-
products
2025-07-30
Framed a vision of 'personal superintelligence' for everyone — putting the power in people's own hands.
«Meta's vision is to bring personal superintelligence to everyone... putting this power in people's hands»
-
jobs
2025-07-30
Contrasted his approach with rivals: against centrally directing superintelligence to automate all valuable work.
«distinct from others who believe superintelligence should be directed centrally towards automating all valuable work»
-
open source
2025-07-30
↺ mengubah posisi
Position shift: acknowledged superintelligence risks and said Meta will be careful about what it open-sources.
«superintelligence will raise novel safety concerns... careful about what we choose to open source»
-
agi timeline
2025-07-30
↺ mengubah posisi
In his superintelligence letter: AI systems have begun improving themselves; superintelligence now in sight.
«Over the last few months we have begun to see glimpses of our AI systems improving themselves»
-
compute
2025-07-14
Pledged hundreds of billions into compute for superintelligence; clusters Prometheus and Hyperion.
«We're also going to invest hundreds of billions of dollars into compute to build superintelligence.»
-
products
2025-01-24
Called 2025 a defining year for AI: Meta AI to reach 1 billion people and Llama 4 to become the leading model.
«Meta AI will be the leading assistant serving more than 1 billion people, Llama 4 will become the leading model»
-
jobs
2025-01-10
On Joe Rogan podcast: in 2025 AI will act as a mid-level engineer able to write code at companies.
«in 2025... an AI that can effectively be a sort of midlevel engineer that can write code»
-
china us
2024-07-23
Says closed development hurts the US: constraining innovation to closed models lowers America's chance to lead.
«constraining American innovation to closed development increases the chance that we don't lead at all»
-
safety
2024-07-23
Argues open source is safer than closed: systems are more transparent and can be widely scrutinized.
«Open source should be significantly safer since the systems are more transparent and can be widely scrutinized»
-
open source
2024-07-23
Manifesto essay: open-source AI is the path forward and the world's best shot at broad benefit and security.
«Open source AI represents the world's best shot... for everyone»
-
agi timeline
2024-01-18
↺ mengubah posisi
First openly stated Meta is building general intelligence (AGI) and will responsibly open-source it.
«building general intelligence, open-sourcing it responsibly, and making it available and useful to everyone»
.jpg){kind=link}