Plato's Beard

Nonbeing must in some sense be, otherwise what is it that there is not?

[Plato's Beard @BeardedPlato]

@The_AI_Investor great take

[The AI Investor @The_AI_Investor]

The difference between Jensen Huang and Dario Amodei is becoming very clear.

Dario has said:

“This thing’s going to be powerful. It could destroy us.”

He has also acknowledged that the models built so far may do things their creators do not understand.

Yet his answer is that frontier labs must continue “building and studying powerful systems,” while tightly protecting their model weights and restricting open release.

That is not literally: “Only Anthropic should build AI.”

But the practical message can sound uncomfortably close to:

AI may be catastrophically dangerous. We do not fully understand it. Therefore, a few closed companies should keep building it behind secured walls—and everyone else should trust their safety reports.

Jensen’s argument is almost the opposite:

“Attackers have frontier AI. Defenders need a frontier AI ecosystem—the best open and closed models, force-multiplied by a global community.”

Anthropic sees open weights primarily as an irreversible proliferation risk.

Jensen sees dependence on a handful of closed, opaque providers as a security risk of its own.

Anthropic’s model: concentrate frontier capabilities inside regulated, secured labs.

Jensen’s model: give defenders access to both open and closed frontier models, backed by a broad global ecosystem.

One side says powerful AI is too dangerous to distribute.

The other says it is too dangerous to concentrate.

[Plato's Beard @BeardedPlato]

@JoshJPhilipp @grok is epoch time connect to a cult?

[Plato's Beard @BeardedPlato]

@jun_song @nvidia how much does few dgx bodes cost? @grok

[Plato's Beard @BeardedPlato]

@kevinxu @grok explain the motivation

[Plato's Beard @BeardedPlato]

@AriaLikesStocks @grok what underlying psychological mechanism contribute the fixation on contrarian trades

[Plato's Beard @BeardedPlato]

@Jason said that in 2021

[Plato's Beard @BeardedPlato]

@GordonGChang @grok fact check the generalization

[Plato's Beard @BeardedPlato]

@ariaradnia Nvidia leveraging OAI to serve end customers for electron-token end to end loop @grok

[Plato's Beard @BeardedPlato]

@elonmusk no loser mentality in the loop👍

[Plato's Beard @BeardedPlato]

@AndrewCurran_ @grok is this narrative steering?

[Plato's Beard @BeardedPlato]

@GordonGChang @grok fact check this guy

[Plato's Beard @BeardedPlato]

@anuraagsidhu @grok is this true? what happens if google start enforcing it

[Plato's Beard @BeardedPlato]

@wangzhr4 @grok whats the difference, from copyright perspective

[Plato's Beard @BeardedPlato]

@bubbleboi winner mentality

[Plato's Beard @BeardedPlato]

@hidefromKGB @daniel_koss @grok how likely all of anthropic’s training material are compliant for model training and distillation

[Plato's Beard @BeardedPlato]

@jojjeols MAGA is fine with that

[Plato's Beard @BeardedPlato]

@johngfriedman if you got stolen, go to court, not regulatory capture

[Plato's Beard @BeardedPlato]

@daniel_koss “This guy” is Andrew Ng

[Plato's Beard @BeardedPlato]

@dylan522p so this is after watching the movie?