carter

princeton . code & networks & DERs & borings

[carter @thecarterjc]

@zeeshanq I have very positive experiences using LLMs in Go codebases. I think the language can be frustrating though. The type system occasionally, yes, but working with the Go GC and memory allocator can be painful for certain perf sensitive apps

[Alina Stefanescu @aliner]

It still gives me goosebumps

[carter @thecarterjc]

@paulnadal_ I recently graduated from Princeton with a minor in English. Sad to miss this - looks fantastic!

[carter @thecarterjc]

@bernardtjoy I read AA in college. Very difficult, very rewarding. One of my favorite novels, and certainly among my favorite final pages.

[Mark W. @DurhamWASP]

“The past is never dead. It’s not even past. All of us labor in webs spun long before we were born, webs of heredity and environment, of desire and consequence, of history and eternity.”

William Faulkner, born 25th September 1897

[carter @thecarterjc]

@litdotsalon I’ve never seen that Dickens cover. That’s Bowen to me.

[carter @thecarterjc]

@asaiiie @modal @amitpr @jonobelotti_IO Awesome work. Learned something new

[carter @thecarterjc]

@cognition Wow, Alex Lombardi. Cool post

[Cian McCarthy @arealmofwonder]

For the last Friday in August...

• Philip Larkin •

[Cloudflare @Cloudflare]

We found a working remote Spectre attack against Cloudflare Workers. We already fixed it. No exploitation in the wild. Blog post and paper out now:
https://t.co/okdulUT1yB

[carter @thecarterjc]

We've been focusing on paintings and certain works on paper. If you have suggestions for other collections, do let me know.

The study feature has slowly been taking shape, so I'm excited to see where things go.

[carter @thecarterjc]

Over 100k artworks in Vasari! Check it out https://t.co/kL2kvfWTqF

[Jeff Dean @JeffDean]

We created a pitch deck to tell a handful of VC firms about us and what we were up to (a fun experience!). Here’s a few slides about our background and some of the things we’ve worked on from the pitch deck (it was fun putting together the list of people in our teams who have gone on to found a whole range of exciting companies). We are delighted to have selected @radicalvcfund
and @khoslaventures to lead our initial funding round, along with participation from @lightspeedvp, @kleinerperkins, Doerr Capital (@johndoerr), and Alphabet (@Google). We’ll be working with them to close our seed round over the next few weeks.

[carter @thecarterjc]

@KobeissiLetter Hiring from OAI’s Jalapeño team

[Brett Adcock @adcock_brett]

Figure robots now have visitor passes!

They have full access to every door on campus. Soon there will be hundreds of them walking around campus, just coming and going

[carter @thecarterjc]

SemiAnalysis' prediction that *annual* hyperscaler capex would spike to 2T sounds more reasonable every day!

[The Kobeissi Letter @KobeissiLetter]

BREAKING: Nvidia is in talks with OpenAI to guarantee $250 billion in financing for a data center in Ohio.

Details include:

1. The guarantees from Nvidia would help OpenAI lease a 10GW project that SoftBank is developing in Ohio

2. In total, the project could cost more than $500 billion, the largest data center project ever announced

3. The power for the project is controlled by the US government and funded separately by Japan under a recent trade deal

4. Commerce Secretary Lutnick is reportedly involved in deciding who will get the power

The data center buildout just hit a whole new level.

[carter @thecarterjc]

+ many thanks to @modal and @Cloudflare for making this possible

[carter @thecarterjc]

Vasari now aggregates works from the CMA, PMA, AIC, + more! Who should we add next?

[carter @thecarterjc]

An interesting, and perhaps promising, take on training a CUA. Is it possible to learn actions directly from pretrain without using SFT after? Seems very scalable with no IDM in the loop.

Visualizing the representation space as a screenshot is neat.

[Induction Labs @induction_labs]

We’re introducing imagination models: a new foundation model architecture that unlocks learning from internet-scale video.

Our first imagination model, Photon-1, learned to use a computer by watching 18 years of screen recording video without action labels.

[carter @thecarterjc]

@natlungfy @midjourney @DavidSHolz @banu__guler I wonder if they’ll port their code from Haskell to something else