Swami Adi

In the land of the blind... The one-eyed man is king.

[Swami Adi @Swami_Adi]

@peterbarnett_ Ideological lunatics.

Akin to a doomsday cult.

"We have a problem"

*Creates the problem

See look guys the problem!

[sdmat @sdmat123]

Anthropic

[Swami Adi @Swami_Adi]

@rizzn @NeelNanda5 Vested interest.

[Swami Adi @Swami_Adi]

@Aella_Girl "We have a problem"

Meanwhile anthropic sets up a wetlab

*Creates the problem

See, Look guys the problem!

[Swami Adi @Swami_Adi]

@peterbarnett_ Ideological lunatics.

Akin to a doomsday cult.

"We have a problem"

*Creates the problem

See look guys the problem!

[Peter Barnett @peterbarnett_]

“but HOw wOuLd tHE AI gET aCcesS tO a wET LAb??”

[Swami Adi @Swami_Adi]

@Chaos2Cured @MTSlive

[Swami Adi @Swami_Adi]

@sephmartinmusic @Polymarket Well atleast Yann isnt a nut.

[Swami Adi @Swami_Adi]

@remarks The ELIZA effect on steroids.

[Remarks @remarks]

JUST IN: Some Anthropic engineers are reportedly "worshipping" Claude as a God.

[Hailey Lennon @HaileyLennonBTC]

Anytime the government talks about crypto or AI, I picture this image

[Swami Adi @Swami_Adi]

@AestherML Sigh...

You need to actually read philosophy. Then study AI for as long as Yann to understand.

What you call "reasoning"

Shannon demonstrated by pure prediction.

Written English is about 75% - 80% redundant.

(Prediction and Entropy of Printed English - 1951)

@ylecun

[Swami Adi @Swami_Adi]

@AstuteRealm @MnehmosTMRI @moiiikaaa If USA is evil to you. China literally installs malware on your cell phone. How's that for suppressing or monitoring dissidence.

Neither are the good guy. But one is clearly leagues ahead of the other. Hint its the one we still have elections. Not a permanent president.

[Swami Adi @Swami_Adi]

@MnehmosTMRI @moiiikaaa Not justifying either.

But the scale and intent are vastly different.

In China they scan your cell phone coming off the plane. Literally install malware on your phone.

Here is a parallel. That is like if they got TSA installing Pegasus on foreigners phones.

[Swami Adi @Swami_Adi]

The EA and Doomerism crowd.

They are playing a game of "What if" not anything backed by facts

They take logically possible, yet improbable scenarios and present them as an actual threat

They are more akin to Rational Egoism frame. In everything they do is some "Great Project"

[Swami Adi @Swami_Adi]

@MnehmosTMRI @moiiikaaa You compared a literal authoritarian surveillance state(China) to a private company putting up cameras in America. OK

[Swami Adi @Swami_Adi]

@DudeWhoInvests @brewmarkets E/acc

[Swami Adi @Swami_Adi]

@boneGPT @micsolana

[Swami Adi @Swami_Adi]

@Sentdex EA's safety philosophy is keep patching holes. Avoiding the actual problem.

Safety today, is waiting for the model to output. Then judging that output in an rl environments.

Complete insanity.

Then the circlejerk with METR lately. More evals and jacking ourselves off safety.

[Satyam @sattyyouneed]

are you actually worried about AI becoming too powerful?

yes or no?

[Swami Adi @Swami_Adi]

@sattyyouneed Not the least bit worried.

Because fundamentally everything it has demonstrated as an issue, is an engineering issue.