This week - Spot explores Chernobyl; Yann LeCun on GPT-3; Apple learns the limits of robotics in manufacturing; and more!
More Than A Human
A newly developed artificial neural connection allowed experimental animals to regain control of a paralyzed hand in a matter of minutes. This result may find its way to help people recovering from strokes to control their hands again.
Yann LeCun, one of the pioneers of deep learning, shares his thoughts on GPT-3 and he is not impressed. "It's entertaining, and perhaps mildly useful as a creative help. But trying to build intelligent machines by scaling up language models is like building a high-altitude airplanes to go to the moon", LeCun writes.
Researchers from MIT applied machine learning to historical linguistic and created a system that can automatically decipher a lost language that’s no longer understood — without knowing its relationship to other languages. The AI was created in response to the rapid disappearance of human languages. Most of the languages that have existed are no longer spoken, and at least half of those remaining are predicted to vanish in the next 100 years. The new system could help recover them. More importantly, it could preserve our understanding of the cultures and wisdom of their speakers.
We can add now screenwriting to the list of things GPT-3 was used to generate.
"In the digital arms race with China, the only thing worse than fearing AI itself is the fear of not having it at all", reads the headline of this article and then it goes arguing that the US has to embrace AI in the military or it will be left behind. The ethics of military AI were only mentioned in one sentence at the end of the article.
Spot went where no robot-dog has been before - into Exclusion Zone territory of the Chernobyl Nuclear Power Plant. The Spot robot, being remotely controlled, walked around the area and inside the New Safe Confinement structure, a steel dome designed to keep hazardous radioactivity inside of it. Spot also had help from drones, ground-based robotic vehicles and sensors to assess the distribution of radioactive materials.
Robotiz3d is a new company developing autonomous robots to repair potholes. "The proposed system will be able to autonomously detect and characterize road defects such as cracks and potholes, assess and predict the severity of such defects and fix cracks so that they do not evolve into potholes," said Robotiz3d's CTO.
Just like Tesla a while back, Apple tried to automate as much on their manufacturing process as possible to learn there are limits to current robotics and sometimes humans are better.
Here a very high-level description of how to make a synthetic cell - either by stripping or replacing genome of an existing cell (the top-down approach) or making the cell from scratch (bottom-up approach).
This article lays out 10 bold predictions for human genomics by 2030 - from DNA sequencing becoming a routine and being securely stored on people's phones to therapies targeting genetic diseases. The list also features such predictions as gene sequencing being regularly featured at school science fairs and moving beyond population descriptors based on historic social constructs such as race.
Just outside Shenzhen, Chinese tech-giant Tencent plans to build an entire mini-city for 80,000 people. Named Netcity, the proposed city will be an extension to Shenzhen and will feature everything that modern technology has to offer to make the most futuristic city in the world. The construction should start this year and take seven years to complete.