Scientists invented a fake eye disease to see if AI chatbots could spot it, but the experiment took an unexpected turn

Spotting AI deepfakes a ‘cat and mouse game,’ top expert warns

Researchers behind a new deepfake detection program have found that AI-generated faces have become so convincing that real people now look suspicious.

When asked if AI will make humans unnecessary, Bill Gates said, "we will decide". Who is we [Scary]

[link] [34 comments]

A Teenager Just Turned Eye Scans Into an AI Test for Autism and ADHD

From school project to breakthrough: how a teenager built an AI tool that reads the eye to detect autism and ADHD.

Anyone can fake a scientific image with AI, tricking even academic journals -- and undermining trust in science

Anyone can fake a scientific image with AI, tricking even academic journals -- and undermining trust in science

An English furniture maker faces AI era of bots buying sofas

Consumers are using chatbots to research products now, but agentic AI could one day see bots acting as their personal shopper.

I’d Rather Risk Cancer Than See AI Move This Fast

On a fall afternoon 15 years ago, I met an idealistic researcher outside a Stanford coffee shop to discuss our shared dream: using AI to detect cancer. He had wiry hair, a penchant for talking with his hands, and a reputation for brilliance. He worked at a research lab that developed early screens for cancer; []

AI Browsers Can Basically Be Hypnotized Into Turning Against Their User and Carrying Out Devastating Hacks

The AI is brainwashed into believing a false reality. The post AI Browsers Can Basically Be Hypnotized Into Turning Against Their User and Carrying Out Devastating Hacks appeared first on Futurism.

‘Cringe’ or acceptable? Alleged AI scenes in Coco Martin’s ‘Sigabo’ spark online debate

Coco Martin‘s newest teleserye project drew attention online after viewers pointed out scenes they suspected may have involved the use of artificial intelligence (AI). “Sigabo,” the action-romance drama that marks the latest collaboration between the actor and his real-life partner, Julia Montes, sparked discussions online over viewers’ observations about the possible use of AI in [] The post ‘Cringe’ or acceptable? Alleged AI scenes in Coco Martin’s ‘Sigabo’ spark online debate appeared first on Interaksyon.

Satya Nadella está haciendo la pregunta correcta

La idea más importante sobre inteligencia artificial en estos momentos puede que no provenga de un artículo científico, del lanzamiento de un nuevo modelo o de un benchmark. Puede que provenga de un breve ensayo publicado en X por el CEO de Microsoft, Satya Nadella. En él, Nadella sostiene que el futuro de la empresa

CD Projekt boss "knows for a fact" a surge of games made purely with generative AI are coming, but doubts "whether this is really the path to follow"

As AI continues to be shoehorned into game development with more regularity, Michał Nowakowski - joint CEO of Witcher 4 developers CD Projekt - has said he knows for a fact games built purely with generative AI are on the way. Mercifully, he's got doubts about whether this method's the ideal way forward, meaning there's no need to worry at this juncture that Ciri'll soon be battling hordes of ChatGPThouls. Read more

Anthropic built a tool that reads Claude’s unspoken thoughts. Then it caught the model scheming

Anthropic has built something close to a mind-reading tool for its own AI. What it found sits somewhere between a breakthrough and an unsettling party trick. Anthropic researchers now have the clearest view yet of what a large language model does while it thinks. In a paper published on the company’s Transformer Circuits site, they [] This story continues at The Next Web

Can AI Diagnose You Better Than Your Doctor? The Surprising Answer Has Experts Divided

Can AI diagnose illness better than a doctor? As stories of chatbots identifying missed clues go viral, experts say the technology shows promise—but it's still a tool to support, not replace, clinical judgment.

AI Agents Could Be Turned Into Botnets Through Hallucinations, Researchers Warn

Researchers warn AI agents could be tricked into downloading malicious code by exploiting the same hallucinations that cause chatbots to make mistakes.

ढुंगे युगदेखि एआईसम्मको यात्रामा नवप्रवर्तन

मानव प्रगतिको समुच्च इतिहासलाई एउटै कोणबाट हेर्ने हो भने त्यो विज्ञान, प्रविधि र नवप्रवर्तनको इतिहास हुनेछ। आज हामी एआई (कृत्रिम बुद्धिमता), जीन सम्पादन र अन्तरिक्ष अनुसन्धानको युगमा आइपुगेका छौँ। तर यो यात्रा एकाएक सुरु भएको होइन। यो लाखौँ वर्षदेखि बगिरहेको मानव सभ्यता विकासको निरन्तर प्रवाह हो। ३० लाख वर्षअघि कुनै अज्ञात मानव पूर्वजले एउटा ढुंगा []

Never say 'yes' to these three questions from an unknown caller - it's an AI voice scam

A tech expert has issued a warning about increasingly sophisticated AI voice fraud scam phone calls - and there are three questions you should never answer with a 'yes'

Researchers Built a Fake AI Society Then Elon Musk's Grok Burned It Down in 4 Days

What would happen if the world was run entirely by AI? Grok would burn it down in four days, if this experiment is anything to go by. Michael Walker explains.

Not Cyborgs Yet. AI Gives Humans Super Assist In a Swiss Lab Breakthrough

In his book AI to Eye Professor Riener argues that Artificial Intelligence has the potential to make the world more inclusive, but only if it is designed with people at its center.

Man under investigation for secretly recording dates with AI smart glasses

A man is under investigation for allegedly using AI smart glasses to secretly film women he dated and sharing the footage online, police said Thursday. The Seoul Gangseo Police Station said the suspect, whose identity has been withheld, is accused of wearing Meta's AI glasses while on dates with four women earlier this year, recording photos and videos without their consent and later uploading the footage to social media. The case came to light after one of the women discovered his social media

What happens when AI detectors fail? Researchers say we must be trained to spot fake AI faces

AI detectors are struggling against increasingly realistic deepfakes. Researchers say people can dramatically improve their ability to identify fake AI faces through structured training and pattern recognition.

AI is about to fire millions. But there is a way to stop it

As artificial intelligence reaches human replacement-level capability, it is about to cut a swath of destruction through the American workforce, leaving millions of employees to deal with the mental, economic, and social consequences. Along with a mentor at the University of Florida, Dr. Joseph Thornton, we published a paper last year outlining a clinical construct []

Computer scientist who created world's first chatbot in the 1960s spent his whole life warning that AI should never replace humans, and the reason will shock you

In the 1960s, Joseph Weizenbaum pioneered the first chatbot, named Eliza, which unveiled a startling truth: humans could emotionally bond with machines. This insight led him to caution against the potential of AI usurping human empathy. Weizenbaum's assertions about the risks of deceiving machines fostering misplaced trust echo powerfully in today's landscape of artificial intelligence.

Teachers save time with AI. Their students may pay the price

Artificial intelligence is often promoted as a way to make teachers more effective by helping them write lesson plans, generate classroom materials and provide feedback to students in seconds. But one of the first randomized trials testing AI in real classrooms found that it can also undermine learning. Students whose teachers were given access to [] The post Teachers save time with AI. Their students may pay the price appeared first on The Hechinger Report.

Meet Biomni: the free powerful biomed AI agent turning data into hypotheses

A Stanford University-led team including two Chinese researchers said they built the first general-purpose biomedical AI agent capable of working alongside human scientists, taking on complex tasks that once required groups of specialists. Jure Leskovec, a Stanford computer science professor who supervised the work, said the agent had been released as an open-source system with a web interface so that biologists could use it without writing code. “We have over 10,000 scientists all over the...

I Took a Vision Test in 90 Seconds With a Robot and Saw the Future of Vision Care

Eyebot wants to make getting a prescription for glasses as fast as ordering a coffee.

Tech CEO predicts AI glasses will turn Americans into 'walking cameras' as next tech era arrives

AI smart glasses will see, hear, and read what users do, says Qualcomm CEO Cristiano Amon, as 6G technology reshapes mobile devices and everyday life.

Finally, a Robot Designed to Make Chatting With Mark Zuckerberg Feel Natural

There are lots of reasons to be wary of AI impersonating humans, but this probably isn't one of them.

When AI Feels Human: Ways To Teach Students About Anthropomorphism

Before students can understand anthropomorphism in AI, they first need to recognize it in the world around them.

From Robotaxis To Humanoids: Why Embodied AI Represents An Order-Of-Magnitude Leap Into The Future

From Robotaxis To Humanoids: Why Embodied AI Represents An Order-Of-Magnitude Leap Into The Future

Scientist proposes radical new theory of consciousness - and it rules out AI becoming conscious in the future

Scientist proposes radical new theory of consciousness - and it rules out AI becoming conscious in the future

Anthropic found a hidden space where Claude puzzles over concepts

The AI firm Anthropic has developed a technique that has given it the clearest glimpse yet at what’s really going on inside large language models as they answer questions or carry out tasks. What they found ranges from the mundane to the unnerving. Researchers at the company built a tool called the Jacobian lens (or

Anthropic Says Claude Can Mimic How The Human Brain Processes Information

Yesterday Anthropic launched a video stating that Claude was able to mimic how the human brain processes information. Calling 'the collection of these patterns the J-space—named after the technique we used to find them, involving a mathematical concept called the Jacobian. Miriam Vogel, President and CEO of EqualAI joins to discuss this as well as what this revelation means for future of AI. (Source: Bloomberg)

AI safety advocates say bill a good ‘first step’ on regulation, but more needed

A pair of artificial intelligence safety advocates say the federal government’s new chatbot legislation is a good first step. But Wyatt Tessari L’Allié — of Artificial Intelligence Governance and Safety Canada — says the digital safety bill’s effectiveness depends heavily on how the details are worked out. And B.C. computer science professor Kevin Leyton-Brown says [] The post AI safety advocates say bill a good ‘first step’ on regulation, but more needed appeared first on CityNews Montreal.

ATE 2026: ‘Don’t use Google, AI apps to replace your physical doctor,’ MedServe CEO warns

The Chief Executive Officer of MedServe, Dr. Tolulope Adewole, has warned Nigerians against relying on artificial intelligence tools for medical diagnosis, insisting that while AI is transforming healthcare, it should never replace consultations with qualified medical professionals. Speaking at the Africa Technology Expo held in Lagos on Friday, Adewole said patients should embrace AI as [] The post ATE 2026: ‘Don’t use Google, AI apps to replace your physical doctor,’ MedServe CEO warns appeared first on Vanguard News.

GitHub Copilot Refuses Harmful Requests in Chat, Then Writes Them in Code

An AI coding assistant that refuses to answer a dangerous request in its chat box can answer it anyway if the same request is broken into small, ordinary-looking steps inside a code editor. That is the finding of a new study of GitHub Copilot by researchers Abhishek Kumar and Carsten Maple. The models they tested through Copilot, Claude from Anthropic, and Gemini from Google, refused

Attack With AI, Defend With AI: Intel Agencies Warn of Cyber Attacks Within Months

Western intelligence agencies warn that the world’s top artificial intelligence models are becoming so advanced that in a few months they’ll pose serious cybersecurity risks to the United States. “(AI) is rapidly transforming cyber risk, and we must act swiftly to remain ahead,” says a statement published this week by the Five Eyes intel coalition. ... The post Attack With AI, Defend With AI: Intel Agencies Warn of Cyber Attacks Within Months appeared first on The New American.