From Stanford TRANSIC: Sim-to-Real Policy Transfer by Learning from Online Correction.
From Stanford.
TRANSIC: Sim-to-Real Policy Transfer by Learning from Online Correction.
Learning in simulation and transferring the learned policy to the real world has the potential to enable generalist robots.
An advanced new technique combines machine-learning algorithms with measurements of vibrations for monitoring tool wear.
A viral two-minute short titled “Air Head,” created in part by using OpenAI’s video-generating Sora tool, required a good bit human editing.
New tech aids stroke recovery:
A groundbreaking new robotic hand that can feel touch like a human one and pave the way for new prosthetics and robots.
“This is a really nice way of incorporating something you know about your physical system deep inside your machine-learning scheme. It goes far beyond just performing feature engineering on your data samples or simple inductive biases,” Schäfer says.
This generative classifier can determine what phase the system is in given some parameter, like temperature or pressure. And because the researchers directly approximate the probability distributions underlying measurements from the physical system, the classifier has system knowledge.
This enables their method to perform better than other machine-learning techniques. And because it can work automatically without the need for extensive training, their approach significantly enhances the computational efficiency of identifying phase transitions.
Summary: Researchers use AI to reveal distinct cellular-level differences in the brains of men and women, focusing on white matter. These findings show AI can accurately identify sex-based brain patterns invisible to human eyes.
The study suggests that understanding these differences can enhance diagnostic tools and treatments for brain disorders. This research emphasizes the need for diversity in brain studies to ensure comprehensive insights into neurological diseases.
Micius is considered quantum’s “Sputnik” moment, prompting American policymakers to funnel hundreds of millions of dollars into quantum information science via the National Quantum Initiative. Bills such as the Innovation and Competition Act of 2021 have provided $1.5 billion for communications research, including quantum technology. The Biden Administration’s proposed 2024 budget includes $25 billion for “emerging technologies” including AI and quantum. Ultimately, quantum’s awesome computing power will soon render all existing cryptography obsolete, presenting a security migraine for governments and corporations everywhere.
Quantum’s potential to turbocharge AI also applies to the simmering technology competition between the world’s superpowers. In 2021, the U.S. Commerce Department added eight Chinese quantum computing organizations to its Entity List, claiming they “support the military modernization of the People’s Liberation Army” and adopt American technologies to develop “counter-stealth and counter-submarine applications, and the ability to break encryption.”
These restrictions dovetail with a raft of measures targeting China’s AI ambitions, including last year blocking Nvida from selling AI chips to Chinese firms. The question is whether competition between the world’s top two economies stymies overall progress on AI and quantum—or pushes each nation to accelerate these technologies. The answer could have far-reaching consequences.
Spur uses computer vision to help companies monitor and predict how a wind farm will impact a local bird population.
Oscar Wilde once said that sarcasm was the lowest form of wit, but the highest form of intelligence. Perhaps that is due to how difficult it is to use and understand. Sarcasm is notoriously tricky to convey through text—even in person, it can be easily misinterpreted. The subtle changes in tone that convey sarcasm often confuse computer algorithms as well, limiting virtual assistants and content analysis tools.
