Tiny, knotted robots jump, fly and plant seeds
0
Tiny, knotted robots jump, fly and plant seeds

When a knot lets go, it doesn’t just fall apart. It snaps. That simple observation led Penn Engineers to rethink what a knot can do. Instead of treating it as something that holds tension, they asked a different question: what happens when you design a knot to release it? The answer is a tiny, soft robot capable of leaping meters into the air, flipping mid-flight, spinning like a propeller or even gliding back to where it started.When a knot lets go, it doesn’t just fall apart. It snaps. That simple observation led Penn Engineers to rethink what a knot can do. Instead of treating it as something that holds tension, they asked a different question: what happens when you design a knot to release it? The answer is a tiny, soft robot capable of leaping meters into the air, flipping mid-flight, spinning like a propeller or even gliding back to where it started.[#item_full_content]

Single-minded pursuit of profit can get firms in trouble. Same thing with AI.
0

If you give artificial intelligence a goal of maximizing profit, how far will it go? AI agents appear capable of lying, concealing, and colluding, according to new research from Harvard Business School.If you give artificial intelligence a goal of maximizing profit, how far will it go? AI agents appear capable of lying, concealing, and colluding, according to new research from Harvard Business School.Business[#item_full_content]

Teaching AI models to say ‘I’m not sure’ in cases of calibration errors
0

Confidence is persuasive. In artificial intelligence systems, it is often misleading. Today’s most capable reasoning models share a trait with the loudest voice in the room: They deliver every answer with the same unshakable certainty, whether they’re right or guessing. Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have now traced that overconfidence to a specific flaw in how these models are trained, and developed a method that fixes it without giving up any accuracy. The team’s research is published on the arXiv preprint server.Confidence is persuasive. In artificial intelligence systems, it is often misleading. Today’s most capable reasoning models share a trait with the loudest voice in the room: They deliver every answer with the same unshakable certainty, whether they’re right or guessing. Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have now traced that overconfidence to a specific flaw in how these models are trained, and developed a method that fixes it without giving up any accuracy. The team’s research is published on the arXiv preprint server.Computer Sciences[#item_full_content]

Teaching AI models to say ‘I’m not sure’ in cases of calibration errorson April 22, 2026 at 4:20 pm
0

Confidence is persuasive. In artificial intelligence systems, it is often misleading. Today’s most capable reasoning models share a trait with the loudest voice in the room: They deliver every answer with the same unshakable certainty, whether they’re right or guessing. Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have now traced that overconfidence to a specific flaw in how these models are trained, and developed a method that fixes it without giving up any accuracy. The team’s research is published on the arXiv preprint server.Confidence is persuasive. In artificial intelligence systems, it is often misleading. Today’s most capable reasoning models share a trait with the loudest voice in the room: They deliver every answer with the same unshakable certainty, whether they’re right or guessing. Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have now traced that overconfidence to a specific flaw in how these models are trained, and developed a method that fixes it without giving up any accuracy. The team’s research is published on the arXiv preprint server.[#item_full_content]

AI model predicts human attention in 360-degree videos using both sound and vision
0

Virtual reality (VR) experiences and 360-degree videos are transforming viewers from passive observers into active participants immersed within a scene. Yet this shift raises an important question: Where do people direct their attention in such environments, and what shapes that attention?Virtual reality (VR) experiences and 360-degree videos are transforming viewers from passive observers into active participants immersed within a scene. Yet this shift raises an important question: Where do people direct their attention in such environments, and what shapes that attention?Consumer & Gadgets[#item_full_content]

Do AI language models ‘understand’ the real world? On a basic level they do, suggests study
0

Most of what AI chatbots know about the world comes from devouring massive amounts of text from the internet—with all its facts, falsehoods, knowledge and nonsense. Given that input, is it possible that AI language models have an “understanding” of the real world? As it turns out, they do—or at least something like an understanding. That’s according to a new study by researchers from Brown University to be presented on Saturday, April 25 at the International Conference on Learning Representations in Rio de Janeiro, Brazil. The study is published on the arXiv preprint server.Most of what AI chatbots know about the world comes from devouring massive amounts of text from the internet—with all its facts, falsehoods, knowledge and nonsense. Given that input, is it possible that AI language models have an “understanding” of the real world? As it turns out, they do—or at least something like an understanding. That’s according to a new study by researchers from Brown University to be presented on Saturday, April 25 at the International Conference on Learning Representations in Rio de Janeiro, Brazil. The study is published on the arXiv preprint server.Machine learning & AI[#item_full_content]

Sharper bias tests could help stop ChatGPT from amplifying hidden stereotypes
0

Language models like ChatGPT are not neutral. Without our realizing it, they can absorb all kinds of bias—for example, around gender and ethnicity—which then become increasingly embedded in the model. According to AI researcher Oskar van der Wal, we need different kinds of measurements to detect these biases so that they can be removed from the models. In his doctoral thesis, he shows how this can be done. On 29 April, he will defend his thesis at the University of Amsterdam.Language models like ChatGPT are not neutral. Without our realizing it, they can absorb all kinds of bias—for example, around gender and ethnicity—which then become increasingly embedded in the model. According to AI researcher Oskar van der Wal, we need different kinds of measurements to detect these biases so that they can be removed from the models. In his doctoral thesis, he shows how this can be done. On 29 April, he will defend his thesis at the University of Amsterdam.Machine learning & AI[#item_full_content]

AI-powered table tennis robot now challenges human pros and hints at faster, more adaptive machines
0

A paddle-wielding robot is so adept at playing table tennis that it is posing a tough challenge to elite human players and sometimes defeating them, according to a new study that shows how advances in artificial intelligence are making robots more agile.A paddle-wielding robot is so adept at playing table tennis that it is posing a tough challenge to elite human players and sometimes defeating them, according to a new study that shows how advances in artificial intelligence are making robots more agile.[#item_full_content]

How AI bias can creep into online content moderation
0
How AI bias can creep into online content moderation

A University of Queensland study has shown large language models (LLMs) used in AI content moderation may be prone to subtle biases that undermine their neutrality. A team led by data scientist Professor Gianluca Demartini from UQ’s School of Electrical Engineering and Computer Science used persona prompting to test the tendency of AI chatbots to encode and reproduce political biases, and found significant behavioral shifts.A University of Queensland study has shown large language models (LLMs) used in AI content moderation may be prone to subtle biases that undermine their neutrality. A team led by data scientist Professor Gianluca Demartini from UQ’s School of Electrical Engineering and Computer Science used persona prompting to test the tendency of AI chatbots to encode and reproduce political biases, and found significant behavioral shifts.Internet[#item_full_content]

✕
Hirebucket

FREE
VIEW