Author | Lightnews

Reposted by William Gunn

Yoshua Bengio @yoshuabengio.bsky.social · 18h

Europe has a chance to shape a safer and more values-aligned future for AI innovation, and needs to. This was my main message at the AI in Science Summit in Copenhagen today.
I also presented Scientist AI, LawZero's approach to create technical guardrails and help accelerate scientific discovery.

1 3 7

William Gunn @metasynthesis.net · 8h

🤯

Victoria Bosch @initself.bsky.social · 20h

Introducing CorText: a framework that fuses brain data directly into a large language model, allowing for interactive neural readout using natural language.

tl;dr: you can now chat with a brain scan 🧠💬

1/n

Reposted by William Gunn

Risto Uuk @ristouuk.bsky.social · 22h

𝗧𝗵𝗲 𝗘𝗨 𝗔𝗜 𝗔𝗰𝘁 𝗡𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿 #𝟴𝟵: 𝗔𝗜 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱𝘀 𝗔𝗰𝗰𝗲𝗹𝗲𝗿𝗮𝘁𝗶𝗼𝗻 𝗨𝗽𝗱𝗮𝘁𝗲𝘀 is live! CEN and CENELEC have announced exceptional measures to speed up the development of European standards supporting the AI Act.

The EU AI Act Newsletter #89: AI Standards Acceleration Updates

CEN and CENELEC have announced exceptional measures to speed up the development of European standards supporting the AI Act.

artificialintelligenceact.substack.com

1 2

William Gunn @metasynthesis.net · 19h

"We're not making some features available to you because some very vocal people who may not even be users of our product don't think anyone should have them".

two men are sitting in front of a crowd and one of them is talking on a phone

ALT: two men are sitting in front of a crowd and one of them is talking on a phone

media.tenor.com

1 1

William Gunn @metasynthesis.net · 1d

You're too modest 😉

1

Reposted by William Gunn

bryan newbold (🇨🇦 IETF/Montreal) @bnewbold.net · 1d

AT service operators and moderation thinkers: I put together an early proposal around infra abuse notices across organizational boundaries.

really looking for feedback on this one, it is bait for counter-proposals and references to prior work!

github.com

5 23 87

William Gunn @metasynthesis.net · 1d

It can take a while to be proven right sometimes!

1 1

William Gunn @metasynthesis.net · 1d

I'm putting a sign on my stove for Known Hot November.

totallynotseth.dev @totallynotseth.dev · 1d

Unfortunately @tangled.org will be down this month in honor of no knot november

brendan @schlage.town · 1d

starting on my research for known hut november

1

William Gunn @metasynthesis.net · 1d

Solving the blank page activation energy problem is a great use case.

David Buchanan @retr0.id · 1d

claude makes me more productive because I look at its output and then I'm like "damn this sucks, I guess I gotta write it myself"

William Gunn @metasynthesis.net · 1d

Disagreeing is fine, if that's what you're into, but the generative question is "How would we know?"

1 1

William Gunn @metasynthesis.net · 1d

Appreciate the exasperation, and also the history of claims of things that humans will always be better at is littered with corpses. Clever turns of phrase and great illustrations don't seem uniquely different such that they'd be resistant. Anyways: manifold.markets/WilliamGunn/...

Will humans remain better at clever turns of phrase and great illustrations, as Wired claims?

50% chance. Global Editorial Director Katie Drummond claims that "unearthing and communicating new and newsworthy information? Landing on the turn of phrase that says exactly what you want, so damn be...

manifold.markets

William Gunn @metasynthesis.net · 1d

My experience has been that they tend to mean computer science & ML research when they talk about automating science. Has anyone made statements explicitly including wetlab research?

William Gunn @metasynthesis.net · 2d

I'll just put this out there. There's a reason we have representative democracy and not direct democracy in the US. Blocklists are direct democracy. Perhaps different atproto servers should be able to elect moderators like we elect representatives?

Kuba Suder 🇵🇱🇺🇦 @mackuba.eu · 2d

Re: blocklists ⬇️

I don't like being added to lists like "MAGA/dipshits/reactionaries/genociders/scammers"… But I also don't want us to decide that "we're rolling back the whole composable moderation, people can't be trusted to self govern, let's have only centralized moderation by Bluesky PBC" 🫤

Me in a thread:

Yeah I really don't know what would be best to do about it… on one hand, this is what composable moderation was meant to be. People, communities moderating for themselves. And now they started hiding a lot of the lists, making them useless for sharing…

On the other hand, a lot of lists are pretty abusive. But it always has two sides, you want to be able to moderate people you find annoying/bad, but you get pissed when you're added to one because you feel you don't deserve it…

It needs to be somehow balanced, so that people and communities can moderate as they want, but it somehow prevents some negatives

William Gunn @metasynthesis.net · 2d

Christmas: reconnecting with family
Thanksgiving: reconnecting with friends
Halloween: connecting with neighbors

1 4

Reposted by William Gunn

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 2d

This labeler is such a clever way of making spam visible. Bluesky wins yet again: bsky.app/profile/stec...

5 39 160

William Gunn @metasynthesis.net · 2d

What fraction of people would you estimate are half-dead or otherwise living lives of quiet desperation?

1

Reposted by William Gunn

Dustin Moskovitz @moskov.goodventures.org · 2d

This dystopia is brought to you by a16z, the firm that is against ethics.

Leif Hancox-Li, PhD @struthious.bsky.social · 3d

This is Bad (source: network.fintechtakes.com/library/3d38...)

5 4 56

Reposted by William Gunn

Brian Nosek @briannosek.bsky.social · 3d

I love asking the question: how much should be spent on scientific replication? When asked, I have suggested that 1-2% of a funding budget felt right. "Felt right" meaning no evidence basis other than intuition.

@jdworkin.bsky.social offers a much more thoughtful approach.

How Much Should We Spend on Scientific Replication? | IFP

A data-driven framework for targeting replication funding where it matters most

ifp.org

2 7 24

William Gunn @metasynthesis.net · 3d

We now finally have actual semantic search, which is huge.

Timothy Gowers @wtgowers.bsky.social · 3d

I crossed an interesting threshold yesterday, which I think many other mathematicians have been crossing recently as well. In the middle of trying to prove a result, I identified a statement that looked true and that would, if true, be useful to me. 1/3

1

Reposted by William Gunn

MIRI @intelligence.org · 4d

”If Anyone Builds It, Everyone Dies” was recently added to the New Yorker's “The Best Books of the Year So Far” list!

newyorker.com/best-books-2...

2 4

William Gunn @metasynthesis.net · 3d

You Can Just Do Things lite.cnn.com/2025/10/30/t...

This veteran’s Halloween spirit gave him a new purpose. The result? $1 million for kids suffering with cancer | CNN

lite.cnn.com

William Gunn @metasynthesis.net · 4d

You could do worse than be a normie.

norvid_studies @norvid-studies.bsky.social · 4d

William Gunn @metasynthesis.net · 4d

Triangulating who someone really is by isolated statements they make is the primary activity of a lot of people here and the lack of epistemic rigor is really telling. 🤔

Reposted by William Gunn

Larry Levitt @larrylevitt.bsky.social · 5d

We estimate the average ACA enrollee will see their out-of-pocket premium increase by 114%. But, it will vary a lot based on age, income, family size, and zip code. I'd be interested in what you're seeing if you're an ACA enrollee and log on to the marketplace. Please reply here.

2 10 13

William Gunn @metasynthesis.net · 5d

So good until that last sentence. Representatives have to represent, and how do you propose they do that while blinding themselves to what their constituents think on the issue?