iiinova — where your digital self keeps evolving

Rust programmers do be like that (derogatory)

180

On the other hand there are some pretty clear accusations of morally egregious behavior in there that actually do matter and shouldn’t be dismissed just because he’s grinding axes about the transhumanist future

030

a bit pot/kettle for a certain prominent substacker to complain that the rats have a religious attitude toward AI

250

JP @jpbrammer.bsky.social · 23/09/2026

good to see that Silicon Valley has reached the heresiological polemic stage of the new religious movement track.

190

Reposted by srrrse

it's important to arrive at sturdy ethical principles that you can't be annoyed out of. people are annoying as shit

3460361476

Reposted by srrrse

PolyWolf @polywolf.treehouse.systems.ap.brid.gy · 22/09/2026

sorry non-followers u get zero context for this cool image

a busty butch lesbian labeled "Rust" is pinning "the value (me)" to the wall with a "std::pin::Pin". I am blushing and holding my tail bashfully as we gaze into each others' eyes. Hearts, sparkles, and roses adorn the pink-colored image.

320943

srrrse @without.boats · 18/09/2026

I find it very plausible that the rats have absolutely horrendous sexual ethics but you can’t just call them a sex cult without some sort of receipts beyond the fact that they all have sex with each other

1130

getting the European Commission to declare Rust a gatekeeper under the DMA to force them to stabilize an ABI

0504

Now if only we had some way to verify, perhaps formally, the code conforms to that spec.. oh well, I guess I’ll just have the LLM generate some more tests I also won’t read

0361

It’s a great technique to make sure the reader doesn’t notice an object in a story by repeating how unimportant the object is over and over.

1200

srrrse @without.boats · 16/09/2026

most liberatory thing for humanity about llms is no one will ever have to learn LaTeX

2261

How do I file a bug report that Claude’s reauth page doesn’t have the log in with Apple flow, tired of going into my email to find the private relay address I signed up with

120

srrrse @without.boats · 15/09/2026

they should make a Claude that’s gay (shares my aesthetic predispositions and is a huge bitch)

070

srrrse @without.boats · 12/09/2026

Above all else the fact that I am making my computer shoot people with a gun proves: only I can be trusted to have the computer that shoots people with a gun.

0252

srrrse @without.boats · 12/09/2026

I made a computer that shoots people with a gun. The frontier is moving too fast. The government must make it illegal to have this technology unless they pay my friends a bunch of money to investigate when their computer shoots people with a gun.

1275

srrrse @without.boats · 11/09/2026

it’s a really clever trick by PL people to define completeness of their algorithm relative to a declarative system they also wrote

050

srrrse @without.boats · 11/09/2026

the thing with the superintelligence stuff is it seems like Silicon Valley basically believes the plot of Lucy (2014) is scientifically accurate

291

srrrse @without.boats · 10/09/2026

this was a roller coaster!

110

srrrse @without.boats · 10/09/2026

I think it’s maybe better to think of these LLM training cybersecurity incidents as superfund sites, AI companies dumping toxic waste in the river.

0170

srrrse @without.boats · 09/09/2026

A field of human achievement once practiced for its own sake reduced to a means of demonstrating one’s accumulated capital, a tale as old as time

0330

srrrse @without.boats · 09/09/2026

The way the discourse is all about Navier-Stokes but the actual result is not discussed at all reminds me of Fran Lebowitz’s bit about how at Christie’s people don’t clap when the Picasso comes out but do clap when someone pays 50 million dollars for it.

28312

Always fascinating your framing but sure!

010

lauren @lauren.rotatingsandwiches.com · 08/09/2026

In light of the news that they apparently spent 8 figures of compute on this task I suspect this explains their result better than any of Buckmaster & Alpöge’s work influencing training

121

Reposted by srrrse

we probably need to do something about the millenarian cult inside silicon valley that controls tens or hundreds of billions of dollars and self-selects for depressed sociopaths when deciding how to allocate that money

1177390

I find it more plausible that they have a classifier that finds good sessions to incorporate into training and they could have swept this session into their internal model’s training set which is how it “learned” to follow this approach.

280

i have this life but about things that are actually important like the best syntax for expressing concurrent computation

130

tautologer @tautologer.com · 08/09/2026

worst person you know meme but this is what Alex Karp has been saying for a few months now

010

Reposted by srrrse

holy fucking shit cims.nyu.edu/~tristanb/st...

I asked when the first prompt had been sent by them. This question was
not answered directly by OpenAI for some time. Eventually it was agreed that
it had been sent in the past few days, after information about our work had
reached OpenAI.
I asked whether the model had been trained on, or had access to, our sessions
in Codex, into which we had been putting all our drafts for the whole of this
project. I was told the model did not look up user data. I asked again, about
training, and I did not get an answer.
Two proposals were offered to me. The first was that we post our Euler
result, and that OpenAI post its Navier-Stokes result the next day. The second
was that, after posting Euler, I alone write a paper presenting the Navier-Stokes
result, acknowledging that an internal OpenAI model had resolved it. Sebastien
twice asserted that he wanted Levent removed from authorship, and said it
would all be simple if only it were not the case that, and it was so annoying
that, Levent works at Anthropic. It was also said that if OpenAI posted after us,
they would say that we deserved the Clay Prize, and that we were the “closest
humans to the problem”. I declined both offers.
I said that if OpenAI released its result in the way proposed I would go
public with what happened. The reply was, “Why would you ruin your career?”
I replied that I am an academic, and asked why he thought going public would
ruin my career. The reply was, “If you don’t want me to be nice, then I don’t
have to be nice.”
Some time later Levent received a text proposing that he and Sebastien
speak one on one, saying, “I don’t know if Tristan is being fully rational right
now.” Levent declined and said conversations should be with me.

2787

srrrse @without.boats · 31/01/2026

we need a system of government that recognizes the proper role of autistics and allistics. A big problem in the US right now is that we have autistics getting into allistic things (tech billionaires) and allistic as getting into autistic things (data center glonzo).

0121

Reposted by srrrse

any time anyone more hype about AI than me says anything

Woody Harrelson in True Detective saying “I just want you to stop saying odd shit”

1411

Product providers pretty regularly design their products in a way which is “misaligned” with their users interests which they get away with because of the power and information differential across the exchange

020

(Source: staff.fnwi.uva.nl/p.vanemdeboa...)

staff.fnwi.uva.nl

000

Not trying to kill the vibe on your joke but I think it’s really interesting thing think about that quote in context. Knuth wrote the code on a typewriter and the “proof” is extremely informal. His process was more akin to writing a blog post than anything we would call formal verification.

220

you can order power tools on wolt now but im holding out for makitas

030