Uber puts self-driving cars on roads in San Francisco and gets into troubles. Google’s self-driving car division gets a new name and becomes a company. Amazon drone delivers the first package in the UK. IEEE will help you build an ethical AI. On AI winter and why we won’t see another one. And more!
More Than A Human
Here’s an infographic from Futurism showing how we can enhance our bodies and minds, and what we could do when we connect our minds directly with the machines.
An interesting talk (27 minutes long) by Dr Martine Rothblatt. She starts with the bold claim that the goal of the technology is to end of the death and then explains why it is true.
IEEE has released a document called Ethically Aligned Design, which includes a series of detailed recommendations based on the input of more than 100 “thought leaders” working in academia, science, government and corporate sectors, in the fields of AI, law and ethics, philosophy and policy. IEEE hopes the document will help technologists build benevolent and beneficial autonomous systems, rather than thinking that ethics is not something they need to be worrying about.
According to this article, many people overhype AI and machine learning. Especially the solutions based deep learning, which work great where there is a lot of data. But if you don’t have enough data or your problem is simpler, it might be better to choose simpler algorithms. The blog post says that the entire AI industry is based on grand hopes and that it eventually bust.
On the other hand, Andrew Ng, chief scientist at Baidu Research and one of the big guys in AI research, claims that we will not see another AI winter because hardware advances will keep AI breakthroughs coming. He admits there is a hype around AI thought, but he believes it will not crash like a couple times before.
Or “How to read the medical AI reports with a critical eye, and see the truth through the hype”. Long, but interesting read.
Garbage in, garbage out. This rule known by everyone who had some contact with computer science is even more important nowadays when it comes to training machine learning systems. Most of the recent AI failures (which were featured in the last issue of H+ Weekly) can be traced to a biased training dataset or how the humans interact and teach the system after the training.
Somewhere in Cambridge, UK, on December 7th, an Amazon drone took off with a package and delivered it 13 minutes later.
Google’s self-driving division gets a new name – Waymo, and becomes a separate entity under Alphabet. It will be interesting to see how a research project becomes a business. So far, the new company announced a partnership with Fiat Chrysler to deploy a ride-sharing service, which would see semi-self-driving Pacifica vans hit the roads to carry passengers as early as the end of 2017.
Meanwhile, Uber unleashed its own mini fleet of autonomous cars to the streets of San Francisco. And hours after they did it, the California Department of Motor Vehicles requested Uber to stop the experiment, because Uber doesn’t play according to the rules and does not have required permits. The cars also run on red lights, but Uber blames humans for this mistake.
Veebot is a robot designed to safely and quickly draw blood samples from humans. It uses computer vision to find veins and ultrasounds to make sure the vein is big enough. Then the robot chooses the right target and sticks the needle in.
Here a view on robotics and automation from an economist’s point of view. China is investing a lot of resources into robotics. Way more than US or Japan. Chinese plan is to automate manufacturing of goods and lowering the costs of production to make sure these goods are still being produced in China. It will have a big impact on economy and shows how a fourth industrial revolution will look like.
There is a robot that can fold your clothes. The future is now!
Zoltan Istvan compares CRISPR to the first nuclear explosion. Not in a destructive way, but how disruptive both technologies are. In his opinion, gene editing and making better humans might lead to a new Cold War between China and US, where the China does not care and enhances humans while the US has serious ethical problems that it has to solve to be still in the game.
The XPRIZE contest to build a real-life Tricoder from Star Trek is down to two finalists. They still have a lot of work ahead, because the next year their product will have to face clinical tests at UC San Diego.