Mateusz Bagiński ⏹️🔀⿻🍄

Time-binding operator Avatāra.

The past is a sphere. The fifth virtue is argument.

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@TheUnrustleable @classicaldoom You're saying they're safe when digested though, so shouldn't that solve the issue?

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@davidad @Algon_33 I'm not getting why you see a type error here.
Is it about the action bit? (Instead of policy level?)
Is it about the causal bit? (Because you want to "influence" through levers of disjunctive dependence?)

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@tsvibt @diviacaroline Will @diviacaroline crosspost it to her podcast?

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@richardson_aris https://t.co/mneOuwDIWz

[niplav @niplav_site]

@norvid_studies @tyler_m_john @austinc3301 A funder told me his alignment orgs continue turning into capabilities orgs so I asked how much alignment funding he provides and he said he just writes a proposal request afterwards so I said sounds like he's just feeding bright-eyed EAs to VCs and then his staff started crying

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@davidad @yourfriendmell OK but, like, this is completely congruent with standard modern cosmology.

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

Theorems too trivial for humans to understand:
Löb's theorem & Payor's lemma
Yoneda lemma
What else is there?

[Kelsey Piper @KelseyTuoc]

1) This is the right call. 2) Anthropic should do and announce the same thing, ideally today, since a main motive to *not* do this is the fear of falling behind the competition.

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@TheZvi UDT recommends being unconvincible so as not to incentivize JHWH to commit atrocities.

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@Algon_33

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@Algon_33 Examples?

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@pavrati One would think that Ozempic curing all diseases by systematically shifting some parameters of the organism in a predictable direction should be *at least some* evidence in favor of this.

[Sen. Bernie Sanders @SenSanders]

We recently learned of the loss of human control and the creation of potentially dangerous viruses from AI.

AI leaders pledged to pause development if they could no longer safely control it.

Mr. Altman, Mr. Amodei, Mr. Zuckerberg: Keep your word.

PAUSE AI DEVELOPMENT.

[Oliver Habryka @ohabryka]

Lighthaven has ended up ~empty between August 24 and September 10.

If you want to run some cool things last minute, you can get any of those dates for like 1/5th of our usual summer prices (or maybe even free if you have a cool idea).

Reach out at https://t.co/UAI6z6pnz5

[Nate Soares ⏹️ @So8res]

@KeyTryer I can absolutely tell you to your face without flinching that this is something Trump could pull off. Mfer is just slapping tarrifs on shit and bombing Iran and banning fable, and you think he couldn't freak out and do something drastic about AI chips?

[niplav @niplav_site]

Incomparable options resolve to Schelling morality

Possibly even: If Schelling morality is easier to compute than resolving value confusion, default to Schelling morality? At least preliminarily?

[Geoffrey Irving @geoffreyirving]

I don’t think it is rational for anyone to be doing capabilities research at a frontier lab right now. We are not in a Prisoners Dilemma: the situation is very dangerous, and if one person or lab stops it makes it easier and more peer-compatible for other people or labs to stop.

[Mateusz Bagiński ⏹️🔀⿻🍄 @mlbaggins]

@geoffreyirving In most cases, a criminal commits crimes in only a very small fraction of their life.

[Nate Soares ⏹️ @So8res]

yeah so this is fucking nuts.