you write an argument that breaks down the 'okay outcomes' into lots of distinct categories, without breaking down internal conjuncts and so on, Reality is very impressed with how disjunctive this sounds and allocates more probability.

You are fooled by at least one option on this list, which out of many tries, ends up sufficiently well-aimed at your personal ideals / prejudices / the parts you understand less well / your own personal indulgences in wishful thinking.

Something wonderful happens that isn't well-described by any option listed.  (The semantics of this option may change if other options are added.)

An outcome is "okay" if it gets at least 20% of the maximum attainable cosmopolitan value that could've been attained by a positive Singularity (a la full Coherent Extrapolated Volition done correctly), and existing humans don't suffer death or any other awful fates.

https://manifold.markets/IsaacKing/if-we-survive-general-artificial-in

https://manifold.markets/EliezerYudkowsky/if-artificial-general-intelligence-539844cd3ba1?r=RWxpZXplcll1ZGtvd3NreQ

 is this same question but with user-submitted answers.

(Please note:  It's a known cognitive bias that you can make people assign more probability to one bucket over another, by unpacking one bucket into lots of subcategories, but not the other bucket, and asking people to assign probabilities to everything listed.  This is the disjunctive dual of the Multiple Stage Fallacy, whereby you can unpack any outcome into a big list of supposedly necessary conjuncts that you ask people to assign probabilities to, and make the final outcome seem very improbable.

So:  That famed fiction writer Eliezer Yudkowsky can rationalize at least 15 different stories (options 'A' through 'O') about how things could maybe possibly turn out okay; and that the option texts don't have enough room to list out all the reasons each story is unlikely; and that you get 15 different chances to be mistaken about how plausible each story sounds; does not mean that Reality will be terribly impressed with how disjunctive the okay outcome bucket has been made to sound.  Reality need not actually allocate more total probability into all the okayness disjuncts listed, from out of all the disjunctive bad ends and intervening difficulties not detailed here.)

https://futureoflife.org/open-letter/pause-giant-ai-experiments/

https://garymarcus.substack.com/p/agi-will-not-happen-in-your-lifetime

Resolves to YES if it happens. Otherwise I'll leave this open for a few weeks after the end of 2023, and if I haven't found any credible reporting of such a thing by then, I'll resolve to NO.

Could include a direct death from a language model being given physical control over some mechanism, or an indirect death from something like a language model providing bad advice and someone acting on that advice and getting themselves killed, committing suicide after a language model told them to do so, or something like that.

Does not include things that aren't clearly the language model's "fault", such as someone bombing an AI research lab, someone crashing their car because they were distracted talking to a language model, a model giving someone a vague idea that they investigate themselves and eventually leads to their death months later, etc.

"Large" is vague, so I'll ignore that qualifer and count any language model.

Gary Marcus GPT-4 predictions

 called for a moratorium on the development of increasingly powerful artificial intelligence (AI) systems, highlighting the potential risks to society and humanity. The letter urged all AI labs to pause the training of AI systems more powerful than GPT-4, including GPT-5, for at least 6 months. It also called for shared safety protocols and improved AI governance. Sam Altman, the CEO of OpenAI, is a prominent figure in the AI community and his support for the open letter could influence its impact and reception.

Before April 5th, 2023, will it be verified that Sam Altman signed the recent open letter calling for a moratorium on the development of AI systems more powerful than GPT-4 (including GPT-5)?

This question will resolve positively if, before April 5th, 2023, credible news sources, official statements, or other verifiable evidence confirm that Sam Altman has signed the open letter, meeting the following criteria:

If credible news sources, official statements, or other verifiable evidence confirm that Sam Altman has signed the open letter meeting the above criteria before April 5th 2023, the question will resolve positively. If no such evidence is provided by the deadline, the question will resolve negatively.

AI Safety

This resolves to YES if Tiktok is banned in the US or forced to change ownership by any US government action by the closing date (Dec 31 2023). This includes any US government action that makes it more difficult for TikTok to operate in the US without changing ownership but does not include the case where TikTok preemptively decides to change ownership without any actual action on the part of any US government action. "US government action" is defined to include both Federal and State-level actions including changes in rules or legislation, executive orders, and lawsuits brought by a US government entity. This also includes federal-level audits and US government changes to the economic playing field that are specifically targetted at TikTok. The edge case where legislation has been set in motion but has not passed yet counts as YES if it seems like it has the votes to pass and NO if it doesn't (based off of my best assessment at the time). "banned in the US" means banned for all users in the US (so for example only for one state or only for government employees doesn't meet the criterion).

Economics

TikTok

I will resolve this based on some combination of how much it gets talked about in elections, how much money goes to interest groups on both topics, and how much of the "political conversation" seems to be about either.

Scott Alexander's 5 year predictions

 called for a moratorium on the development of increasingly powerful artificial intelligence (AI) systems, highlighting the potential risks to society and humanity. The letter urged all AI labs to pause the training of AI systems more powerful than GPT-4, including GPT-5, for at least 6 months. It also called for shared safety protocols and improved AI governance. The letter suggested that if the AI labs do not enact a pause voluntarily, governments should step in and institute a moratorium.

Before January 1st, 2024, will any of the top AI labs or governments commit to a moratorium on the development of AI systems beyond a certain capability level?

This question will resolve positively if, before January 1st, 2024, credible news sources, official statements, or legal documents confirm that at least one of the top AI labs or governments has committed to a moratorium on the development of AI systems more powerful than GPT-4, including GPT-5, meeting all of the following criteria:

If credible news sources, official statements, or legal documents confirm that at least one of the top AI labs or governments has committed to a moratorium meeting all of the above criteria before January 1st, 2024, the question will resolve positively. If no such commitment is made by the deadline, the question will resolve negatively.

Note: This question focuses on the commitment to a moratorium, not the actual implementation or enforcement of the moratorium or the development of safety protocols and governance systems. Additionally, the question is not concerned with the potential impact of the moratorium on AI research, technological progress, or society at large.

Science

Something a-la ChatGPT, as in it has to be able to conversational and open domain. Google Assistant doesn't count primarily because you can't ask it stuff like "write a Shakespearean sonnet about Bill Gates".

 homepage. Has to be open to general public, and easily accessible from Google search home page. Not something you have to get whitelisted for, be in a small experiment or meticulously look for in some settings dialog.

Separate app counts, as long as it's promoted on 

 homepage to the general public. "Random 1% of users get an invite" doesn't count. "All US users get access but it's not in other locales" does count.

If Assistant (as it's integrated into phones, smart speakers etc.) were to just quietly get this capability, but it weren't promoted on 

, that wouldn't count. If it were promoted, it would.

Gracefully winding down doesn't count, it must be somewhat catastrophic, similar to FTX. Happening unexpectedly in a short time frame, leaving other people at a disadvantage, etc.

Cryptocurrency

Binance

If you're required to pay to use Twitter as it currently exists (even if you have some free time), this market will be solved to YES.