- all
- popular
- trending
- most recent
Will AI Start Forgetting What It Knows? The Risks of Recursion in Model Training
Researchers warn of "model collapse": a slow AI degradation from over-relying on generated data. Can real data save generative models from losing their human-like edge?
DeepMind makes AlphaFold3 open source
Google DeepMind has made AlphaFold3, an AI model for predicting protein structures, available as open-source for non-commercial use.
Self-Predictive AI: Reshaping Reinforcement Learning through Self-AIXI
Can AI predict itself? Self-AIXI transforms reinforcement learning, replacing exhaustive planning with foresight, promising efficiency and adaptability for AI-driven decision-making.
AI gets a keen sense of touch
Researchers have found a way to teach AI how to feel surfaces, by leveraging surface quantum interactions with laser light.
Science and technology under Trump: what can we expect?
Donald Trump will be the next President of the USA. What can we expect from his administration for science and technology?
AI finds cross-disciplinary connections and suggests research ideas
AI finds connections between different scientific topics, offers insights, and suggests promising research ideas for further study.
DeepSeek R1 reportedly thinks and reasons like OpenAI o1
A new Chinese AI considers each question carefully, trying to avoid mistakes by essentially fact-checking itself.
Faster hurricane modeling with machine learning
A new machine learning method uses equations from atmospheric physics to predict wind patterns inside hurricanes.
Generative AI for genomes
A new generative AI can predict how changes in DNA might affect cells and even design new DNA sequences to change how cells work.
Researchers argue that LLMs don't understand the world
In a new study researchers argue that, despite their impressive performance at certain tasks, LLMs don't really understand the world.