"Beachgoers warned against swimming after pollution alert." The "bathing season" finishes at the end of the month, after that you're on your own, none of our beaches will be tested again until next May. You have been warned. bbc.co.uk/news/articles/cm27…

‘We have never seen anything like this’ Fire chiefs warn wildfires are increasingly threatening homes and businesses, as Sky News analysis finds almost a fifth of this year's large wildfires came dangerously close to built-up areas. news.sky.com/story/we-have-n…

Paying peer reviewers = 85% faster editorial decisions. And payments gave high-quality reviews. (the story is described in Nature) At Biology Open, peer reviewers were paid for high quality reviews. $290 per peer review. Miraculously, the time to editorial decisions DROPPED: - Unpaid reviews: 38 days. - Paid reviews: 5-6 days. And review quality WENT UP. 📍How it works: “When researchers apply to become paid reviewers for Biology Open, we screen their CVs and check whether they’re legitimate and have peer-review experience. We currently have 500 pre-contracted reviewers from all over the world. Ideally, what I envision is a system where reviewer applicants undertake a manuscript-reviewing test. And we’d review that review, so if it scored high enough, we’d hire them.” - Editor-in-chief Daniel Gorelick. Nature: “Do you think paying reviewers could work in other journals and in other disciplines?” Editor-in-chief: “Absolutely. There’s been such a profound aversion to paying peer reviewers, whether it’s on principle or because people think it will never work. We’ve demonstrated that it works.” So... At last, there’s light at the end of the tunnel. __

A biology journal trialled paying their reviewers and found it led to faster first editorial decisions – and increased review quality. go.nature.com/4v7XkSv

Someone needs to start selling a clip on or Velcro on outdoor window screen that can cover windows to keep out the sun when it’s very hot. Needs to be easy to remove when not needed and cheap! They’d make a mint 💴

Good news. London has doubled the area of wildflower verges beside its roads in the last two years to boost biodiversity. Butterfly populations are booming as a result. ianvisits.co.uk/articles/bee…

Probably the best example of a rapid bottom-up energy transition: Pakistan imported more Chinese solar last year than any country on earth. No big subsidy, no rooftop scheme, no feed-in tariff. Solar is now around a quarter of electricity. More here👇 thenewworld.co.uk/jan-roseno…

Cardiovascular complications from covid-19 have been seen widely since the early days of the pandemic. Katharine Lang reports on why covid-19 is “a vascular disease masquerading as a respiratory one” bmj.com/content/392/bmj.s31?…

Fei-Fei Li says language models are extremely limited. This @GoogleDeepMind paper makes almost the same point, just in the world of video. The models are just very advanced pattern matchers. They can recreate what looks like reality because they’ve seen so much data, but they don’t know why the world works the way it does. The models can generate clips that look stunningly real, but when you test whether they actually follow basic physics, they fall apart. The Physics-IQ benchmark shows that visual polish and true understanding are two completely different things. Here, the authors build Physics-IQ, a real-video benchmark spanning solid mechanics, fluids, optics, thermodynamics, and magnetism Each test shows the start of an event, then asks a model to continue the next seconds. They compare the prediction to the real future using motion checks for where, when, and how much things move. Scores then roll into a single Physics-IQ number that caps at what 2 real takes agree on. Across popular models, even the strongest sits far below that cap, while multiframe versions usually beat image-to-video versions. Sora is hardest to tell apart from real videos, yet its physics score stays low, showing realism and physics are uncorrelated. Some cases work, like paint smearing or pouring liquid, but contact and cutting often fail. arxiv. org/abs/2501.09038

The 2025 BEHAVIOR Challenge, launched by @drfeifei's team at Stanford, is a global competition to train robots in household tasks, testing reasoning, navigation, and manipulation in simulated homes. ⦿ 50 tasks, 1,000 activities (e.g., cooking, cleaning) ⦿ 10,000 expert demos (1,200+ hours) ⦿ Baselines methods like ACT, OpenVLA, and π0 The submission deadline is November 15th. behavior.stanford.edu/challe…

"In 2010, the IEA estimated that there would be 410GW of solar panels installed around the world by 2035. There is already more than 4 times that capacity, with about half of it in China" - 10 years earlier ft.com/content/c4068478-c091…

It’s official!! 200 London schools just got a lot better! Massive investment into improving the air quality of hundreds of London schools has just been made. Well done @MayorofLondon!!! Way to take the lead on this incredibly important issue!

🚨 LIVE right now from the UN Headquarters in NYC HEALTHY INDOOR AIR: A GLOBAL CALL TO ACTION Global leaders, policymakers, scientists and health experts will come together on the sidelines of the UN General Assembly for this event. Live broadcast here: webtv.un.org/en/asset/k1v/k1…

A research team from València has developed an innovative biosensor capable of detecting airborne viruses in real time and at low cost, without the need for chemical markers or laboratory procedures. This sort of technology could be a total game-changer! upv.es/noticias-upv/noticia-…

In a world of answers how do we teach better questions? Teaching Dialogic Intelligence with AI open.substack.com/pub/rupert…

The amazing Lydia Cao, now working at Toronto University, has a new paper in Computers and Education on how to use simulations to help prepare teachers for the 'contingent responsiveness' that is required by dialogic education sciencedirect.com/science/ar…

My new baby!

Have you been diagnosed with Lobular Breast Cancer? Join us on 14 February in London where Future Dreams’ in-house Breast Care Nurse will answer any questions about treatment & Claire and Emma of LBCUK will share the latest Lobular Breast Cancer research. buff.ly/4gSG2Ta

Take a look at our book! Artificial Intelligence in Education: The Intersection of Technology and Pedagogy link.springer.com/book/10.10… #AIinEducation #EdTech #AIPedagogy #TeachingWithAI #FutureOfEducation

Looking forward to attending this event tomorrow where I will be speaking briefly about the intersection of CI and AI - cic.uts.edu.au/events/collec…