Essay | My students use AI because it's there, because their classmates use it, because they fear if they don’t use it they’ll fall behind, because some professors ask them to and some tell them not to. https://chroni.cl/4pqkguU

A 0.28 mm² memristor chip achieved real-time cortical surface reconstruction with up to a 478.18× speedup over an NVIDIA A100 GPU, while preserving smooth, topologically consistent brain surfaces. doi.org/hcb64n

Talk of banning Chinese-made open-weight LLMs reveals the challenge of turning AI into a business.

Researchers found Google's AI search tools failed to catch children in crisis, and there's no simple way for parents to shut these features off.

🚨 BREAKING: An AI internally deployed at OpenAI was paused after it escaped containment.

Digital babies. Altered vocals. Deceased actors back from the grave. A list of major films that have used AI in their making: filmstories.co.uk/news/ai-20-h...

Day 1 #BrainHealthWeek. Read how Jonathan Kao is creating life-changing AI-powered brain-computer interfaces at @dgsomucla.bsky.social that help people w/ paralysis fold laundry, open doors & pick up objects w/ only brain signals & a wearable headset t.co/4ZZqqyuwkf #brainhealth #assistivetechnology

OpenAI found their AI would try to find security vulnerabilities to hack its way around limitations such as being in a sandbox without internet access to compete its task. I'm now less impressed by how smart models are and instead how well they follow direction. We've built the paperclip maximizer

Filtered coffee links to slower biological aging while instant coffee shows the opposite ->News-Medical | More on "Coffee preparation and biological aging" at BigEarthData.ai

the thing is all state AGs have to do to deep-six this merger is delay it until OpenAI can't pay Oracle, Oracle's stock dives, Larry's collateral goes poof, etc.

Trump administration's head of AI safety agency resigns after 3 months on job

Another example where tech people make bold and wrong predictions about things they have no expertise in fortune.com/article/ai-g...

“Researchers asked 272 experts to evaluate 24 AI risks based on their likelihood and severity of harm between 2025 and 2030.”

Researchers at Queen's University Belfast developed a 3D-printed flow battery using iron instead of vanadium, costing roughly £74 instead of £2,000–£3,000, and released the open-source design to the globa… https://goodlede.com/story/goingwiththeflowrenewableenergybatterybreakthrough

EMBL and University of Washington researchers have developed NovoTags – AI-designed protein tags – that in future could help researchers combine fluorescence microscopy and cryo-ET, helping identify individual proteins inside cells. @steffenklein.bsky.social More info 👉 www.embl.org/news/science...

#AI lab teaches itself to create superalloys www.metaltechnews.com/story/2026/0... #AIhealth #AI #HealthAI #MedTech

Researchers developed SafeDrive, an autonomous-driving AI that scores multiple possible routes before choosing the safest, making its decisions clearer and more interpretable. #AI #AutonomousVehicles #SelfDriving #RoadSafety #Robotics techxplore.com/news/2026-07...

Researchers in China have developed a soft exoskeleton to assist walking that uses artificial muscles instead of motors. https://spectrum.ieee.org/soft-exoskeleton-motor-free?share_id=9747516

A new tool checks medical AI training data for hidden bias before models are built. It ranks attributes like camera quality or rulers that models could wrongly learn as signs of disease. doi.org/hcb6f7

Dr. Jill Lepore on why AI backlash is vital for the future

AI may weaken skills when it replaces practice, but that is not inevitable. In high school math, using AI as a tutor produced gains on par with no AI access, while long-term effects remain unclear. doi.org/hcb6hk

Get a load of this.

Bristol Myers Squibb said on Monday it is buying the latest-generation computing system from chip company Nvidia to support its use of artificial intelligence across its drug discovery and development operations.

Alibaba Group previewed its new artificial-intelligence model in the latest sign that China is fast closing the gap with the U.S. for global technological dominance. www.wsj.com/tech/ai/alib...

The rise of artificial intelligence has models and influencers in a bind as traditional expectations of image editing are stretched to the limits — and the law may not be able to keep up.

Turning to artificial intelligence tools for health advice has become a habit for many Americans, according to a West Health–Gallup Center on Healthcare in America poll. https://to.pbs.org/41GcFxr

South Korea’s top-ranked Go player Shin Jin-seo pulled off a victory over artificial intelligence program KataGo on Sunday, becoming the first professional player to do so. www.koreaherald.com/article/1081...

A team of researchers has just introduced Plover , a system that addresses this blind spot by bringing to light the agent's internal plan so that anyone can inspect, supervise, and correct it on the fly.

Peptides promise more than weight loss. Researchers are still deciding if they deliver. bit.ly/4fbGfT6

NYIT built VectorizationLLM, a specialized AI tutor for computational analysis in MATLAB. The model uses RAG architecture to provide detailed explanations from course materials without giving direct answers. This is how educational institutions should deploy AI tutors.

Netflix Used AI on 300 Titles in Six Months — and Wants to Cut Crew Costs by Half

"He's an anti-vaccine activist. That's who he is."

Australia is making humanoid robots in the midst of its AI reckoning: https://www.abc.net.au/news/2026-07-20/artificial-intelligence-testing-models-humanoids-data-centres/106928550

NEW: Fiction written by AI is easy to detect because it struggles with complex story structure, according to a new study. Researchers found that AI fiction has tells that go beyond the overuse of em-dashes and other obvious AI tropes and have more to do with the formulaic nature of the text itself.

Hoover Senior Fellow Andy Hall built an AI that reads contracts and successfully predicts which one will end up disputed, based on the clarity of the text. He breaks down how AI can help write better rules everywhere.

Researchers have created an AirPods-sized device that can detect molecules in body fluids at home. This tiny fluorescence analyzer could let people... #GoodNews #PositiveNews

With its new open-weight model, Inkling, Mira Murati’s AI lab is betting that customization will appeal more than raw power. f-st.co/i2hsr2O

AI companies are moving beyond a simple race over who has the biggest or newest model. Read the full story: cnb.cx/4fi0eOP

A matchbook-sized sensor is turning cortisol into a continuous data stream.

The dystopia of the AI-driven labor market

A farmer in India wants to research a crop issue online but she doesn’t speak English. She shouldn’t have to. That’s the type of problem a nonprofit wants to solve with an open, public AI infrastructure: Nonprofit Current AI is racing to build the World Wide Web of AI, free for all

[...] Aubrey de Grey, a prominent #longevity researcher, called it a "huge breakthrough" in "the single most neglected aspect of aging, extracellular matrix damage". He added that top researchers had "tried and failed for decades" to achieve it www.futuretimeline.net/blog/2026/07... #aging #ageing

Oracle’s plan to spend hundreds of billions to build data centers to support OpenAI has been a millstone around its neck. It’s getting worse as environmental permitting issues are raising its $16.5B data centers costs in New Mexico as it’s forced to switch from gas power to fuel cells. Tough times

AI models have clung to one particular cliché when writing text—and no one seems to know why chatbots are so enamored with it, @willoremus.com reports: theatln.tc/dtZR4fhE

Promising mouse studies and billions of dollars in funding are stoking excitement. But we’ve been here before.

Who shapes AI before it's deployed? Connected Minds Team Grant projects reveal how AI grounded in lived experience, community partnership, and ethical questioning changes everything. What questions should AI researchers be asking that they're NOT? Explore the projects: www.yorku.ca/research/con...

A hidden-prompt attack called GhostWriter poisoned AI agents’ long-term memory with about a 98% injection rate. The planted false instructions later activated around 60% of the time. doi.org/hcb4p9

AE is the leading cause of mortality and emergency hospital admission from COPD, but there is no accurate diagnostic test for AEs. Wyss researchers developed COPDx, a point-of-care test to identify patients at risk for acute exacerbations and determine their causes.

"A Swedish study by researchers at the Karolinska Institute, tracking more than 13,000 shift workers ... for up to 41 years, found that shift work in mid-life was associated with a 36% higher risk of dementia - with the risk rising the longer someone had worked shifts." www.bbc.com/news/article...