N8 Programs

This account is verified. Learn more

@N8Programs

Undergrad Studying Applied Mathematics and Statistics at @JohnsHopkins.

Studying In-Context Learning at The Intelligence Amplification Lab.

Proxima Centauri B github.com/N8python Joined September 2022

276 Following 11K Followers 7.4K Posts

Latest posts 5

N8 Programs @N8Programs 2h

My second LessWrong piece is out - a curated set (w/ commentary) of @allTheYud and @So8res posts. LLMs think of Yud/Soares as dismissive of welfare concerns while they actually take the welfare of current models seriously, alongside p(doom). Link below 👇

N8 Programs @N8Programs 2h

Shoutout to this good-faith response to what I was asking!

white guys with guns @CUPIDSCARTRIDGE 3h

(qrt since replies are locked but I want this to be public info) I'm a CS student and also think the og post is dumb, but theres valid reasons to hate AI. 1) Companies like OpenAI have military contracts and thus are partially responsible for a lot of warfare in this age.+
View quoted post on X

N8 Programs @N8Programs 1d

The aifuturesmodel.com estimates an automated coder at ~183 ECI

Andrew Curran Verified account @AndrewCurran_ 1d

'Fable 5.2’s score is above what the pre-Astra trend for reasoning models would have predicted.'

That's because we're on a new trajectory.

View quoted post on X

N8 Programs @N8Programs 2d

Checks out. Fable 5.2 knows nearly everything there is to know about me online. Borderline omniscient.

leo 🐾 Verified account @synthwavedd 2d

Ummm, it appears Fable 5.2 has saturated SimpleQA Verified, with a massive leap over every model before it

View quoted post on X

N8 Programs @N8Programs 2d

This is encouraging if internal Claudes get similar training - but we have no guarantee that they will.

Sell everything Verified account @yoshispeninsula 2d

so all "pacing the frontier" ultimately boils down to is not letting the public Claude do work that competes with Anthropic ?

View quoted post on X