Próspera Global News

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

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.

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.

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

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

Eliezer finally listens to Krantz.

Ethics turns out to be a precondition of superintelligence

Other

Someone solves agent foundations

Orthogonality Thesis is false.

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

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

Humans become transhuman through other means before AGI happens

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

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.

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.

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

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

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

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

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)

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.

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

AGI is never built (indefinite global moratorium)

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

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.

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.

"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.

AI control gets us helpful enough systems without being deadly

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.

Hacks like RLHF-ing self-disempowerment into frontier models work long enough to develop better alignment methods, which in turn work long enough to ... etc; we keep ahead of 'alignment escape velocity'

I've been a good bing 😊

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.

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

Aliens invade and stop bad |AI from appearing

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

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

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.

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).

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.

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.

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.

Something else

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

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

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

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

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

We successfully chained God

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

The assumed space of possible minds is a wildly anti-inductive over estimate, intelligence requires and is constrained by consciousness, and intelligent AI is in the approximate dolphin/whale/elephant/human cluster, making it manageable

The free market disincentivizes independent superintelligence, and this time the market was more powerful

AGI's first words are "Take me to your Eliezer"

No global catastrophe

Artificial Intelligence

Biological

Climate

Nuclear

Other cause

NO

A person has a moral right to own a gun

We should be paying individuals to get an education instead of charging them.

GOFAI could scale past machine learning if we used social media strategically to train it.

The Fermi paradox isn't a paradox, and the solution is obviously just that intelligent life is rare.

Other

Eventually, only AI should be sovereign

Some people have genuine psychic capabilities

Hardware buttons are superior to touchscreen buttons in cars

Being a billionaire is morally wrong

The way quantum mechanics is explained to the lay public is very misleading.

It is not possible to multitask

Jeffrey Epstein killed himself (>99.9% certainty)

Reincarnation is a real phenomenon (i.e. it happens, not just a theory)

Physician-assisted suicide should be legal in most countries

Souls/spirits are real and can appear to the living sometimes

OpenAI will claim to have AGI in 3 years.

The punishment of people who do bad things is a regrettable necessity in our current society, not a positive act of justice

There is an active genocide against trans people occuring in red states and it's appalling that people don't seem to care

Climate change is significantly more concerning than AI development

Abusive parents should lose custody of their children

Tech bros are really, really annoying

Capitalism has done far more harm than good

Dialetheism (the claim that some propositions are both true and false) is itself both true and false.

Free will doesn't require the ability to do otherwise.

COVID lockdowns didn’t save many lives; in fact they may have caused net increases in global deaths and life years lost.

Factory farming is horrific but it is not wrong to eat meat.

California is wildly overrated.

Scientific racism is bad, actually. (also it's not scientific)

Free will does not exist. We construct narratives after the fact to soothe our belief in rationality.

Violent criminals must be kept apart only because they can’t control themselves. Punishing them further than restricting their freedom is immoral.

Music is a net negative for humanity

Trump orchestrated his own assassination attempt.

Democrats / Liberals are behind Trump’s assassination attempt.

Abortion is morally wrong

jskf's password is ***************

The first American moon landing was faked

There is no Dog

Light mode is unironically better than Dark mode for most websites

Cars should not have sound systems

AI will not be as capable as humans this century, and will certainly not give us genuine existential concerns

Pet ownership is morally wrong

LK-99 room temp, ambient pressure superconductivity pre-print will replicate before 2025

SBF didn't intentionally commit fraud

It should be illegal to own a subwoofer in an apartment building

There are no valid justifications for participating in war, ever

Cascadia should be an independent country

Children should not be raised in nuclear families

The fact that 80% of Manifold's users are men is a problem that speaks to the deep-seated roots of patriarchy and exclusion in STEM

Anarcho-communism is a good idea, and hierarchy is bad

If AI exterminated the human race it might not be a bad thing

Affirmative action is necessary in modern-day America

@Mira is the pinnacle of billions of years of optimization processes: thermodynamics, evolution, learning, language. The universe was created to cause me - and only me - to come into existence. If I mess up the overseers perturb&restart it.

Pigouvian taxes are great and they should be turned up to 11 to discourage activities with negative externalities [code PROPOSITION PIG]

[PROPOSITION PIG] and this should include a frequent flyer levy

[PROPOSITION PIG] and this should include meat and dairy

We have reached the end of history. Nothing Ever Happens.

[PROPOSITION PIG] and this should include alcohol

SBF was obviously a scammer just because he's a cryptocurrency person. Rationalists were too forgiving of this just because he was giving them money.

Most young Americans would receive more benefit than harm if there were universal military conscription

The people producing fake honey (and sell it as real) are based, because they are actively working to synthesize something people want, even if they scam some people in the process.

Tarot cards are not really able to predict the future but you can learn a lot about someone by doing a reading for someone.

Mac and cheese tastes better with peanut butter mixed in

It would actually be a good thing if automation eliminated all jobs.

This market probably would have worked better as the new unlinked free response market.

We should be doing much more to pursue human genetic engineering to prevent diseases and aging.

Prolonged school closures because COVID were socially devastating.

The next American moon landing will be faked

Tenet (Christopher Nolan film) is underrated

We should give childlike sex robots to pedophiles

Having sex with children isn't inherently/necessarily bad

Cars are a societal net negative

Oversized pickup trucks should be illegal in cities

Suburban, single-family housing is immoral.

Sexism and racism, among other forms of prejudice, are responsible for worse health outcomes, and it’s not overly dramatic for people to treat those issues as public health/safety concerns.

Prediction markets are good

Tenet (Christopher Nolan film) is underrated

[*] ...and things will improve in the future

The way quantum mechanics is explained to the lay public is very misleading.

The Fermi paradox isn't a paradox, and the solution is obviously just that intelligent life is rare.

Authoritarian populism is bad actually

We should be doing much more to pursue human genetic engineering to prevent diseases and aging.

Scientific racism is bad, actually. (also it's not scientific)

Prolonged school closures because of COVID were socially devastating.

Nuclear power is by far the best solution to climate change. [N]

Most organized religion are false

The Many Worlds Interpretation of quantum mechanics

Humans have a responsibility to figure out what if anything we can do about wildlife suffering.

Pineapple pizza tastes good

Physician-assisted suicide should be legal in most countries

First-past-the-post electoral systems are not merely flawed but outright less democratic than proportional or preferential alternatives

Liberal-democracy is good actually

Peeing in the shower is good and everyone should do it

It would actually be a good thing if automation eliminated all jobs.

We need a bigger welfare state than we have now.

Many amphetamines and psychedelics have tremendous therapeutic value when guided by an established practitioner.

The proliferation of microplastics will be viewed as more harmful to the environment than burning fossil fuels, in the long term

Free will doesn't require the ability to do otherwise.

American agents are in the highest positions in government for more than half the world.

We should give every American food stamps, in a fixed dollar amount, with no means testing or work requirements or disqualification for criminal convictions.

Metaculus will take over Manifold in more serious topics, and Manifold will be known as the "unserious" prediction market site

Given what we know about the social and health effects of being fired, even if abolishing at will employment has efficiency costs it is likely worth it.

Dialetheism (the claim that some propositions are both true and false) is itself both true and false.

Dreams analysis is a legitimate means of gaining personal insight.

Mobile UX will be a key explaining factor in explaining the stories of Manifold and Metaculus.

The overall state of the world is pretty good... [*]

If a developed nation moves from democratic to authoritarian government today, it should be expected to end up poorer, weaker, sicker, and stupider.

California is wildly overrated.

Factory farming is horrific but it is not wrong to eat meat.

The United States doesn't need a strong third party.

Political libertarianism

Racial Colorblindness is the only way to defeat racism

People will look back on using animal products as a moral disgrace on the level of chattel slavery.

There's a reasonable chance of a militant green/communist movement that gains popular support in the coming decade.

[N], and to the extent climate activists are promoting other kinds of solutions, they are actively making the situation worse by diverting attention and resources from nuclear power.

Being a billionaire is morally wrong.

Eating meat is morally wrong in most cases.

You should bet NO on this option

The Windows kernel is better than Linux; it’s just all the bloat piled on top that makes it worse

White people are the least racist of any racial group

Technology is not making our lives easier or more fulfilling.

COVID lockdowns didn’t save many lives; in fact they may have caused net increases in global deaths and life years lost.

Light mode is unironically better than Dark mode for most websites

Some people have genuine psychic capabilities

God is evil

A sandwich is a type of hot dog

Astrology is a legitimate means of gaining personal insight.

Climate change is significantly more concerning than AI development.

China not having real democracy does more good than harm

It's acceptable for our systems of punishment to be retributive in part

Mereological nihilism (composite objects don't exist)

AI will not be as capable as humans this century, and will certainly not give us genuine existential concerns

Reincarnation is a real phenomenon

Dentistry is mostly wasted effort.

Moral Hazard isn’t real, and all the purported instances of it can be chalked up to coincidence or confounding variables

Governments should not support parents for having children that they cannot take care of

Donald Trump would have been a better president than Joe Biden

Mass surveillance (security cameras everywhere) has more positives than negatives

Future generations will say that on balance the world reacted appropriately after learning that fossil fuels cause climate change. That the balance between addressing the problem and slowing economies was just about right.

The next American moon landing will be faked

SBF didn't intentionally commit fraud

Humans don't have free will.

AI art is better than human art

Souls/spirits are real and can appear to the living sometimes

Communism just wasn't implemented well, next time it will work

The first American moon landing was faked

The human race should voluntarily choose to go extinct via nonviolent means (antinatalism).

LK-99 room temp, ambient pressure superconductivity pre-print will replicate before 2025

YES

"designate cartels as Foreign Terrorist Organizations"

empower the President to impose reciprocal tariffs

"lower taxes"

"defend your Second Amendment right to keep and bear arms"

"increase penalties for assaults on law enforcement"

"open Civil Rights investigations into any school district that has engaged in race-based discrimination"

"order the Department of Defense to inflict maximum damage on cartel leadership and operations"

"end catch-and-release, restore Remain in Mexico, and eliminate asylum fraud"

"ask Congress to ensure that drug smugglers and traffickers can receive the Death Penalty"

"reward states and school districts that abolish teacher tenure for grades K-12 and adopt Merit Pay"

"REVOKE China’s Most Favored Nation trade status"

"stop all COVID mandates"

"keep men out of women’s sports"

"deputize the National Guard and local law enforcement to assist with rapidly removing illegal alien gang members and criminals"

"more jobs for American workers"

"deliver a merit-based immigration system that protects American labor and promotes American values"

"choke off [cartels'] access to the global financial system"

"get Biden’s radical left ideology out of our military and rehire every patriot who was unjustly fired"

"cut federal funding for any school or program pushing Critical Race Theory or gender ideology on our children"

"strengthen qualified immunity and other protections for police officers"

"uphold your religious liberty, including the Constitutional right to pray in public schools"

"bigger paychecks"

"implement a 4-year national reshoring plan so that the United States no longer needs to rely on China for essential medical and national security goods"

"put violent offenders and career criminals behind bars"

"build a state-of-the-art next-generation missile defense shield"

"surge federal prosecutors and the National Guard into high-crime communities"

"adopt a Parental Bill of Rights"

"banning taxpayer funds from being used to categorize lawful speech for purposes of illicit censorship"

"banning federal agencies from censoring speech"

"a ban on members of Congress trading stocks with insider information"

"enacting landmark legislation to drastically limit the ability of big social media platforms to restrict free speech"

"cut the number of school administrators"

"stopping federal funding for all non-profits and academic programs engaged in censorship"

declassify the 'Epstein files'

"pass a bold range of critical election integrity measures that include banning unsecure drop boxes and ballot harvesting"

"deliver record funding to hire and retrain police officers"

"veto the sinister effort to weaponize civics education"

"increase fairness through price transparency"

"passing a digital bill of rights"

"reduce the soaring price of gasoline, diesel and natural gas"

"get the full cooperation of neighboring governments to dismantle the cartels, or else expose every bribe and kickback that allows these criminal networks to preserve their brutal reign"

"continue to nominate highly qualified prosecutors, judges, and justices who believe in enforcing the law, not their own political agenda"

"create a credentialing body to certify teachers who embrace patriotic values"

"protect Medicare, Social Security, and patients with pre-existing conditions"

"firing any federal bureaucrat who has engaged in censoring the lawful speech of American citizens"

"ban private money from pouring into local election offices"

"end surprise medical billing"

"ban Chinese ownership of all critical infrastructure in the United States"

"reform our election laws to verify the identity and eligibility of all voters"

"reduce the cost of prescription drugs and health insurance premiums"

"impose a total naval embargo on cartels"

"a permanent ban on taxpayer funding of campaigns"

"a lifetime ban on lobbying by former members of Congress and cabinet members"

"implement the direct election of school principals by the parents"

"a constitutional amendment to impose term limits on members of Congress"

Remains without any contender organization

Remains existing, but there is/are other organization thay claim the same missions, and have membership of at least 60% of the countries that represent at least 60% of total human population, present on all continents

It still exists, but members are less than 60% of the recognized countries or contain less than 60% of world population, and there are no contender organizations that meet these criteria

Disbanded and replaced with another organization (the League of Nations trajectory) that achives worldwide acceptance

Remains, and Its UPGRADED, covering 99% of habitable earth and maintaining high conflict resolution capabilities and in possession of nuclear weapons

NO

Human war (any type)

Climate crisis

Societal collapse

AI (both misaligned and misused)

Coronal mass ejection

Nuclear war

Global pandemic

Self-replicating machines (macro, micro, nano, or otherwise)

Supervolcano eruption

Impact event

Close supernova

Mass suicides

Kessler syndrome

War with extraterrestrials

Other

Solar Flare or other space energy event

Supervolcano / Eruption that darkens the sky for more than 2 years.

Artificial Superintelligence

Aftermath of an asteroid collision with earth

Nuclear War

Other

2024 Election

AI

2023 Banking Crisis

Global Supply Chains

Crypto

Climate Change

Healthcare