Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI

The human brain, just like whatever you’re reading this on, uses electricity to function. Neurons are constantly sending and receiving electrical signals. Everyone’s brain works a bit differently, and scientists are now getting closer to establishing how electrical activity is functioning in individual patients’ brains and how to stimulate it to treat neurological and psychiatric disorders. Some scientists are even using advanced AI predictive technology to enhance their brain stimulation therapy methods.

Read more | >

A $75000 robot dog seen leaving a Manhattan housing project has outraged city residents who heard less than a month ago that the pricey Boston Dynamics bots would be banned from being used as weapons.

The dog was seen exiting the housing project on Monday with a group of human police officers, who acted almost subservient to the shiny blue metal creature while escorting it out of the Kips Bay building. While the New York Police Department insisted the faux-canine had not played an “active role” in the operation, which allegedly involved removing an armed man from an apartment belonging to a woman and her baby, its appearance seemed to cast a spell over the assembled residents.

One resident, tenant organization head Melanie Aucello, rushed back home after receiving a handful of calls about “police in the building” and was wildly disturbed by what she saw there – not just the dog, but its effects on people.

Read more | >

Words categorize the semantic fields they refer to in ways that maximize communication accuracy while minimizing complexity. Recent studies have shown that human languages are optimally balanced between accuracy and complexity. For example, many languages have a word that denotes the color red, but no language has individual words to distinguish ten different shades of the color. These additional words would complicate the vocabulary and rarely would they be useful to achieve precise communication.

A study published on 23 March in the journal Proceedings of the National Academy of Sciences analyzed how develop spontaneous systems to name colors. A study by Marco Baroni, ICREA research professor at the UPF Department of Translation and Language Sciences (DTCL), conducted with members of Facebook AI Research (France).

For this study, the researchers formed two artificial neural networks trained with two generic deep learning methods. As Baroni explains: “we made the networks play a color-naming game in which they had to communicate about color chips from a continuous color space. We did not limit the “language” they could use, however, when they learned to play the game successfully, we observed the color-naming terms these artificial neural networks had developed spontaneously.”

Read more | >

The Lightmatter photonic computer is 10 times faster than the fastest NVIDIA artificial intelligence GPU while using far less energy. And it has a runway for boosting that massive advantage by a factor of 100, according to CEO Nicholas Harris.

In the process, it may just restart a moribund Moore’s Law.

Or completely blow it up.

Read more | >

Spot was apparently being used for reconnaissance.

Pictures of the exercises were shared on Twitter by France’s foremost military school, the École Spéciale Militaire de Saint-Cyr. It described the tests as “raising students’ awareness of the challenges of tomorrow,” which include the “robotization of the battlefield.”

A report by French newspaper Ouest-France offers more detail, saying that Spot was one of a number of robots being tested by students from France’s École Militaire Interarmes (Combined Arms School), with the intention of assessing the usefulness of robots on future battlefields.

Boston Dynamics’ vice president of business development Michael Perry told The Verge that the robot had been supplied by a European distributor, Shark Robotics, and that the US firm had not been notified about this particular use. “We’re learning about it as you are,” says Perry. “We’re not clear on the exact scope of this engagement.” The company says it was aware that its robots were being used with the French government, including the military.

Read more | >

Halloween is forever if you’re a Japanese black bear.

The Japanese city of Takikawa is using robot wolves to prevent bear attacks. Bear encounters have been on the rise as cities have grown and acorns — a key part of black bears’ pre-hibernation diet — have become harder to find. Since these robots were installed in Takikawa, there have been no attacks in the city’s surrounding area.

Read more | >

The tendency of many cellular proteins to form protein-rich biomolecular condensates underlies the formation of subcellular compartments and has been linked to various physiological functions. Understanding the molecular basis of this fundamental process and predicting protein phase behavior have therefore become important objectives. To develop a global understanding of how protein sequence determines its phase behavior, we constructed bespoke datasets of proteins of varying phase separation propensity and identified explicit biophysical and sequence-specific features common to phase-separating proteins. Moreover, by combining this insight with neural network-based sequence embeddings, we trained machine-learning classifiers that identified phase-separating sequences with high accuracy, including from independent external test data.

Intracellular phase separation of proteins into biomolecular condensates is increasingly recognized as a process with a key role in cellular compartmentalization and regulation. Different hypotheses about the parameters that determine the tendency of proteins to form condensates have been proposed, with some of them probed experimentally through the use of constructs generated by sequence alterations. To broaden the scope of these observations, we established an in silico strategy for understanding on a global level the associations between protein sequence and phase behavior and further constructed machine-learning models for predicting protein liquid–liquid phase separation (LLPS). Our analysis highlighted that LLPS-prone proteins are more disordered, less hydrophobic, and of lower Shannon entropy than sequences in the Protein Data Bank or the Swiss-Prot database and that they show a fine balance in their relative content of polar and hydrophobic residues.

Read more | >

They weren’t scheduled to return to Earth until April 28th at the earliest, so why did NASA astronauts Michael Hopkins, Victor Glover, and Shannon Walker, along with Japan Aerospace Exploration Agency (JAXA) astronaut Soichi Noguchi, suit up and climb aboard the Crew Dragon Resilience on April 5th? Because a previously untested maneuver meant that after they closed the hatch between their spacecraft and the International Space Station, there was a chance they weren’t going to be coming back.

On paper, moving a capsule between docking ports seems simple enough. All Resilience had to do was undock from the International Docking Adapter 2 (IDA-2) located on the front of the Harmony module, itself attached to the Pressurized Mating Adapter 2 (PMA-2) that was once the orbital parking spot for the Space Shuttle, and move over to the PMA-3/IDA-3 on top of Harmony. It was a short trip through open space, and when the crew exited their craft and reentered the Station at the end of it, they’d only be a few meters from where they started out approximately 45 minutes prior.

The maneuver was designed to be performed autonomously, so technically the crew didn’t need to be on Resilience when it switched docking ports. But allowing the astronauts to stay aboard the station while their only ride home undocked and flew away without them was a risk NASA wasn’t willing to take.

Read more | >

Those that have grown up with open source in the past 20 years know that open source is popular. It’s popular because of a number of reasons including that it fosters innovation, speeds up delivery, and helps us all collectively learn from each other.

We ourselves at the AGI Lab have just assumed this was a good thing. We believe that Open Source research helps everyone. Many research groups in AGI research are already open sourcing including Open Cog, Open Nars, and more.

From an ethical standpoint, we use a system called SSIVA Theory to teach ethics to systems we work on such as Uplift and so we assumed we should release some of our code (which we have here on this blog and in papers) and we planned on open sourcing a version of the mASI or collective system that we work on that uses an AGI Cognitive Architecture.

Read more | >