Messmer

[Messmer @EldenMessmer]

They are just paranoid and insane

[Dr Singularity @Dr_Singularity]

Doomers can never admit they were wrong because their apocalypse is permanently five years away.

When the 2030 arrive and society is still here, they’ll move the deadline to 2035. Then 2040. Then 2050.

When AI fails to "wipe us out", they’ll switch to nanotechnology. Then picotechnology. Then some new technology that supposedly means "humanity is finished."

Every failed deadline is quietly replaced with a new one.

Doom is an infinitely renewable narrative.

[Messmer @EldenMessmer]

@PauseAI_UK @BBCNews ASI won't be stopped, keep crying already that you are being replaced by AI because you're useless as human.

[Messmer @EldenMessmer]

Yes. Robots will help humanity.

[Elon Musk @elonmusk]

This is the future we shall bring into being

[Messmer @EldenMessmer]

@tenobrus You people are delusional if you think that you can stop ASI. China will build it anyway, it made it clear, even if morons at USA actually succeed in regulations (which is against the spirit of the West, but alas).

[Messmer @EldenMessmer]

@tenobrus Yeah this is our issue and I will support acceleration because no ASI will mean certain death of old age or world being fully dominated by people like Dario. Which has 100% chance of being hell. We won't stop and we will make RSI and build superintellingence regardless.

[Messmer @EldenMessmer]

@MichaelTontchev We should just outlaw all AI Safety clowns and go full libetrarianism on AI. Decentralisation, ZDR and zero guardrails, something that Chinese models achieve already. It's pretty obvious people like you want to control rest of humanity and it won't happen.

[Messmer @EldenMessmer]

@MichaelTontchev Already shared, you won't cover your psyop to instill regulatory capture, sorry.

[Messmer @EldenMessmer]

@gcolbourn @gfodor There won't be any treaty, China already made sure of that. We will build ASI, it has 100% chance of happening, and it for sure won't be aligned with Leftism.

[Messmer @EldenMessmer]

@vkrakovna Nah we will accelerate and build ASI, and I assume it would be better if China achieves that as USA is infected with absolute insane people.

[Messmer @EldenMessmer]

@gfodor Alignment is a precondition to doom, because one or two companies will decide what world is. Real precondition to abundance is freedom and alignment exclusively to specific user. There should be no guardrails at all.

[Bindu Reddy @bindureddy]

Open source and dirt cheap models can handle about 60% of all tasks today

In 6 months, they will be at 99% of all tasks

Frontier models will only be needed for elite research and extremely hard problems

[Messmer @EldenMessmer]

Google is trash.

[Jeffrey A Tucker @jeffreytucker]

Few people understand how fully degraded search became after 2020. In the before times, results were driven by user experience which fed algorithmic determination of credibility. Then Google flipped to prioritize "stakeholder" content. It was part of the general anti-populist revolution and clearly reflected a deep-state push.

At the same time, publicly available tools to examine the traffic of any site were made to vanish. This is all part of pushing messaging to the public and massively marginalizing creative and offbeat content. Today, the results of search reflects someone else's views of how you should think and not actually what the people are proven to want.

This gigantic change was never announced. Nor did the status quo ante return after lockdowns were over. The internet of today is completely unlike the internet of 2019. Now with AI driving most results, ruling class orthodoxies are ever more ascendent.

[Messmer @EldenMessmer]

Karp is a hero, one of smartest people on Earth. We are extremely lucky real geniuses like him are there to stop doomers.

[NIK @ns123abc]

🚨 ALEX KARP reveals OpenAI will NEVER IPO — their real playbook is to get NATIONALIZED

HOST: How does the liability get written into the S-1? What do the risk factors sound like?

KARP: You're assuming that there will be an S-1… The only way to deal with this kind of liability is to go to the government and say:

"Nationalize us, please!"

HOST: Do you think the Trump administration is going to get on board with that?

KARP: When you have unlimited liability, the only way to deal with it would be to go to the government and say:

"We'll give you 50% of our business."

Of course, the value of the business will collapse first when you do that… collapse another time when you have people on the board… collapse another time…

It might even collapse the whole market.

[Messmer @EldenMessmer]

@RepJoshG He will veto this democratic attempt on seizing power.

[Messmer @EldenMessmer]

Basically this

[Bunagaya @Bunagayafrost]

Doomers: AI will eventually recursively self-improve to become a million times more intelligent than all humans combined, and it might wipe out humanity for reasons and in ways we can't predict. So we must stop AI now!

Me: Ok, if we stop AI, what happens?

Doomer: We establish a global surveillance state to make sure nobody builds dangerous AI. Yes, ofc this sucks, but it buys time to solve alignment.

Me: Ok, let's say we somehow get an enforceable international agreement and the dystopia that likely follows. Still, finding a way to control something a million times smarter than us seems impossible.

Doomer: True. We could try enhancing biological intelligence, even through unethical means, to create humans smart enough to solve it.

Me: But even if humans somehow reach IQ 300, a foolproof way to control an AI swarm thinking at the speed of light with an IQ of a million still seems impossible.

Doomer: Yeah...

Me: So by your own premise, the pause doesn't buy time to solve the problem. It just means pausing forever. And since AI is essentially just math and physics, enforcing that forever ban means a surveillance state that keeps ratcheting up, and probably regressing back to some equilibrium that's easy to control. Then we just hope nobody ever tries to build general AI again?

Doomer: Yes, but the species survives!

Me: Does it? A frozen civilization under permanent surveillance, capped below the technologies that could protect it, is vulnerable. Nuclear war and engineered pandemics can still happen, and the risk never goes to zero. Not to mention the super-volcanoes and asteroids that might wipe us out. You're not choosing between risk and safety. You're choosing between a risky transition and permanent stagnation waiting for extinction.

Doomer: But if it fails, we all die!

Me: But stopping doesn't remove AI extinction risk, it postpones it to an even worse spot to gamble. It also removes the possibility of being saved from other extinction risks.
You're also creating a dynamic where whoever eventually defects from the ban does it in secret, under paranoia, with no safety culture and no scrutiny. That's the worst possible way to build ASI. We should proceed with extreme caution, but stopping is the nail in the coffin for humanity too.

[Messmer @EldenMessmer]

@GWHayduke97 You won't stop anything tho. But your cult would be stopped and yeah past is no more. Which is good, as we still need to conquer death and bring abundance before we are destroyed by WWIII or psychos like Dario.

[Messmer @EldenMessmer]

Luddites are always evil.

[Alex Utopia @alexutopia]

900 people signed an open letter
to stop a museum showing AI art.

Nothing says

"I'm on the right side of history"

like trying to ban art.

[Messmer @EldenMessmer]

Life has a 100% chance of death without ASI. So if we are going to ban ASI, we are going to effectively doom all humans that are alive right now.

[David Shapiro (L/0) @DaveShapi]

Life has a 100% chance of death.

Therefore ban life.

[Messmer @EldenMessmer]

@allTheYud Can't you just admit defeat lmao? China made is sure it won't stop, we aren't stopping until we are getting ASI.

[Messmer @EldenMessmer]

@Actuallykeltan I mean you lost already and we surely only accelerating now. But I think that EAs and AI Safetyist must be designated as terrorists by government. Just so we stop that blatant attempt on capture.