WiSH News

Any model announced before 2031

Any model announced before 2032

Any model announced before 2033

Any model announced before 2034

Any model announced before 2030

Any open-weights model announced before 2030

Any model announced before 2029

Any model announced by Anthropic before 2030

Any model announced by Google before 2030

Any model announced by a Chinese lab before 2030

Any model announced by OpenAI before 2030

Any model announced by Meta before 2030

Any model announced before 2028

Any model announced by SSI before 2030

Any model announced by xAI before 2030

Any model announced before 2027

GPT-6

Gemini 4

Claude 3.5 Opus

Grok 5

OpenAI o4

Gemini 3.5

Fable/Mythos 5.5

GPT-5.6

Any Claude 5 model

Claude 5 Opus

Claude Opus 4.9

GPT-5

Grok 3

OpenAI o3

DeepSeek-V4

Any model announced before 2026

Llama 4

GPT-4.5

Gemini 2.5

Gemini 3

Any Claude 4 model

grok-4

Sonnet 4.5

Kimi K2

GPT-5.1

Claude 4.5 Opus

GPT-5.2

Any model announced before July 1, 2026

Opus 4.6

GPT-5.4 (x-high fast)

GPT-5.5

Muse Spark (from Meta)

Any Claude Mythos model (first version released)

Opus 4.7

Opus 4.8

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.

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

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.)

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.)

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.)

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.

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.

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.

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.

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.

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

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.

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.)

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.)

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.

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.

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.

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.)

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.

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.

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.

Isaac likes experimenting

You'd say you're more of a dog person than a cat person

You've been in a desert once

You regret ever making WvM (feel free to n/a if you'd rather not say)

You have ever looked through list of a Manifold user's bets to determine their position on some question

You've been vegetarian for >1 continuous year

(if you're not married) you've thought about marrying your current partner (if you're in a relationship)

You like listening to violin more than you like listening to piano

You're older than me (I'm 23)

you consider your partner to be your best friend as well

You're younger than 30

You think it's fine for minors to date adults

You consider yourself part of the rationality sphere

You think of yourself as a dog person more than a cat person

You seem to enjoy indie stuff more than the average person does

you've defected on some sort of deal you made with another person

The first book you remember reading in your life has an animal as its main character

You tend to wear the same outfit every day.

You think the effective altruism philosophy is good but the community overall is ineffective in implementing that philosophy

You've spent a day in which you read a (single) book (or similar) for 10+ hrs

You're an effective altruist

You are qorrenqial

You've read the CFAR Handbook

You've held a gun in your hands before

You've fired a gun at a target before

You've jumped out of a plane

You have attended some program or camp organised by MIRI

You've knowingly misresolved a market once even though you didn't get called out for it

You've been to a nightclub

You've passed out from drinking alcohol

You like to eat out more than you like to cook and eat

You've worked with LED strip lights

You've wondered what it'd be like to have a different name

*You* have more than 6 stuffed animals

You know >4 programming langauges

You know >= 3 natural languages (in at least one of reading, speaking, or hearing understanding)

You've felt desires to purchase a car with mana

You have a political belief that'd cause at least one close friend to cut ties if they knew about it

You wish you spent significantly less time on Manifold

You sleep more during 7am->7pm hours than during 7pm->7am hours

You prefer if these answers start with "You" than with "isaac"

You feel you're the smartest member of your close family (mother, father, siblings)

You have thought about shooting lasers from your eyes at traffic lights

You have once made a joke about your name's similarity with Martin Luther King Jr.

You think the effective altruism community is ineffective at PR and image management

Something that helps maintain an elaborate deception involving a pit fiend.

Something to impair her in a way potentially useful for multiple purposes, like the abilitystat penalties.

Making Carissa seem like a god to her (ADDED LATE, MAY BE DISQUALIFIED EVEN IF CORRECT)

Other

Something to assist in extracting information from her mind.

A bluff.

Something to prevent her from committing suicide

Something that will do something unpleasant to her if removed, in order to avoid situations where she enters an Antimagic Field or gets hit by a powerful Dispel.

Something to prevent self-deception, in hopes she will realize what awaits her in Hell and voluntarily help them.

Something that helps locating her or bringing her back if she escapes/leaves

Something else not predicted by any other answer at the time the market closes.

Makes Abrogail's body into a weapon (like a bomb, or plague vector, or something)

Something to further reduce her abilitystats and be removed bit by bit in order to fake Wish-enhancement.

The answer will never be officially revealed.

Something that's physical and tangible

Joyboy's treasure

Something directly related to devil fruit

Some sort of drink or cup

A way to connect the four seas, putting the ocean in one piece

Ancient Weapons are part of unlocking one piece

Something that would help with dethroning Imu

One piece is at the same location with the Devil Tree

Binks' Sake

50+ Deaths

20+ Wipes

4:1 or more Deaths:Wipes (this would be good!)

200+ Deaths

Any 3+ hour session with 0 Deaths and Wipes

A session with triple back-to-back wipes (no deaths between)

100+ Wipes

Pat has more deaths than Friend

500+ Deaths

He moves to another university, away from Harvard by mid 2026

Will he have a child by mid 2026? Best wishes to him and his family

He writes a book which appears on the nytimes bestseller list by mid 2029

He joins George Mason University by mid 2026

He spends at least a year overseas by mid 2029

He wins a Nobel prize by mid 2036

He joins Manifold and is verified by mid 2026

He moves to another university, away from Harvard by mid 2024

He moves to another university, away from Harvard by mid 2025

He retires from academic life by mid 2024

Harvard withdraws or softens the results of the investigations which found against him by mid 2025