Never bet against the naked emperor. He might be naked but while everybody knows it, everybody bets longer on him than you can stay solvent...

A Moroccan Tajine, from initial preparation to cooking to serving (courtesy of my mother-in-law).

Day 3 of #CCN2026. Alona Fyshe kicks us off with a keynote questioning the idea that LLMs are 'mere' stochastic parrots

European vacation is all fun and games but at some point you don't want the GPU queues to dry up ...

Rebuttals and discussions need to be reduced an extreme minimal length. CVPR style, 1 single PDF page. No discussion. Else, human ACs and reviewers will have a disadvantage compared to the world of agents out there now.

When you need to remember a 6-digit number for a couple of seconds or up to a minute and then recall it, do you (a) just remember it, (b) see it visually, or (c) replay the sound of the sequence? For me it is (c).

This is an excellent read on the (possible) future of mathematics in the age of AI.

Clément Canonne@ccanonne.github.io · 2d ago

FWIW, here are Terence Tao’s slides at the ICM on maths and AI:  teorth.github.io/tao-web/slid... @teorth.bsky.social (I am not endorsing nor criticizing the content, but this is a useful and thoughtful set of points and views in the discussion, from someone who has deeply engaged with the matter)

Surfing in Morocco in a nutshell (son and his cousin).

Academic bibliometrics in a nutshell.

Tom Gauld@tomgauld.bsky.social · 3d ago

My latest books cartoon for @theguardian.com

"We're all worried," as what it means to do research (in my field, Theoretical CS) seems to be shifting, and shifting fast. What to do? Senior researchers must lead by example, knowing that not everything will pan out. What I'm suggesting below may not work everywhere, but here's my own advice: 1/

I need to take ownership of my mistakes. I used computer. I think this happened because I'm a normal person. I know I've disappointed many of my fans who only watch my videos between praying and feeding the hungry. I'm sorry for being such a piece of shit

It turns out that 46 deg Celsius is REALLY hot. Being outside is just unbearable. Drove in a car w/o AC: holding your arm out of the open windows is also not possible. It feels like putting it into an air dryer. Jumping into the water: priceless 💦

Doing some of my famous relatable humor

The gaming industry is breaking apart because of the RAM prices driven up by AI 😬 (1 datacenter uses as much RAM as 8.3 million PS5) www.youtube.com/watch?v=Eo5w...

UEFA 1 : FIFA 0 Interestingly Italy had already anticipated this boycott...

ICLR 2027 has authorship quotas: - No more than 20 submissions per author - No more than 1 submission where no author has a paper previously accepted to a major ML conference From iclr.cc/Conferences/...

Reviewer #2

LLMs do a really good job of finding weaknesses, they are better then many humans in this. However, they are TERRIBLE in assessing whether weaknesses are minor and major and whether a paper is worth publishing or not. Review forms should enforce this through structure, minor vs. major fields.

😳

When in the last year I read 'Twitter is so toxic' I thought, ok the BSky crowd, they exaggerate and are special. I hadn't left Twitter bc of the language there but bc of Musk. However, I now went back a bit lately and, OMG, why would anybody read that? Horrible interactions, propaganda bots...

Evening reading. I’ve been looking forward to reading this paper. @chriswolfvision.bsky.social @weinzaepfelp.bsky.social

This is one of the rare ways Europeans can flex: real vacation 💪

Agreed but I still like reading real paper books and I don't give it up... (From The Expanse series)

Everyone who knows how to use a computer is now saying “well, why shouldn’t I use AI to write a paper and send it to a conference? Worst case it’s rejected.” The wave is coming.

Become a partner of #3DV2027 in Thessaloniki - #CallForSponsors - Your company logo prominently displayed - Complimentary conference registrations - Exhibition space - Many more options… 💡More info: 3dvconf.github.io/2027/partner...

1/6 Excited to share that our paper on model merging was accepted at ECCV 2026! 🎉 We introduce an efficient, decoder-free proxy that makes model selection faster, simpler and practical across vision tasks. 📄 arxiv.org/abs/2604.12935 🌐 europe.naverlabs.com/task-alignment 🧵👇

I absolutely love cycling and the Tour de France but I hate that cyclists need to carve space through the crowds because of the obnoxious behaviour of fans.

Goodhart's law states that "When a measure becomes a target, it ceases to be a good measure". Conference peer-review is a measure, arguably now mostly implemented through LLMs, and will become a target. Ergo, future academic research will target what Claude finds interesting and acceptable.

Thanks @elliot-eu.bsky.social for the short interview on why egocentric vision is the right place to start your exploration into understanding human actions youtu.be/DfbxlgWpKSQ

Why first-person AI matters, by Dima Damen

YouTube video by ELLIOT Project

youtu.be

ELLIOT Project@elliot-eu.bsky.social · 2w ago

How can AI learn everyday actions from a person’s point of view? 👀 In this ELLIOT interview, @dimadamen.bsky.social (University of Bristol) explains why first-person vision matters and why translating human actions into language remains a major challenge for multimodal AI. Link in comments 👇

We may have to give up precise understanding and focus on the broader picture if we want to keep up progress. In the sense that you cannot build knowledge of what a fluid is doing by looking at every single molecule, or build knowledge of what a DNN is doing by looking at every single activation.