Lifeboat Foundation

Our approach to analyzing and mitigating future risks posed by advanced AI models.

Google DeepMind has consistently pushed the boundaries of AI, developing models that have transformed our understanding of what’s possible. We believe that AI technology on the horizon will provide society with invaluable tools to help tackle critical global challenges, such as climate change, drug discovery, and economic productivity. At the same time, we recognize that as we continue to advance the frontier of AI capabilities, these breakthroughs may eventually come with new risks beyond those posed by present-day models.

Today, we are introducing our Frontier Safety Framework — a set of protocols for proactively identifying future AI capabilities that could cause severe harm and putting in place mechanisms to detect and mitigate them. Our Framework focuses on severe risks resulting from powerful capabilities at the model level, such as exceptional agency or sophisticated cyber capabilities. It is designed to complement our alignment research, which trains models to act in accordance with human values and societal goals, and Google’s existing suite of AI responsibility and safety practices.

Many more places with Life are likely to exist outside the so-called Habitable Zone (HZ) and they are getting more and more… New post on BigThink, direct link to it at.

Posted on Big Think.

OpenAI’s Superalignment team was promised 20% of the company’s compute resources, according to a person from that team.

A source reveals that OpenAI’s Superalignment team, which was created to develop was to control ‘superintelligent’ AI, wasn’t set up for success.

In organisms, fluid is what binds the organs, the blood vessels and the musculoskeletal system as a whole. For example, hemolymph, a blood-like fluid in a spider’s body, enables muscle activation and exoskeleton flexibility. It was the cucumber spider inhabiting Estonia that inspired scientists to create a complex soft robot, where soft and rigid parts are made to work together and are connected by a liquid.

According to Indrek Must, Associate Professor of Soft Robotics, the designed soft robot is based on real reason. “Broadly speaking, our goal is to build systems from both natural and artificial materials that are as effective as in wildlife. The robotic leg could touch delicate objects and move in the same complex environments as a living spider,” he explains.

In a research paper published in the journal Advanced Functional Materials, the researchers show how a robotic foot touches a primrose stamen, spider web, and pollen grain. This demonstrates the soft robot’s ability to interact with very small and delicate structures without damaging them.

Intel’s upcoming CPUs for thin-and-light laptops will be its first with on-package memory.

A team of scientists led by the Department of Energy’s Oak Ridge National Laboratory has found an unconventional way to improve catalysts made of more than one material. The solution demonstrates a path to designing catalysts with greater activity, selectivity and stability.

Although the Tesla Cybertruck has only been available for about six months, it is already making a major splash in the electric pickup scene.

According to data from S&P Global Mobility, Tesla’s first all-electric pickup had more registrations in March than one of its most notable foes, the Rivian R1T.

March was the fourth month the Tesla Cybertruck was available for purchase. Production is still ramping, and after Tesla built 1,000 units in a single week during Q1, it is evident the company is starting to get the hang of things.

But out of everything Meta announced, one particular demo blew my mind. Meta AI comes with its own image-generation tool called Imagine, which is available in beta to some WhatsApp and web users. The new Meta AI feature can do something OpenAI’s ChatGPT can’t: It creates images instantly with no waiting necessary.

This is the second time an AI product has blown my mind this week. Earlier, I showed you Microsoft’s VASA-1 tool, which generates talking video clips out of a portrait image and a voice recording. VASA-1 isn’t made for the public though, and we might never get access to this particular AI. Anyone could create misleading fakes with it, so Microsoft is only showing off a proof of concept.

Sign up for the most interesting tech & entertainment news out there.

The tension between quantum mechanics and relativity has long been a central split in modern-day physics. Developing a theory of quantum gravity remains one of the great outstanding challenges of the discipline. And yet, no one has yet been able to do it. But as we collect more data, it shines more light on the potential solution, even if some of that data happens to show negative results.

That happened recently with a review of data collected at IceCube, a neutrino detector located in the Antarctic ice sheet, and compiled by researchers at the University of Texas at Arlington. They looked for signs that gravity could vary even a minuscule amount based on quantum mechanical fluctuations. And, to put it bluntly, they didn’t find any evidence of that happening.

Continue reading “Scientists Test for Quantum Gravity” »