Note: This is a joint distillation of both Iterated Distillation and Amplification by Ajeya Cotra (summarizing Paul Christiano) and A Generalist Agent by DeepMind.
Category: robotics/AI – Page 1,479
An autonomously oscillating supramolecular self-replicator
Oscillations are widespread throughout the natural world and a number of fascinating inorganic oscillating reactions are known—but the formation and control of oscillating, self-replicating synthetic systems has remained challenging. Now, it has been shown that chemically fuelled oscillations within a network of organic replicators can drive supramolecular assembly and disassembly.
Computable Artificial General Intelligence
If you are interested in artificial general intelligence (AGI), then I have a panel discussion to recommend. My friend, David Wood, has done a masterful job of selecting three panelists with deep insight into possible regulation of AGI. One of the panelists was my friend, Dan Faggella, who was eloquent and informative as usual. For this session of the London Futurists, David Wood selected two other panelists with significantly different opinions on how to properly restrain AGI.
An artificial general intelligence (AGI), by one definition, is an agent that requires less information than any other to make an accurate prediction. It is arguable that the general reinforcement learning agent AIXI not only met this definition, but was the only mathematical formalism to do so. Though a significant result, AIXI was incomputable and its performance subjective. This paper proposes an alternative formalism of AGI which overcomes both problems. Formal proof of its performance is given, along with a simple implementation and experimental results that support these claims.
Autonomy: the missing AGI ingredient?
Here are some things I would expect any AGI to be able to do:…
Epistemic status: trying to feel out the shape of a concept and give it an appropriate name. Trying to make explicit some things that I think exist implicitly in many people’s minds. This post makes truth claims, but its main goal is to not to convince you that they are true.
The big new idea for making self-driving cars that can go anywhere
The mainstream approach to driverless cars is slow and difficult. These startups think going all-in on AI will get there faster.
Sick of picking up toys? Dyson’s future home robots want to do it for you
Company expects its robots to be doing your household chores in a decade.
Researchers create digital humans that learn complex movements
Researchers at Meta’s Artificial Intelligence Research Lab (Facebook) in the U.S. and at the University of Twente’s Neuromechanical Modelling and Engineering Lab in the Netherlands (led by Prof.dr.ir Massimo Sartori), have co-developed the open-source framework MyoSuite, which combines advanced musculoskeletal models with advanced artificial intelligence (AI). The AI-powered digital models in MyoSuite can learn to execute complex movements and interactions with assistive robots, that would otherwise require long experimentations on real human subjects.
Modeling and simulation are now as important to human health technologies as they have been for the advancement of modern automotive industry. Prof. Massimo Sartori: “If we could predict the outcome of a robotic therapy beforehand, then we could optimize it for a patient and deliver a truly personalized and cost-effective treatment.”
MyoSuite supports the co-simulation of AI-powered musculoskeletal systems physically interacting with assistive robots such as exoskeletons. With MyoSuite you can simulate biological phenomena, e.g., muscle fatigue, muscle sarcopenia, tendon tear and tendon reaffirmation. Moreover, you can simulate how assistive robots could be designed and controlled to restore movement following impairment.
OpenAI punished dev who used GPT-3 to ‘resurrect’ the dead — was this fair?
Predicting it now. 2030s there will be Tons of this, and not just chat bots of dead people, but making them seem alive, 24/7 in VR world meta whatever. There will probably be shops that cater to this and try and make it as close and realistic as possible, will probably mostly be underground.
The recent case of a man making a chatbot based on his deceased fiancée raises ethical questions: Is this something we want? property= description.
A helping hand for robotic manipulator design
Robotic copy of a human hand still needs a lot of ummm work, and funding.
MIT researchers have created an interactive design pipeline that streamlines and simplifies the process of crafting a customized robotic hand with tactile sensors.
Typically, a robotics expert may spend months manually designing a custom manipulator, largely through trial-and-error. Each iteration could require new parts that must be designed and tested from scratch. By contrast, this new pipeline doesn’t require any manual assembly or specialized knowledge.
Akin to building with digital LEGOs, a designer uses the interface to construct a robotic manipulator from a set of modular components that are guaranteed to be manufacturable. The user can adjust the palm and fingers of the robotic hand, tailoring it to a specific task, and then easily integrate tactile sensors into the final design.