gnarmis

[gnarmis @gnarmis]

@monitoringbias for eg,

[gnarmis @gnarmis]

@monitoringbias Trash take. The actual document is instead acknowledging most math questions have correct answers, and suggesting to focus on building interpretation skills on things like complex word problems.

[gnarmis @gnarmis]

@JosephPolitano public entities are using their authority to backstop deposits and are not bailing out shareholders of banks that were essentially gambling. Banking regulation should assume even smaller/regional banks can pose systemic risk, but actions here have been narrow

[gnarmis @gnarmis]

@JosephPolitano preventing insured and un-insured depositors from loss of their deposits, by effectively an insurance mechanism that banks pay for (not taxpayers), and not covering shareholders of this bank or management = “bailout”? This is just a hot take.

[gnarmis @gnarmis]

@xlr8harder “AI will kill everyone” isn’t even the belief, though? It’s more “unaligned AI is more likely to be harmful than aligned AI” — and aren’t these LLM chatbots (maybe shittily so far) aligned AI systems too? What’s your belief, alignment isn’t necessary to prevent harms?

[gnarmis @gnarmis]

@derek_j_morris Congrats!!

[gnarmis @gnarmis]

@CB2 oh fun, so great to see unsubscribing *specifically* broken!

[gnarmis @gnarmis]

@SeanGransee already texted but it's Ondine :)