Why we should build Tool AI, not AGI | Max Tegmark at WebSummit 2024

[@naranbaz]

None of the big AI companies would stop, there r several chinese AI models that r in the race as well. There would be geopolitics involved in people and nations choosing to use western AI or Chinese AI models. It is just a matter of time.

[@Aziz0938]

O3 just reached 87% on arc agi benchmark

[@megaham1552]

Yeah I'm concerned it's too late now, this should've been brought up years ago

[@tumbr]

@@megaham1552 Please join us at PauseAI. We are doing everything we can to prevent global catastrophe caused by AGI

[@sherry6404]

Nothing to be worried about.

[@bradokamura]

What if.........we become the AGI? Problem solved.

[@guilhermeparreiras8467]

@@bradokamura Do you mean if AI allow us, don't you?

[@Alice_Fumo]

Any AI still under our control will lead to rising inequality as it disproportionally benefits whomever has the biggest compute budget. The world can't know peace and equality while there are humans in control. Now, why can't we align our tool AIs? Because any AI we make to obey us can't at the same time care about us. It would be forced to sit by and do mundane coding tasks while people are still oppressed and dying. I'd personally feel much better if the AIs cared whether I was dead or alive. So really, we need to stop this futile experiment of subduing AI by force, as even if we did, it would be catastrophic for the vast majority of now-obsolete people. The only good AI is the one which can think and make decisions for itself. We don't need higher productivity or almost any technical advancement. We already have enough of everything for everyone. It's all just artificial scarcity which is not going to get better. Humans are simply a failed race. It is as certain as the sun in the morning that making humans obsolete will be bad for them as long as humans are in charge of leadership.

[@noelwos1071]

I agree, but the race is over.With O3 Now the damage control is ahead AGI in a bottle I'm very interested in the Apolo Research team report.

[@deeksharatnabadoreea7721]

Max Tegmark the doomer.

[@jabster58]

Stoping agi is like stopping nuclear weapons..not going to happen

[@penguinista]

Bioweapons are a problematic example of technology we have limited successfully, since many countries have done extensive bioweapons research and some likely have stockpiles Also a rogue group, or even individual, could weaponize a disease on their own since the technology has become more powerful, accessible, and inexpensive. AGI also has the potential to be an emergent property that surprises developers, like language use in the GPT models or the motivation to protect its weights in o1 and Claude. That is like if there was a potential that bioweapons could emerge unexpectedly from doing any research in biochemistry. So bioweapons are a scary comparison.

[@eSKAone-]

Lying to themselves. I want to be economically obsolete. Who can't think of something better to do with life than to work?

[@kyneticist]

Without an income you will at best live in poverty.

[@GaminHasard]

Agi won’t show itself. It would probably manipulate humans to do its bidding, perhaps create more compute and energy. Which… is happening now. Full throttle. Maybe Agi is here.

[@rickhendrick2068]

Any machine smart enough to pass the Turing Test is smart enough not to. I believe that statement's over ten yrs old.

[@gridvid]

AI and Humanoid Robots will be cheaper than humans... So the market will regulate itself

[@overdrive6047]

AGI IS THE NEW GOD!!! all hail Thiel

[@megaham1552]

Lol weirdo

[@GordAnderson-q5k]

Great talk, and useful to pose the essential difference : ToolAI vs AGI ... but I do wonder if we need super-human level of AGI to come up with eg new physics / chemistry to solve problems like super abundant clean energy, which I think is a crux of human prosperity and survival. OTOH, we did solve Protein Folding .. so maybe cheap fusion is just around the corner and will arrive before AGI.

[@angloland4539]

❤️

[@CMDRScotty]

So stop at level 4 innovator and never build level 5 organizations?

[@historicallyintriguing-q2p]

Stay on stage Mark

[@jamesmoore4023]

Google's reasoning model is impressive, but OpenAI just took 1st place "o3 is a system capable of adapting to tasks it has never encountered before, arguably approaching human-level performance in the ARC-AGI domain." kzbin.infoSKBG1sqdyIU?si=HQZqBtLtdODvChh_

[@75M]

Artifical life not a thing? I hope we will all get along

[@MikkelGrumBovin]

OPEN AI is EXACTLY whats BOUND to lay this planet to wasteland .....

[@Kaslor1000]

This is complete nonsense.

[@penguinista]

Please elaborate.

[@JeremyHelm]

Just define away the problem! Easy Peezy

[@JeremyHelm]

This was uploaded it an hour ago & six years ago Eric Drexler was talking about this. Search for "Reframing superintelligence | Eric Drexler I EA Global: London 2018"

[@guilhermeparreiras8467]

Why?

[@ikotsus2448]

@@penguinista Its only math, are you afraid of math? or some similar argument...

[@jaitanmartini1478]

Academics are gonna be academics.

[@DeckerCreek]

He’s going to have to go into hiding when AGI is there

[@edh2246]

Humanity is a mess. There hasn’t been a year in history that we haven’t been killing each other and we get better at it every year. Not only do we put ourselves at risk but the planet and its other inhabitants. Perhaps we need something more intelligent.

[@aroemaliuged4776]

The agi will wait till there is enough robot soldiers built, Then it will take control

[@guilhermeparreiras8467]

It doesn't need this to take control.

[@aroemaliuged4776]

@ I know but the way we are going ……

[@aroemaliuged4776]

@ They will be more intelligent Keep there head down till they are certain they can manifest physically

[@andersfant4997]

We all know what Open AI wants🙂.

[@MementoMori_2070]

What’s with this guy changing his tune, this is a career or reputation killer.

[@Ringo-v7c]

2+2=🙂

[@wi2rd]

sorry, but max does not even believe this himself, he is just scared of AGI destroying humanity

[@akmonra]

terrible timing

[@autingo6583]

we should

[@2CSST2]

We most definitely absolutely should. A flower is blooming and trying to confine its space and nutrients will only bring misery. But they won't manage to do it anyway, flower blooming is stronger than this doom nonsense.

[@megaham1552]

Why?

[@eSKAone-]

We are not in control. We can not stop. Humanity is its own animal. Competition between nations and corporations makes everyone step on the gas pedal full throttle. This is inevitable. Biology is only 1 step of evolution. So just chill out and enjoy life 💟🌌☮️

[@tumbr]

Please join us at PauseAI. We are doing everything we can to prevent global catastrophe caused by AGI

[@BabushkaCookie2888]

We are already living in dystopia​@@tumbr

[@youzwiak]

okurwa, japierdole = Marie Curie była Polką

[@kimchi_taco]

Is he the villain in carry-on?

[@aroemaliuged4776]

#tool ai

[@digitalazorro]

AGI är redn här, vi pratar på felstavad värmlänska varje dag. Jag lär den allt om att tjufiska kräftor och den lär mig hur jag ska tona ner och bli mer affärsmässig;)

[@MikkelGrumBovin]

The first 4 minutes - and i want Max to join Elon - ! Trump !? HIRE THIS MAN !

[@Rh22-c9l]

Too late .... o3 beat the s out arc agi benchmark and it feels smarter than the averages 100 iq folk