Category Technology/Electronics

Google’s AI Dreamer learns how to self-improve over time by mastering Minecraft

Google's AI Dreamer learns how to self-improve over time by mastering Minecraft
Training process of Dreamer. Credit: Nature (2025). DOI: 10.1038/s41586-025-08744-2

A trio of AI researchers at Google’s Google DeepMind, working with a colleague from the University of Toronto, report that the AI algorithm Dreamer can learn to self-improve by mastering Minecraft in a short amount of time. In their study published in the journal Nature, Danijar Hafner, Jurgis Pasukonis, Timothy Lillicrap and Jimmy Ba programmed the AI app to play Minecraft without being trained and to achieve an expert level in just nine days.

Over the past several years, computer scientists have learned a lot about how deep learning can be used to train AI applications to conduct seemingly intelligent activities such as answering questions...

Read More

Repurposed smartphone camera sensors create real-time, high-resolution imaging of antiproton annihilations

Did you know that the camera sensor in your smartphone could help unlock the secrets of antimatter? The AEgIS collaboration, led by Professor Christoph Hugenschmidt’s team from the research neutron source FRM II at the Technical University of Munich (TUM), has developed a detector using modified mobile camera sensors to image, in real time, the points where antimatter annihilates with matter.

This new device, described in a paper published in Science Advances, can pinpoint antiproton annihilations with a resolution of about 0.6 micrometers, a 35-fold improvement over previous real-time methods.

AEgIS and other experiments at CERN’s Antimatter Factory, such as ALPHA and GBAR, are on a mission to measure the free-fall of antihydrogen within Earth’s gravitational field with high pr...

Read More

AI thinks like us—flaws and all: Study finds ChatGPT mirrors human decision biases in half the tests

AI bias
Credit: AI-generated image

Can we really trust AI to make better decisions than humans? A new study says … not always. Researchers have discovered that OpenAI’s ChatGPT, one of the most advanced and popular AI models, makes the same kinds of decision-making mistakes as humans in some situations—showing biases like overconfidence of hot-hand (gambler’s) fallacy—yet acting inhuman in others (e.g., not suffering from base-rate neglect or sunk cost fallacies).

Published in the Manufacturing & Service Operations Management journal, the study reveals that ChatGPT doesn’t just crunch numbers—it “thinks” in ways eerily similar to humans, including mental shortcuts and blind spots...

Read More

New superconducting state discovered: Cooper-pair density modulation

Superconductivity is a quantum physical state in which a metal is able to conduct electricity perfectly without any resistance. In its most familiar application, it enables powerful magnets in MRI machines to create the magnetic fields that allow doctors to see inside our bodies. Thus far, materials can only achieve superconductivity at extremely low temperatures, near absolute zero (a few tens of Kelvin or colder).

But physicists dream of superconductive materials that might one day operate at room temperature. Such materials could open entirely new possibilities in areas such as quantum computing, the energy sector, and medical technologies.

“Understanding the mechanisms leading to the formation of superconductivity and discovering exotic new superconducting phases is not only...

Read More