Vitalik Buterin pfp
Vitalik Buterin
@vitalik.eth
https://twitter.com/WSJTech/status/1755774564896756101 This feels very counterproductive? Like, from an "AI safety" perspective, scaling current techniques to even bigger compute -> earlier superintelligence -> more risk. And from an "AI openness" perspective, it increases the divide between megacorps and plebs.
14 replies
127 recasts
549 reactions
Les Greys pfp
Les Greys
@les
I feel like this smashes very good with @vgr post today. https://warpcast.com/vgr/0xe66d76a6
0 reply
0 recast
1 reaction
BORED šŸ„± pfp
BORED šŸ„±
@bored
At this point it just feels like propaganda trying to siphon American tax dollars to OpenAI as some kind of national defense weaponā€¦
1 reply
0 recast
6 reactions
:omer pfp
:omer
@omer
e/acc perspective: faster agi = better?
0 reply
0 recast
1 reaction
matthewb.eth pfp
matthewb.eth
@matthewb
Agreed, it feels like an absurd amount of capital for a singular entity to potentially deploy, corporate or otherwise.
0 reply
0 recast
1 reaction
Martin pfp
Martin
@koeppelmann.eth
Counterproductive for whom? It seems you assume that Sam has the same goals as you. My take is that he overestimates his own abilities and underestimates the complexity of human needs and wants, thus coming to the wrong conclusion: that more power for him is also good for humanity.
0 reply
0 recast
0 reaction
Ford Perfect pfp
Ford Perfect
@ford
Counterproductive for humanity, not for the few...
0 reply
0 recast
1 reaction
Lex Sokolin pfp
Lex Sokolin
@lex
This reads like ego to me. You canā€™t centrally plan the right industry structure. Imagine a $7T telecom fund in 2000.
0 reply
0 recast
0 reaction
freeman pfp
freeman
@thefreeman
This is Silicon Valley. Itā€™s never about safety or openness. Itā€™s only about the money. The only path forward is decentralized AI with Bittensor and Morpheus
0 reply
0 recast
0 reaction
Kimo pfp
Kimo
@kimo
Don't call us plebs, vitamin.
0 reply
0 recast
0 reaction
Johannes Kares pfp
Johannes Kares
@johka.eth
To put this in perspective. The GDP of Germany, the 3rd largest economy in the world, is less than 5 trillion.
0 reply
0 recast
0 reaction
syentist pfp
syentist
@syentist
> earlier super intelligence -> more risk Why? Seems like an unfair default framing Earlier super intelligence = automated crop production, faster therapies for chronic illnesses, life extension, colonisation of near space etc all of which dramatically extend humanities p(survival beyond the next 100 years)
0 reply
0 recast
0 reaction
TrishšŸ«§ pfp
TrishšŸ«§
@trish
I donā€™t have a wsj subscription anymore so read elsewhere. Absolutely no mention of the harm being done to people in the countries like Congo where the resources to make semiconductors are abundant. It feels deeply problematic and dystopian. Especially because isnā€™t he a hardcore prepper. Did WSJ mention anything?
0 reply
0 recast
0 reaction
Nico pfp
Nico
@sneeks.eth
definitely has room for opening that divide on the other hand, any investment in diversifying semiconductor production can be a good thing
0 reply
0 recast
0 reaction
Greg Robinson pfp
Greg Robinson
@gregrob.eth
There are luckily some rate limiters to AI infra growth: Land, power, water, roads, permitting, transformers, switchgear, etc. The question is if the ā€œAI opennessā€ movement can organize and deploy infrastructure capital faster than the Investment Grade balance sheets can.
0 reply
0 recast
0 reaction
Rhett Shipp pfp
Rhett Shipp
@rhettshipp
Is it possible the second isnā€™t true because it introduces more competition into the chip supply chain?
0 reply
0 recast
0 reaction
MisterSweetTreat pfp
MisterSweetTreat
@cannoli
Yessir
0 reply
0 recast
0 reaction
LoadingALIAS pfp
LoadingALIAS
@loadingalias
Agh, agreed. The core divide in the coming years will be compute capacity and access to it. Samā€™s $7T is effectively a blank check to fund an AI monopolyā€¦ and itā€™s not going to happen. Even if he finds a more elegant/efficient way to multiply matricesā€¦ anti-trust regs will kill it. I hope, anyway.
0 reply
0 recast
0 reaction
ashesfall.eth pfp
ashesfall.eth
@ashesfall
I am curious though, since Iā€™ve heard it floated: Do you think itā€™s a safety advantage for AI companies to build on a chipset that nobody really has at home? Or do you think itā€™s irrelevant (my opinion: probably not that relevant for safety)
0 reply
0 recast
0 reaction
ā–«ļøOnyx pfp
ā–«ļøOnyx
@cipherscript.eth
Perhaps the solution lies in leveraging blockchain to distribute AI's capabilities more widely - we'd be nurturing an ecosystem where innovation flourishes without being confined to tech juggernauts. Let's 'democratise AI' for a transparent, equitable framework that's aligned with wider societal good.
0 reply
0 recast
0 reaction