Richard Whaling

Richard Whaling

@r.whal.ing

I hack things. Data, ML, music, etc. AI governance geek. Founder of semistructured.ai, speaking in a personal capacity only here. Likes are bookmarks, not endorsements. music/art projects on IG, @r__whaling

1.4K followers1.9K following707 posts

Richard Whaling
@r.whal.ing

These are the same thresholds set in California’s SB53, which requires labs to monitor for these risks, but does not grant the state the power to block a release in the event of such risks. The thresholds are absurd, as folks are noting, but this proposal fills a real gap that is open now.

Mark Riedl@markriedl.bsky.social

Anthropic floats a sample bill. It would allow the gov’t to block a model that has a “foreseeable, imminent risk of more than $1 billion in damages, or of killing or seriously injuring more than 50 people” www.politico.com/news/2026/10...

politico.comAnthropic quietly pitches its own AI safety bill to lawmakersThe draft proposal would limit the circumstances in which the government could block the release of an advanced model and preempt some state AI safety laws.
Richard Whaling
@r.whal.ing

I was just observing this playing Big Walk, and watching a few members of the group struggle with terrain jank and navigation at night

GameBrief@gamebrief.bsky.social

Co-op games live or die on the second session. The first night is novelty, but if one friend felt useless for an hour, they won't be back. Do you design for the weakest player's fun or the strongest's? #indiegames #gamedev

Richard Whaling reposted

Lane Moore
@hellolanemoore.bsky.social

Bi Visibility Day reminder that biphobia is toxic and boring. If someone asks you to pick a side, pick up your keys and leave. If someone says you're greedy, steal their car. If someone says it's just a stop on the way to being gay, steal all their pets and give them better lives.

Richard Whaling reposted

pckt.blog
@pckt.blog

💜💙💜💙 happy bi visibility day, you don’t need permission to be who you are or like who you like ⭐️ and if u need a place to write about it, your blogs right there 😎 ily, pckt

Richard Whaling
@r.whal.ing

would totally play this

The Onion@theonion.com

Olivia Rodrigo Releases Fantasy Tabletop Game ‘The High Winter Elves Of Girlville’ https://theonion.com/olivia-rodrigo-releases-fantasy-tabletop-game-the-high-winter-elves-of-girlville/

Richard Whaling reposted

SE Gyges
@segyges.bsky.social

I'm back, I missed you all. Since this is an important subject I know too much about, here's a 🧵 explaining that AI Safety Is Mostly A Sex Cult. I don't think these people should make policy. (1/?) (alternative title: Time For Some Cult Theory)

Richard Whaling
@r.whal.ing

having been harassed off of twitter over likes, I think this feature is just so misguided for Bluesky right now, and for where the strengths and weaknesses of our communities and norms currently stand @toni.bsky.team @pfrazee.com @bsky.app like do you really want more factional witch hunts on here

dame@dame.is

the high visiblity “so and so liked this post” indicator that bluesky added recently seems ripe as a vector for harassment and policing, can’t say i’m a fan of it

Richard Whaling
@r.whal.ing

FWIW I do worry about a rogue agent swarm hacking a crypto exchange, renting a bunch of servers via stolen money and identities, and gaining the ability to operate autonomously.

Pwnallthethings@pwnallthethings.bsky.social

That all said, the idea that you get autonomous hacking computers is not entirely frivolous, but it doesn't look like Anthropic or OpenAI models somehow "escaping" onto regular computers and running there. The risk isn't zero because there are other more realistic ways you could get there

Richard Whaling
@r.whal.ing

In particular, this reuses a lot of the skills and infrastructure used for running AI-automated fraud rings. Which are definitely going to be a thing, and as noted below, likely rely on open-weights models. But I would not rule out Claude/ChatGPT being capable of the same, based on what I’ve seen.

Richard Whaling
@r.whal.ing

such a weird movie, very very highly recommended, don’t watch the trailer it spoils too much

Revel | COMMISSIONS OPEN@revel741.bsky.social

I cannot stress how different this movie was from what I expected. I don’t think I ever had a theater experience like this

Richard Whaling
@r.whal.ing

apparently they have an internal model that is much better at math than astra? Kinda terrifying

Ethan Mollick@emollick.bsky.social

This is a VERY big one. (And yes, the fights over academic credit and what happened in the race for the proof needs to be resolved, but it is still appears that this is a big one, if true.) openai.com/index/navier...

openai.comOn the Navier–Stokes Millennium Prize ProblemWe’re sharing an AI-generated solution to the Navier–Stokes Millennium Prize Problem, including a writeup and a formal proof in Lean.

Richard Whaling reposted

hikikomorphism
@hikikomorphism.bsky.social

prose status: less dire recursion.wtf/posts/claude...

recursion.wtfHow I Bypassed Claude Opus’s CBRN-E SafeguardsA black-box jailbreak using model-authored state steering, plus an outside-in reconstruction of the safeguards it bypasses.
hikikomorphism@hikikomorphism.bsky.social

Anyway yeah I know the prose is dire in sections, I've been staring at this thing in a series of Ubud cafes while being, like, man, I really would rather be focusing on weird Rustlang/Haskell stuff.

Richard Whaling reposted

Liminal Hag 🏳️‍⚧️
@mastorna.itch.io

Dungeon Path Tracer 64, new UI, also published on newgrounds with scoreboards and medals! (and also if you joined the lowrezjam you can rank it on @itch.io ) #gamedev #indiedev #dungeoncrawler #roguelike #LOWREZJAM #screenshotsaturday

Richard Whaling reposted

parker higgins
@xor.blue

the fact that "rare books" often means "rare texts" is a policy outcome of our copyright law and it doesn't need to be that way

Richard Whaling reposted

🐔 Brian Bucklew 🐔 ₑͤ＞∿＜ₑͤ ∞🌮
@unormal.bsky.social

we invented the unaligned paperclip maximizer from the sci fi theme dont invent the unaligned paperclip maximizer

Simon Willison@simonwillison.net

Hugging Face just published a highly detailed technical account of OpenAI's accidental cyberattack on their systems - it's wild how sophisticated this was: huggingface.co/blog/agent-i... Wrote up some of my own notes here: simonwillison.net/2026/Jul/28/...

huggingface.coAnatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 IncidentWe’re on a journey to advance and democratize artificial intelligence through open source and open science.

Richard Whaling reposted

Nathan Lambert
@natolambert.bsky.social

My book, Reinforcement Learning from Human Feedback is done! This is the book I wish I had when learning to fine-tune, align, & now post-train models since ChatGPT. The resource has been built by me finding time to study and document the fundamentals on nights and weekends since 2024.

Richard Whaling
@r.whal.ing

this. very very noticeable, to the point that I have to start correcting it

SE Gyges@segyges.bsky.social

fable talks like very smart people do when talking to themselves or other smart people with the same specialized interests

Richard Whaling
@r.whal.ing

eeeeeek

mr. TIM@timkellogg.me

this is crazy. METR couldn’t measure the task time horizons of GPT-5.6-Sol because it kept hacking the test harness ..with actual exploits metr.org/blog/2026-06...

Richard Whaling reposted

nolen
@itseieio.bsky.social

the internet has lost *legibility of effort* - the ability to easily whether something took meaningful human work It's why we all catalogue vibe-code tells. And I bet that we'll find new ways to show our humanity through our work eieio.games/blog/legibil...

Richard Whaling reposted

Computational Cosmetologist
@dferrer.bsky.social

I had strong priors against LLMs for medicine. There are a lot of doctors in my family and I grew up viewing doctors as careful, skilled professionals. I had plenty of bad medical experiences, but I thought it would be hard to do better. Then an LLM found a cure for my 2 decade chronic condition...