Lifeboat Foundation

Cells contain various specialized structures—such as the nucleus, mitochondria or peroxisomes—known as “organelles.” Tracing their genesis and determining their structure is fundamental to understanding cell function and the pathologies linked to their dysfunction.

Carnegie Mellon University (CMU) researchers have developed H₂O – Human2HumanOid – a reinforcement learning-based framework that allows a full-sized humanoid robot to be teleoperated by a human in real-time using only an RGB camera. Which begs the question: will manual labor soon be performed remotely?

A teleoperated humanoid robot allows for the performance of complex tasks that are – at least at this stage – too complex for a robot to perform independently. But achieving whole-body control of human-sized humanoids to replicate our movements in real-time is a challenging task. That’s where reinforcement learning (RL) comes in.

Continue reading “How blue-collar workers will train the humanoids that take their jobs” »

Egineers develop a new hybrid robot that hops, flies, adjusts jump heights, and executes tight turns with high frequency and agility.

Humanoid robots leaped from a curiosity to the next big thing after Nvidia Chief Executive Jensen Huang touted the emerging technology.

On April 9, 1959, reporters and news media crammed into the ballroom of the Dolley Madison House in Washington—the location of NASA Headquarters at that time—to learn the names of the first American astronauts who came to be known as the Mercury 7. Public Information Director Walter Bonney kicked off the announcement by pointing to the seven men sitting on stage. “These are our astronaut volunteers,” he announced. “Take your pictures as you will, gentlemen.” One of those men on the dais, Deke Slayton, a test pilot from Edwards Air Force Base, recalled the pandemonium he witnessed. “I’ve never seen anything like it, before or since.” He described the event as, “a frenzy of light bulbs and questions…it was some kind of roar.” His colleague, Wally Schirra, a test pilot from Naval Air Station Patuxent River, called the media’s interest scary because he soon came to realize that their, “private lives were in jeopardy.”

Sanctuary AI announced that it will be delivering its humanoid robot to a Magna manufacturing facility. Based in Canada, with auto manufacturing facilities in Austria, Magna manufactures and assembles cars for a number of Europe’s top automakers, including Mercedes, Jaguar and BMW. As is often the nature of these deals, the parties have not disclosed how many of Sanctuary AI’s robots will be deployed.

The news follows similar deals announced by Figure and Apptronik, which are piloting their own humanoid systems with BMW and Mercedes, respectively. Agility also announced a deal with Ford at CES in January 2020, though that agreement found the American carmaker exploring the use of Digit units for last-mile deliveries. Agility has since put that functionality on the back burner, focusing on warehouse deployments through partners like Amazon.

For its part, Magna invested in Sanctuary AI back in 2021 — right around the time Elon Musk announced plans to build a humanoid robot to work in Tesla factories. The company would later dub the system “Optimus.” Vancouver-based Sanctuary unveiled its own system, Phoenix, back in May of last year. The system stands 5’7” (a pretty standard height for these machines) and weighs 155 pounds.

SpaceX launched a military weather satellite designed to replace aging satellites from a program dating back to the 1960s. The United States Space Force-62 (USSF-62) mission featured the launch of the first Weather System Follow-on Microwave (WSF-M) spacecraft.

Groundbreaking observations by the James Webb Space Telescope of an early galaxy merger indicate faster and more efficient star formation than previously understood, revealing complex stellar populations and challenging current cosmological theories. Galaxies and stars developed faster after t.

Language models often need more exposure to fruitful mistakes during training, hindering their ability to anticipate consequences beyond the next token. LMs must improve their capacity for complex decision-making, planning, and reasoning. Transformer-based models struggle with planning due to error snowballing and difficulty in lookahead tasks. While some efforts have integrated symbolic search algorithms to address these issues, they merely supplement language models during inference. Yet, enabling language models to search for training could facilitate self-improvement, fostering more adaptable strategies to tackle challenges like error compounding and look-ahead tasks.

Researchers from Stanford University, MIT, and Harvey Mudd have devised a method to teach language models how to search and backtrack by representing the search process as a serialized string, Stream of Search (SoS). They proposed a unified language for search, demonstrated through the game of Countdown. Pretraining a transformer-based language model on streams of search increased accuracy by 25%, while further finetuning with policy improvement methods led to solving 36% of previously unsolved problems. This showcases that language models can learn to solve problems via search, self-improve, and discover new strategies autonomously.

Recent studies integrate language models into search and planning systems, employing them to generate and assess potential actions or states. These methods utilize symbolic search algorithms like BFS or DFS for exploration strategy. However, LMs primarily serve for inference, needing improved reasoning ability. Conversely, in-context demonstrations illustrate search procedures using language, enabling the LM to conduct tree searches accordingly. Yet, these methods are limited by the demonstrated procedures. Process supervision involves training an external verifier model to provide detailed feedback for LM training, outperforming outcome supervision but requiring extensive labeled data.