Homechevron_rightArchiveschevron_rightScience and Research

Archives

Explore Stories by Date

19 Jun 2026

AI discovers hidden antibiotic candidates inside prion proteins linked to brain diseases

Researchers at the University of Pennsylvania used AI to uncover a surprising source of antibiotics: prion proteins known for causing fatal brain diseases. The deep-learning platform identified over 1,000 antimicrobial peptides called prionins, with dozens showing strong activity against drug-resistant bacteria in lab and mouse tests.

2 Sources

Korean researchers use light color to teach artificial synapse how to remember and forget

Researchers at Sungkyunkwan University have developed an artificial synapse that uses light color to control memory—near-infrared light strengthens connections while blue light weakens them. By turning a semiconductor defect into a feature, the team achieved brain-like balanced learning that remained stable over 1,000 training rounds, compared to conventional systems that failed after 200.

2 Sources

AI helps Boston Children's Hospital diagnose 18 kids whose rare diseases stumped doctors for years

Boston Children's Hospital partnered with OpenAI to analyze 376 unsolved pediatric cases, leading to 18 new rare disease diagnoses. The o3 model synthesized genetic data and medical literature to identify conditions including neurodevelopmental disorders and neuromuscular diseases. OpenAI has committed $50 million to support the hospital's AI initiatives, which have already saved 60,000 hours of work valued at over $7 million.

2 Sources

18 Jun 2026

AI deskilling threatens professional expertise as doctors and engineers lose skills to automation

New research reveals artificial intelligence tools are causing skill erosion among healthcare workers and software developers. Studies show physicians' adenoma detection rates dropped from 28.4% to 22.4% after AI adoption, while software engineers using AI assistants scored 17% lower on learning tests. The findings highlight growing concerns about over-reliance on AI tools and the need for deliberate skill retention strategies.

2 Sources

MIT researchers built DAAAM, a robot memory system that remembers where you left your keys

MIT researchers developed DAAAM, a robot memory system that combines computer vision and 3D mapping to give robots long-term spatial memory. The AI-powered system for robots lets them remember object locations over time, answering questions like 'where did I leave my wallet?' The breakthrough addresses a fundamental gap in making physical AI systems useful in real-world environments.

2 Sources

Deep learning identifies promising antibiotic candidates against drug-resistant gonorrhea

Researchers at Harvard's Wyss Institute used deep learning to screen 6 million compounds and identified two promising candidates against multi-drug resistant Neisseria gonorrhoeae. The lead compound, called A1, targets alanine racemase, an enzyme critical for bacterial cell wall formation. The study marks the first use of Organ Chip technology for antibiotic preclinical testing, offering a faster validation pipeline.

2 Sources

17 Jun 2026

Nvidia ENPIRE lets AI coding agents train robots to install GPUs and master precision tasks

Nvidia's GEAR lab has developed ENPIRE, an agent harness framework that enables AI coding agents to autonomously direct robot training without human supervision. The system achieved 99% success rates on tasks like GPU installation and pin insertion, with eight-robot teams completing training in just two hours.

3 Sources

AI analyzes retinal photos to predict Alzheimer's disease risk decades before symptoms appear

University of Florida researchers trained machine learning models on retinal photographs from over 40,000 patients to identify early Alzheimer's disease risk factors. The AI-powered analysis detected biological and lifestyle indicators like blood pressure, smoking, and insomnia by examining the retina's arteries and optic nerve, offering a low-cost alternative to expensive brain scans.

2 Sources

16 Jun 2026

Brain-Computer Interface and AI Let Speechless ALS Patient Speak and Work Full-Time

A UC Davis brain-computer interface has enabled Casey Harrell, a paralyzed ALS patient, to communicate independently for over 3,800 hours across two years with 99% accuracy in controlled tests. Using AI-driven speech synthesis, the BCI implant translates neural activity into speech, allowing him to work full-time as an environmental advocate and have conversations with his daughter who's never heard his voice.

2 Sources

Just 13 words on Reddit can manipulate AI search results, Cornell researchers reveal

Cornell University researchers discovered that a tiny snippet of just 13 words on user-generated content platforms like Reddit, Wikipedia, or Quora can consistently manipulate AI search tools including ChatGPT and Google Gemini. The study exposes how easily brands and scammers can poison AI outputs through AI-engine optimization, raising urgent questions about information reliability.

2 Sources

KAIST's internal chip cooling tech slashes energy use by 90% for AI data centers

Researchers at KAIST developed a liquid cooling technology that embeds microscopic channels directly inside silicon semiconductor chips, cutting cooling energy to one-tenth of current levels. The system achieved a Coefficient of Performance of 106,000—ten times higher than the previous world record—using ordinary room-temperature water and existing manufacturing processes.

2 Sources

Van der Waals crystal mimics human neurons, advancing AI hardware with light-based learning

Sungkyunkwan University researchers developed an optoelectronic synaptic device that mimics human neuronal and synaptic functions using a designable van der Waals crystal. The device, created through a single-step sulfurization process, learns and stores information with light, achieving 96.24% accuracy on image recognition tasks and showing 34.7% better retention efficiency than conventional materials.

2 Sources

AI Writing Tools Create Fluency Trap Where Polished Output Masks Shallow Thinking, Study Finds

New research published in Computers and Composition reveals that AI writing tools like ChatGPT and Claude create a "fluency trap" where polished, confident-sounding text gives writers false confidence. A two-semester study of 38 undergraduate students found that AI-assisted writing demands more thought from students, not less, requiring trial and error, human judgment, and genuine purpose to be effective.

2 Sources

AI-powered deep brain stimulation adapts in real time to improve gait in Parkinson's disease

Researchers have developed adaptive deep brain stimulation that adjusts electrical pulses in real time based on patient movement. The system uses machine learning to decode brain signals during walking, standing, and obstacle navigation, delivering personalized neuromodulation that improves walking in Parkinson's patients where conventional therapies fall short.

3 Sources

11 Jun 2026

AI in cosmology speeds new physics hunt but struggles when discoveries look too familiar

A groundbreaking study reveals AI could uncover new physics faster using transfer learning, cutting simulation needs by over tenfold. But researchers discovered a critical flaw: when AI learned how the universe works through standard models, it sometimes became too reliant on prior knowledge, missing genuinely new phenomena that resembled familiar patterns—a problem called negative transfer.

2 Sources

Get all updates into your inbox 💌, Stay Up-to-date for Free 🥳

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.

Explore

News Categories

Technology Business Policy Startups Health Science Entertainment

Terms Privacy Content Contact Us