Synthetic Intelligence
In This Manhattan Lab, AI Designs Supplies From ScratchAdele Peters | Quick Firm
“The lab makes use of customary supplies science gear, nevertheless it’s nearly all automated and run by AI; if it has a brand new concept at 4am, it begins working once more. It will probably run as many as 50 experiments in a day, and the group is aiming to extend that to 100 experiments a day by the tip of the summer season. A human supplies scientist, Krause says, may do 50 experiments in a yr.”
Biotechnology
One-and-Performed Coronary heart Illness Prevention? Scientists Present It Might Be Attainable.Gina Kolata | The New York Occasions ($)
“In a small, preliminary research, an experimental gene-editing therapy dramatically lowered levels of cholesterol, maybe completely, after only one infusion, scientists reported on Monday. If confirmed in bigger research, researchers hope the findings could result in a one-and-done option to stop coronary heart illness in giant numbers of individuals.”
Robotics
3D-Printable Humanoid Legs Let Robotics Experiments Run WildJeremy Hsu | Ars Technica
“A $2,500 pair of humanoid robotic legs constructed from 3D-printed components and off-the-shelf parts will not be going to win marathons simply but. However such comparatively cheap {hardware} may allow researchers to extra simply check and practice AI-powered robotics software program in a bodily physique throughout real-world experiments.”
Biotechnology
Pancreatic Most cancers Halted by Virus Injection in Three SufferersAlice Klein | New Scientist ($)
“Additional analysis is required in bigger trials, however the early outcomes are encouraging, particularly since solely small doses of the virus have been administered for preliminary security testing. ‘We solely injected one-tenth of the dose we’re finally aiming at, so the efficacy is best than I anticipated, particularly as that is pancreatic most cancers,’ says Masato Yamamoto on the College of Minnesota, who led the event of the viral therapy.”
Future
A Actuality Test on the AI Jobs HysteriaDavid Rotman | MIT Expertise Evaluation ($)
“Haven’t you heard? White-collar jobs are going away, decimated by AI. …However earlier than you give up your job as a software program developer or monetary analyst—or tech journalist—and look to hitch the plumbers’ union, it’s value contemplating as we speak’s financial analysis on whether or not synthetic intelligence has truly begun to devour white-collar work. The brief reply is: No.”
Synthetic Intelligence
The AI Superstars Who Say a ‘Vibe Slop’ Disaster Is ComingChristopher Mims | The Wall Road Journal ($)
“Two engineers who constructed the core of the massively widespread OpenClaw AI agent have a stark warning: The synthetic intelligence supposedly able to changing well-paid software program builders is flooding the world with dangerous, probably even harmful, code. It’s a phenomenon they name ‘vibe slop’—a mix of ‘vibe coding,’ creating software program with AI instruments by describing it in plain English, and ‘AI slop,’ the limitless, low-value AI-generated content material throughout social media.”
Future
Mirror Life: Scientists Conflict Over Risk of Lab-Engineered Micro organismJames Woodford | New Scientist ($)
“Microbes based mostly on mirror pictures of molecules within the pure world would have a tough time surviving exterior the laboratory, in keeping with a modeling research. To take action, they would wish a prepared provide of ‘mirror meals,’ or some novel option to feed themselves. However the analysis has drawn a backlash from different specialists within the discipline who warn that it could underestimate the grave dangers posed by so-called mirror life.”
Tech
Uber President Says AI Spending Is Getting ‘Tougher to Justify’Jess Weatherbed | The Verge
“After reportedly exhausting its annual AI price range simply 4 months into 2026, Uber is now questioning whether or not it’s truly seeing significant returns on its investments. In an interview with Fast Response, Uber president and chief working officer Andrew Macdonald mentioned the corporate isn’t seeing a connection between rising token consumption for Claude Code and extra helpful options being delivered to shoppers.”
Future
Illinois Lawmakers Simply Handed America’s Strongest AI Security InvoiceMaxwell Zeff | Wired ($)
“The Illinois Home of Representatives handed a invoice on Wednesday requiring frontier AI labs like OpenAI, Anthropic, and Google DeepMind to have their security practices audited by a 3rd social gathering. If signed into regulation, AI security specialists inform Wired, it will be the nation’s main examine on the ability of main AI corporations.”
Synthetic Intelligence
RSI Is the New AGI—and It is Simply as Onerous to Pin DownRussell Brandom | TechCrunch
“The phrase ‘recursion’ is the most recent buzzword in AI circles. Two separate startups have taken on the identify, and plenty of extra have began referencing recursive self-improvement (RSI) of their roadmaps. Like AGI earlier than it, RSI has grow to be a three-letter byword for a cataclysmic AI takeoff—even when there’s nonetheless somewhat disagreement about what it precisely means.”
Synthetic Intelligence
I am a Skilled Reality-Checker. AI Is Flawed Extra Than You Suppose.Meghan Herbst | Wired ($)
“Over the previous yr or so, increasingly more individuals have checked out me with nice pity. Certainly a fact-checker at {a magazine} isn’t lengthy for this AI-upgraded world. Name me silly, however I’m not that anxious. Little or no of humanity’s collective information, I’ve concluded, lives on the web. And in keeping with my analysis, AI is much more mistaken than individuals may suppose.”
Area
Thousands and thousands of Planets May Kind Round Supermassive Black HolesJonathan O’Callaghan | New Scientist ($)
“Finally planets would start to develop in enormous numbers, and with unusual properties. ‘This can be a actually superb new pathway to type very alien planets,’ says McKernan. ‘If these items exist, they’re fairly in contrast to planets that we all know and love.'”
