Large language models (LLMs) sometimes lose confidence when answering questions and abandon correct answers, according to a new study by researchers at Google DeepMind and University College London.
Category: robotics/AI – Page 42
Elephant robot demonstrates bioinspired 3D printing technology
A cheetah’s powerful sprint, a snake’s lithe slither, or a human’s deft grasp: Each is made possible by the seamless interplay between soft and rigid tissues. Muscles, tendons, ligaments, and bones work together to provide the energy, precision, and range of motion needed to perform the complex movements seen throughout the animal kingdom.
“Not a Single Mistake”: World’s First Autonomous Surgical Robot Completes Complex Procedure With 100% Accuracy and Zero Human Intervention
IN A NUTSHELL 🤖 The SRT-H robot, developed at Johns Hopkins University, performed a successful gallbladder surgery with human-like precision. 📚 Trained on surgical videos, the robot can learn and adapt in real-time, enhancing its ability to handle complex procedures. 🛠️ Equipped with machine learning technology similar to ChatGPT, SRT-H responds to voice commands and
Quantum on the Cusp—What to Prepare for in the Emerging Ecosystem of Quantum Tech and Computing
Dear Friends & Colleagues, Like artificial intelligence, quantum technologies will transform our world as we know it. This issue focuses on what quantum constitutes and the promises and challenges of this emerging technology.
Kira (Short Film on Human Cloning)
My new AI-assisted short film is here. Kira explores human cloning and the search for identity in today’s world.
It took nearly 600 prompts, 12 days (during my free time), and a $500 budget to bring this project to life. The entire film was created by one person using a range of AI tools, all listed at the end.
Enjoy.
~ Hashem.
Instagram: / hashem.alghaili.
Facebook: / sciencenaturepage.
Other channels: https://muse.io/hashemalghaili
British-built Hawk-Eye software goes dark during Wimbledon match
Wimbledon’s new automated line-calling system glitched during a tennis match Sunday, just days after it replaced the tournament’s human line judges for the first time.
The system, called Hawk-Eye, uses a network of cameras equipped with computer vision to track tennis balls in real-time. If the ball lands out, a pre-recorded voice loudly says, “Out.” If the ball is in, there’s no call and play continues.
However, the software temporarily went dark during a women’s singles match between Brit Sonay Kartal and Russian Anastasia Pavlyuchenkova on Centre Court.
Lab-grown Human Brains Stacked into Towers for Biocomputers
Further reading.
https://www.sciencedirect.com/science/article/abs/pii/S0956566325005779
https://journals.lww.com/medmat/fulltext/2025/06000/organoid…res.1.aspx.
https://link.springer.com/article/10.1007/s10015-024-00969-0
https://www.igi-global.com/chapter/synthetic-biology-and-ai/366024
#sciencenews #science #news #scienceexplained #artificialintelligence #ai #robots #explained #research #organoid
“Engineering Agential Materials” by Michael Levin
This is a ~18 min talk plus ~10 min Q&A on a top-down approach to bioengineering and robotics that I gave at the Biohybrid Robotics Symposium in Switzerland in July 2025 (https://biohybrid-robotics.com/).
Stephen Wolfram on AI, human-like minds & formal knowledge
Want more on knowledge hypergraphs? Sign up for my newsletter at Open Web Mind https://www.openwebmind.com/newsletter/ or subscribe to my YouTube channel htt…
Korean researchers develop energy-efficient NPU technology that cuts AI cloud power consumption by 44%
Researchers at the Korea Advanced Institute of Science and Technology (KAIST) have developed energy-efficient NPU technology that demonstrates substantial performance improvements in laboratory testing.
Their specialised AI chip ran AI models 60% faster while using 44% less electricity than the graphics cards currently powering most AI systems, based on results from controlled experiments.
KAIST researchers develop energy-efficient NPU technology that runs AI models 60% faster while using 44% less power than current GPU systems.