If humanity is still around on Jan 1, 2030, then this market resolves to NO.

If humanity is not around but AI(s) are then this market resolves to YES.

If neither humanity nor AIs are around then this market does not resolve.

AI risk

Existential Risk

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.

Technology

This market predicts whether GPT-4 will be able to solve "easy" Sudoku puzzles 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.

Consider this market a challenge. Also see the group for this and related markets(

GPT-4 Sudoku Challenge (2023) | Manifold Markets

With at most 50 "Assistant" 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. You're allowed to drop sections from the chain when sending context to GPT-4 , as long as the decision to drop does not depend on the contents of any section.

Candidate solutions will be converted to code and run using a script(

). You are not required to interact with this script when submitting a solution, but 

 will attempt to use it to judge your solution so it may help in understanding the format.

ChatGPT 

Technical AI Timelines

OpenAI

GPT-4 speculation

Contests

Showcase

New Year's Resolutions 2024

GPT-4 Sudoku Challenge (2023)

Twitter permanently banned former president of the USA, Donald Trump, from the platform in January 2021 during the final days of his term. His handle @realDonaldTrump had over 88.9 million followers.

This market will resolve to "Yes" if Donald J. Trump's verified Twitter account (@realDonaldTrump) tweets at least once in 2023 (ET timezone). Otherwise, this market will resolve to "No".

The resolution source for this market will be Donald J. Trump's verified Twitter account: 

Please note, only the @realDonaldTrump verified Twitter account counts for this market, regardless of the URL for this profile. If Donald J. Trump tweets from another account, it has no bearing on the resolution of this market.

Donald Trump

MAGA-land

Entertainment and Pop Culture

Twitter

When researching GPT-4 speculations, the estimated parameter count ranged anywhere from 175 billion (same as GPT-3) to over 100 trillion.

This market will resolve YES if GPT-4 has over 1 trillion parameters, otherwise it will resolve NO. The market close date will be indefinitely extended until an official figure for the total number of trainable parameters of the largest released model at initial release has been announced.

😅 Today in "tailcalled makes weird markets"...

Ever since finding gender conservatism/gender progressivism as major explanatory factors in some surveys, I've been really into analyzing things as ideologies lately. Probably a good way of understanding what I mean is by contrasting to what I believed about ideologies before discovering gender conservatism/gender progressivism.

If you think of your stereotypical extreme conservative, then there are all sorts of things they've historically wanted to cancel. Gnosticism, rock and roll, divorce, gays, lesbians, boyfriends for teen girls, mild drugs, dungeons and dragons, video games, Harry Potter, therapy, etc.. Often, the justification for this cancellation is that it is a "bad influence" in some way. E.g. Harry Potter being said to promote witchcraft.

Previously, inspired by twin studies and I suspect ultimately informed by people like Steven Pinker, I had dismissed this sort of thing, with the justification being that it seems lacking in statistical evidence, that if there are supposed correlations with the claimed stuff then it's probably genetically confounded, and that it doesn't really make sense on priors. Because I suspect this answer was popularized by Pinker's The Blank Slate, I call this answer Pinkerism (though I haven't read The Blank Slate so it may be a total misnomer; I think I got it from Scott Alexander, and that Scott Alexander directly got it from The Blank Slate, and also got it from various communities that indirectly got it from The Blank Slate).

Anyway, I now reject Pinkerism, because I now see the potential for various social groups to be carriers for ideologies. Not only that, but I also think ideologies have the potential to influence actions a lot. An examples is in order. I know a guy who is heavily into PUA/macho ideology, and he considers it perfectly reasonable to trick women into sleeping with him to increase his body count. Meanwhile, I think someone who is into feminist ideology might believe that using deception to get sex is coercive. (Though the exact opinion seems to vary, since we live in a culture that performs widespread deception as a matter of course - another matter that is probably down to other ideologies.)

It is probably necessary to define the term "ideology" more precisely. Usually people associate it with classical political perspectives such as communism, conservatism and capitalism. And I agree that these (with mild clarifications) are ideologies by my definition.  But I have a very encompassing definition of ideology:

An ideology is a system of ideas, especially ideas about relevant policies, purposes, conflicts, grand societal facts, etc.. Being a system of ideas, it is socially contagious; by Aumann's agreement theorem, if you trust someone then you should adopt their ideas. And some of the ideas have some direct implications for your behavior, so the entire system of ideas can have grand effects on your behavior.

Other than classical ideologies, what other ideologies are there?

Well, consider our society's culture. Culture is basically the dominant system of ideas within a group of people. So, culture is synonymous with the dominant ideology.

Culture includes norms about how to socially interact, beliefs about how people work, and information about status relations and similar. These are all ideas, and so by my definition they constitute an ideology. And it seems to me that if someone was not aware of some of those ideas, then they would face social difficulties while interacting with others who have been encultured into them, in perhaps exactly the way autistic people do.

So strictly speaking it would not be autism that is an ideology. Rather, it would be allism that is an ideology, and autism that is its negation, reaction or ignorance. I experimented with this, talking to ChatGPT about autism symptoms and trying to identify underlying ideological drivers of allism. I came up with an 8-item scale:

Below, you see a set of statements about how communication can or should work, or related cultural opinions. For each of the statements, please evaluate the extent to which you agree with them. If you think a statement is good or true, choose agree/agree strongly. If you think a statement is bad or false, choose disagree/disagree strongly.

+ You can tell whether someone is interested by their facial expression

+ If a person who is usually out of touch is insisting that they know of something important that most people haven't noticed, people are entitled to ignore them

+ Talking with people about their weekend plans, current events and similar shows that one cares about them

+ You can learn a lot about a person by asking about things like their weekend plans

+ If you want to know whether your presence is wanted, you should not ask directly, but instead bring up a conversation (e.g. "What are you doing?"); if they simply answer the question without inviting you further, you are probably not welcome

+ If someone proposes an idea and gives an argument for it, it is rude to point out flaws in that argument, as it shows that you are opposed to their idea

+ If someone says something that sounds extreme, it is probably a misunderstanding or a joke

- Trends are not important enough that it's worth following them if it means getting mildly inconvenienced

- If one of your friends comes to you for support with a conflict they have with a stranger, you should try to figure out whether the stranger is right, rather than simply supporting your friend

- If you know a topic in-depth that someone else doesn't know, it is kind to step back and begin teaching them lots of specifics of the topic in-depth when it comes up

- If you don't have anything informative or productive to do with a person you care about, then you shouldn't be trying to hang out with them, to avoid wasting their time

(+ means that it is positively coded so allists would believe it, - means that it is negatively coded so autists would believe it. Scale order was randomized, + and - were not shown to participants.)

I call this an ideological scale because it primarily is about norms/policies (e.g. "If you want to know whether your presence is wanted, you should not ask directly, but instead bring up a conversation (e.g. "What are you doing?"); if they simply answer the question without inviting you further, you are probably not welcome") and abstract sociological facts (e.g. "You can tell whether someone is interested by their facial expression").

I tried collecting a small dataset on it, and while I haven't analyzed numerically, just eyeing the data it seems to me that the above scale was a near-complete failure. (I should probably do a proper statistical test, though.) I don't think it really measures autism, and ChatGPT suggests that it instead measures INTP (which admittedly I am pretty confused about how is different from autism, though perhaps that confusion is due to me being both INTP and autistic).

If I create an ideological scale to measure autism, and I discover it to be highly correlated with childhood diagnosis of autism (arbitrarily for market objectivity, say Cohen's d>1), then I will write up my methods in detail. If nobody pokes any serious holes in my methods, I will resolve this market YES.

If 2023 passes without me finding it to be an ideology, I will resolve this market NO. If I somehow collect the data during 2023 but the publishing/critique isn't done by the end of 2023, this market may also resolve YES.

Wait huh, didn't you just say your ideological scale didn't work?

Most strongly, I currently intend to replace ChatGPT with actual autistic people, asking them about social situations they've found confusing, and seeing if I can interpret it in terms of the dominant ideology, and construct a new scale. (If you are autistic, and you have found some social situation confusing, it would be neat if you could describe that situation in the comments! I might try to interpret it ideologically.)

Secondly, many of the items are quite abstract, so I suspect some people might find them unclear or misleading. I might choose to revise the items to make them easier to understand, which might change my original scale from not working to working.

Also maybe eyeing the data instead of running a proper statistical test was a bad idea and actually the current scale does work.

Differential Psychology

Science

Culture

If humanity is still around on Jan 1, 2040, then this market resolves to NO. Otherwise resolves to YES. 