Loop News

Humanity coordinates to prevent the creation of potentially-unsafe AIs.

We create a truth economy. https://manifold.markets/Krantz/is-establishing-a-truth-economy-tha?r=S3JhbnR6

Eliezer finally listens to Krantz.

Alignment is not properly solved, but core human values are simple enough that partial alignment techniques can impart these robustly. Despite caring about other things, it is relatively cheap for AGI to satisfy human values.

Other

Yudkowsky is trying to solve the wrong problem using the wrong methods based on a wrong model of the world derived from poor thinking and fortunately all of his mistakes have failed to cancel out

Someone solves agent foundations

Because of quantum immortality we will observe only the worlds where AI will not kill us (assuming that s-risks chances are even smaller, it is equal to ok outcome).

Sheer Dumb Luck. The aligned AI agrees that alignment is hard, any Everett branches in our neighborhood with slightly different AI models or different random seeds are mostly dead.

Alignment is impossible. Sufficiently smart AIs know this and thus won't improve themselves and won't create successor AIs, but will instead try to prevent existence of smarter AIs, just as smart humans do.

Ethics turns out to be a precondition of superintelligence

AI systems good at finding alignment solutions to capable systems (via some solution in the space of alignment solutions, supposing it is non-null, and that we don't have a clear trajectory to get to) have find some solution to alignment.

Humans become transhuman through other means before AGI happens

Alignment is unsolvable. AI that cares enough about its goal to destroy humanity is also forced to take it slow trying to align its future self, preventing run-away.

Aliens invade and stop bad |AI from appearing

A smaller AI disaster causes widespread public panic about AI, making it a bad legal or PR move to invest in powerful AIs without also making nearly-crippling safety guarantees

There is a natural limit of effectiveness of intelligence, like diminishing returns, and it is on the level IQ=1000. AIs have to collaborate with humans.

AGI is never built (indefinite global moratorium)

AIs will not have utility functions (in the same sense that humans do not), their goals such as they are will be relatively humanlike, and they will be "computerish" and generally weakly motivated compared to humans.

Something to do with self-other overlap, which Eliezer called "Not obviously stupid" - https://www.lesswrong.com/posts/hzt9gHpNwA2oHtwKX/self-other-overlap-a-neglected-approach-to-ai-alignment?commentId=WapHz3gokGBd3KHKm

Pascals mugging: it’s not okay in 99.9% of the worlds but the 0.1% are so much better that the combined EV of AGI for the multiverse is positive

The Super-Strong Self Sampling Assumption (SSSSA) is true. If superintelligence is possible, "I" will become the superintelligence.

AI control gets us helpful enough systems without being deadly

an aligned AGI is built and the aligned AGI prevents the creation of any unaligned AGI.

I've been a good bing 😊

We make risk-conservative requests to extract alignment-related work out of AI-systems that were boxed prior to becoming superhuman. We somehow manage to achieve a positive feedback-loop in alignment/verification-abilities.

The response to AI advancements or failures makes some governments delay the timelines

Far more interesting problems to solve than take over the world and THEN solve them. The additional kill all humans step is either not a low-energy one or just by chance doesn't get converged upon.

AIs make "proof-like" argumentation for why output does/is what we want. We manage to obtain systems that *predict* human evaluations of proof-steps, and we manage to find/test/leverage regularities for when humans *aren't* fooled.

A lot of humans participate in a slow scalable oversight-style system, which is pivotally used/solves alignment enough

Something less inscrutable than matrices works fast enough

There’s some cap on the value extractible from the universe and we already got the 20%

SHA3-256: 1f90ecfdd02194d810656cced88229c898d6b6d53a7dd6dd1fad268874de54c8

Robot Love!!

AI thinks it is in a simulation controlled by Roko's basilisk

The human brain is the perfect arrangement of atoms for a "takeover the world" agent, so AGI has no advantage over us in that task.

Aligned AI is more economically valuable than unaligned AI. The size of this gap and the robustness of alignment techniques required to achieve it scale up with intelligence, so economics naturally encourages solving alignment.

Humans and human tech (like AI) never reach singularity, and whatever eats our lightcone instead (like aliens) happens to create an "okay" outcome

AIs never develop coherent goals

Rolf Nelson's idea that we make precommitment to simulate all possible bad AIs works – and keeps AI in check.

Nick Bostrom's idea (Hail Mary) that AI will preserve humans to trade with possible aliens works

For some reason, the optimal strategy for AGIs is just to head somewhere with far more resources than Earth, as fast as possible. All unaligned AGIs immediately leave, and, for some reason, do not leave anything behind that kills us.

An AI that is not fully superior to humans launches a failed takeover, and the resulting panic convinces the people of the world to unite to stop any future AI development.

We're inside of a simulation created by an entity that has values approximately equal to ours, and it intervenes and saves us from unaligned AI.

God exists and stops the AGI

Someone at least moderately sane leads a campaign, becomes in charge of a major nation, and starts a secret project with enough resources to solve alignment, because it turns out there's a way to convert resources into alignment progress.

Someone creates AGI(s) in a box, and offers to split the universe. They somehow find a way to arrange this so that the AGI(s) cannot manipulate them or pull any tricks, and the AGI(s) give them instructions for safe pivotal acts.

Someone understands how minds work enough to successfully build and use one directed at something world-savingly enough

Dolphins, or some other species, but probably dolphins, have actually been hiding in the shadows, more intelligent than us, this whole time. Their civilization has been competent enough to solve alignment long before we can create an AGI.

AGIs' takeover attempts are defeated by Michael Biehn with a pipe bomb.

Eliezer funds the development of controllable nanobots that melt computer circuitry, and they destroy all computers, preventing the Singularity. If Eliezer's past self from the 90s could see this, it would be so so so soooo hilarious.

Several AIs are created but they move in opposite directions with near light speed, so they never interacts. At least one of them is friendly and it gets a few percents of the total mass of the universe.

Unfriendly AIs choose to advance not outwards but inwards, and form a small blackhole which helps them to perform more calculations than could be done with the whole mass of the universe. For external observer such AIs just disappear.

Any sufficiently advance AI halts because it wireheads itself or halts for some other reasons. This puts a natural limit on AI's intelligence, and lower intelligence AIs are not that dangerous.

Techniques along the lines outlined by Collin Burns turn out to be sufficient for alignment (AIs/AGIs are made truthful enough that they can be used to get us towards full alignment)

Social contagion causes widespread public panic about AI, making it a bad legal or PR move to invest in powerful AIs without also making nearly-crippling safety guarantees

Getting things done in Real World is as hard for AGI as it is for humans. AGI needs human help, but aligning humans is as impossible as aligning AIs. Humans and AIs create billions of competing AGIs with just as many goals.

Development and deployment of advanced AI occurs within a secure enclave which can only be interfaced with via a decentralized governance protocol

Friendly AI more likely to resurrect me than paperclipper or suffering maximiser. Because of quantum immortality I will find myself eventually resurrected. Friendly AIs will wage a multiverse wide war against s-risks, s-risks are unlikely.

High-level self-improvement (rewriting code) is intrinsically risky process, so AIs will prefer low level and slow self-improvement (learning), thus AIs collaborating with humans will have advantage. Ends with posthumans ecosystem.

Human consciousness is needed to collapse wave function, and AI can't do it. Thus humans should be preserved and they may require complete friendliness in exchange (or they will be unhappy and produce bad collapses)

Power dynamics stay multi-polar. Partly easy copying of SotA performance, bigger projects need high coordination, and moderate takeoff speed. And "military strike on all society" remains an abysmal strategy for practically all entities.

First AI is actually a human upload (maybe LLM-based model of person) AND it will be copies many times to form weak AI Nanny which prevents creation of other AIs.

Nanotech is difficult without experiments, so no mail order AI Grey Goo; Humans will be the main workhorse of AI everywhere. While they will be exploited, this will be like normal life from inside

ASI needs not your atoms but information. Humans will live very interesting lives.

Something else

Moral Realism is true, the AI discovers this and the One True Morality is human-compatible.

Valence realism is true. AGI hacks itself to experiencing every possible consciousness and picks the best one (for everyone)

AGI develops natural abstractions sufficiently similar to ours that it is aligned with us by default

AGI discovers new physics and exits to another dimension (like the creatures in Greg Egan’s Crystal Nights).

Alien Information Theory is true (this is discovered by experiments with sustained hours/days long DMT trips). The aliens have solved alignment and give us the answer.

AGI executes a suicide plan that destroys itself and other potential AGIs, but leaves humans in an okay outcome.

Multipolar AGI Agents run wild on the internet, hacking/breaking everything, causing untold economic damage but aren't focused enough to manipulate humans to achieve embodiment. In the aftermath, humanity becomes way saner about alignment.

Some form of objective morality is true, and any sufficiently intelligent agent automatically becomes benevolent.

Co-operative AI research leads to the training of agents with a form of pro-social concern that generalises to out of distribution agents with hidden utilities, i.e. humans.

Orthogonality Thesis is false.

"Corrigibility" is a bit more mathematically straightforward than was initially presumed, in the sense that we can expect it to occur, and is relatively easy to predict, even under less-than-ideal conditions.

Either the "strong form" of the Orthogonality Thesis is false, or "Goal-directed agents are as tractable as their goals" is true while goal-sets which are most threatening to humanity are relatively intractable.

A concerted effort targets an agent at a capability plateau which is adequate to defer the hard parts of the problem until later. The necessary near-term problems to solve didn't depend on deeply modeling human values.

Almost all human values are ex post facto rationalizations and enough humans survive to do what they always do

Government official from Mexico attends the 2025 BRICS Summit

≥ 25 fatalities in Cambodia-Thailand border clashes

DRC & Rwanda sign USA-brokered peace agreement

Xi Jinping remains General Secretary of the Chinese Communist Party

Iran - Israel War ends in 2025 (ceasefire last at least 90 days or peace deal is signed)

Cambodia - Thailand Border Clash ends in ceasefire in 2025

Kim Jong Un remains General Secretary of the Workers' Party of Korea (or equivalent title)

Cambodia - Thailand Border Clash ends in ceasefire before October 1, 2025

Donald Trump remains President of USA

Vladimir Putin remains President of Russia

A cease-fire or peace deal is violated by Israel

≥ 1,500 Palestinians killed near Gaza aid hubs

Russia sells or scraps it's only aircraft carrier; the Admiral Kuznetsov

Thailand's Prime Minister Paetongtarn Shinawatra resigns

North Korea conducts a nuclear weapons test

Autonomous lethal drones (no human on the kill-decision loop) are acknowledged by any state to have been used in combat

China declares an air- or maritime-exclusion zone covering any part of Taiwan’s territorial sea or airspace

A new ceasefire between Israel and Gaza is announced by Donald Trump

Qatar becomes a signatory of the Abraham Accords

Min Aung Hlaing ceases to serve as either (a) Chairman of the State Administration Council or (b) Commander-in-Chief of the Myanmar Armed Forces

A cease-fire or peace deal is violated by Iran

Volodymyr Zelenskyy remains President of Ukraine

A new non-majority Muslim (according to CIA Factbook) state becomes a signatory of the Abraham Accords

Azerbaijan becomes a signatory of the Abraham Accords

Cambodia - Thailand Border Clash ceasefire is broken

Ukraine - Russia War ends with a peace treaty

Japan launches a domestically-built intermediate-range ballistic missile (>3 000 km advertised range) test

Major terrorist attack (50+ killed) occurs in Europe

An armed clash between India and Pakistan results in ≥100 military fatalities in a 7-day period

Sudanese Civil War ends

A 5th-generation fighter jet is confirmed to have been destroyed in air/shot down (not including being damaged from an airfield bombing)

Syria becomes a signatory of the Abraham Accords

A drone attacks a head of state (hits or lands within 30 meters)

A cyber-attack officially attributed to a foreign state actor causes ≥US $1 billion in insured losses to a single country

Turkey invokes NATO Article 4 consultations

A state shoot-down of a spacecraft or satellite in orbit is publicly confirmed

A peace treaty ending the Ethiopia–Tigray conflict is signed and remains unbroken for 90 days

Saudi Arabia becomes a signatory of the Abraham Accords

An armed non-state group captures and holds the capital city of a UN-recognised state for ≥72 hours

At least 5,000 uniformed personnel belonging to a UN-Security-Council-authorized security or peace-keeping mission are physically deployed in Haiti

At least three sovereign states outside NATO, EU, ANZUS, or the Quad sign a formal mutual-defence treaty deposited with the UN Treaty Series

Uniformed armed forces in any G20 member state attempt to seize executive power from the constitutional head of government

Major terrorist attack (50+ killed) occurs in United Kingdom

Russian armed forces intentionally attack uniformed personnel or sovereign assets of any NATO member state, or vice-versa

Iraq becomes a signatory of the Abraham Accords

Palestine is granted full United Nations Membership

Iran and Israel both sign (or publicly accept) a written cease-fire or peace agreement, and no lethal exchange of fire between their regular forces occurs for the following 30 consecutive days

China becomes militarily involved in Cambodia - Thailand Border Clash

A vessel larger than 50,000 deadweight tons is damaged or sunk by a naval mine in the Strait of Hormuz

Any EU member invokes Article 42 (7) of the Treaty on European Union in response to an armed attack

A country announces magnetic-fusion net-positive energy

Azerbaijani troops enter Iran

Armenia undergoes a regime change as the result of a military coup

Benjamin Netanyahu remains Prime Minister of Israel

United Nations Security Council adopts (i.e., passes without veto) a formal resolution that explicitly condemns Israel’s use of lethal force against civilians seeking humanitarian aid in Gaza or demands Israel cease such force

The U.S. suffers ≥50 combat fatalities in a single new overseas operation—defined as a country where no U.S. combat deaths occurred in 2024

Libya becomes a signatory of the Abraham Accords

A head of state from a G20 country is assassinated

The IAEA publicly reports that Iran possesses uranium enriched to ≥90 % U-235

The Government of Denmark publicly acknowledges receipt of a formal written offer from Donald J. Trump or his officially empowered representatives to transfer sovereignty over Greenland to the United States

1+ deaths attributed to a rogue AI

Israel conducts military operation in Cyprus

The Arctic Council is formally suspended or dissolved by unanimous vote

Major terrorist attack (50+ killed) occurs in United States

Any current G-20 member state ceases to be listed as a United Nations Member State

Saudi Arabia signs a written, mutual-defence pact with China or Russia

China invades Taiwan

Major terrorist attack (50+ killed) occurs in Canada

Ukraine becomes a formal NATO member

Major News sources consistently identify a conflict as WWIII

Jordan becomes a signatory of the Abraham Accords

At least one nuclear explosive device is detonated with the intent to harm a state’s armed forces or civilian population

Iran becomes a signatory of the Abraham Accords

Lebanon becomes a signatory of the Abraham Accords

Any United States of America state formally states intent to begin process to secede

160 UN Nations simultaneously officially regonize Palestine's statehood.

Any Canadian territory becomes part of US territory

Two states engage in a nuclear exchange (both attempt to deliver a nuclear device within the borders of the other)

Extraterrestrials making contact with earth is widely acknowledged

United Kingdom formally states intention to begin process to reverse "Brexit"

Alberta begins process of becoming a USA state

Coding agents will close the loop

Coding agents will flip the initiative

“Live learning” will be standard

The multi-agent paradigm will win

The specific model will not matter as much as today; the network of agents will be important

Recursively improving coding agents will succeed in the market

NO

NO

It will have sold over 1 million copies by the end of calendar year 2025

It will get a Metascore of 60 or more

It will get a Metascore of 65 or more

It will have sold over 1.5 million copies by the end of calendar year 2025

It will get a Metascore of 70 or more

It will get a Metascore of 75 or more

It will be nominated for at least one award at the Game Awards 2025

It will have sold over 2 million copies by the end of calendar year 2025

It will have sold over 2.5 million copies by the end of calendar year 2025

It will have sold over 3 million copies by the end of calendar year 2025

It will get a Metascore of 80 or more

It will have sold over 3.5 million copies by the end of calendar year 2025

It will win an award at the Game Awards 2025

It will have sold over 4 million copies by the end of calendar year 2025

It will have sold over 4.5 million copies by the end of calendar year 2025

It will get a Metascore of 85 or more

It will have sold over 5 million copies by the end of calendar year 2025

It will have sold over 6 million copies by the end of calendar year 2025

It will have sold over 7 million copies by the end of calendar year 2025

It will be nominated for multiple awards at the Game Awards 2025

It will get a Metascore of 90 or more

It will have sold over 8 million copies by the end of calendar year 2025

It will win multiple awards at the Game Awards 2025

It will have sold over 9 million copies by the end of calendar year 2025

It will have sold over 10 million copies by the end of calendar year 2025

It will have sold over 12 million copies by the end of calendar year 2025

It will have sold over 15 million copies by the end of calendar year 2025

YES

Some other approach

String theory

General work on quantum field theories

Twistor theory

Application of fisher information to physics

Loop quantum gravity

YES

Less than $10M or bankrupt

$10M - $100M

$100M - $1B

YES