California has forced Uber to remove its self-driving vehicles from the road, canceling the company's controversial pilot program in San Francisco after a week of embarrassing reports of traffic violations and repeated legal threats from state officials. The department of motor vehicles (DMV) announced late Wednesday that it had revoked the registration of 16 autonomous Uber cars, which the corporation deployed without proper permits last week and which were caught on numerous occasions running red lights. Uber, which had previously declared that its rejection of government regulations was an "important issue of principle", confirmed that it has stopped its pilot in a statement, adding: "We're now looking at where we can redeploy these cars but remain 100 percent committed to California and will be redoubling our efforts to develop workable statewide rules." DMV officials and state attorney attorney general Kamala Harris have noted that Uber must get a testing permit to test its Volvo XC90s, which are navigated by a computer system but have a driver in the front seat who can intervene when needed. "It was determined that the registrations were improperly issued for these vehicles because they were not properly marked as test vehicles," the DMV said in a statement.

I've been thinking lately about why there is so much hype recently among lawyers and legal market commentators about artificial intelligence (AI). I was not ready to put hands to keyboard for a full explanation, so on Sunday I recorded a short video on Twitter stating my preliminary views. It's always hard to predict reactions to a single Tweet. This video generated a lot of discussion on Twitter and at least one direct reply blog post. I capture here some of that that commentary. By the way, I don't rehearse my videos and I don't have notes when I record them.

You ask your smartphone virtual assistant to make an appointment for you. You receive a message alert from your bank enquiring if you made a certain transaction. You receive recommendations for music or movies or online purchases based on your past behaviour. These are all examples of Artificial Intelligence (AI) entering your daily life. There is no widely accepted definition of the term or what constitutes AI. Definitions are usually based on some variation of computerized systems or computers exhibiting behaviour or thought that is normally demonstrated by humans or requires intelligence (which itself is hard to define). It could involve rationally solving complex problems or taking appropriate actions to achieve objectives in real world circumstances.

Smile Vector is a Twitter bot that can make any celebrity smile. Its results aren't perfect, but they're created completely automatically, and it's just a small hint of what's to come as artificial intelligence opens a new world of image, audio, and video fakery. Imagine a version of Photoshop that can edit an image as easily as you can edit a Word document -- will we ever trust our own eyes again? "I definitely think that this will be a quantum step forward," Tom White, the creator of Smile Vector, tells The Verge. "Not only in our ability to manipulate images but really their prevalence in our society."

There was the voice recognition software that struggled to understand women, the crime prediction algorithm that targeted black neighbourhoods and the online ad platform which was more likely to show men highly paid executive jobs. Concerns have been growing about AI's so-called "white guy problem" and now scientists have devised a way to test whether an algorithm is introducing gender or racial biases into decision-making. Mortiz Hardt, a senior research scientist at Google and a co-author of the paper, said: "Decisions based on machine learning can be both incredibly useful and have a profound impact on our lives ... Despite the need, a vetted methodology in machine learning for preventing this kind of discrimination based on sensitive attributes has been lacking." A beauty contest was judged by AI and the robots didn't like dark skin The paper was one of several on detecting discrimination by algorithms to be presented at the Neural Information Processing Systems (NIPS) conference in Barcelona this month, indicating a growing recognition of the problem. Nathan Srebro, a computer scientist at the Toyota Technological Institute at Chicago and co-author, said: "We are trying to enforce that you will not have inappropriate bias in the statistical prediction."

Privacy groups have filed a complaint about My Friend Cayla dolls to the Federal Trade Commission, arguing that they spy on children. Privacy groups have filed a complaint about My Friend Cayla dolls to the Federal Trade Commission, arguing that they spy on children. Tech toys have become popular holiday gifts. Many are interactive, some even claim educational benefits. But one such toy has privacy advocates very worried this year.

This hunger for data, or "statistical inefficiency" is perhaps the most significant practical limitation of current deep learning technology. Many of our clients at Scyfer have problems that could be solved by deep learning, but don't have large annotated datasets. Scyfer Active Learning Platform: once integrated, our system will passively observe the work of a domain expert (whether that's a medical doctor diagnosing patients or a factory worker identifying defective products). As the system is starting to learn how to imitate the expert, it will identify its own weaknesses and ask for guidance from the expert, thereby greatly accelerating its learning without requiring so many examples. Data-efficient deep networks: by building in prior knowledge, like "a rotated teddy bear is still a teddy bear", we can drastically reduce the number of examples required to learn a new concept.

And while the word is out that law enforcement agencies from California to New York have used the devices to monitor citizens for years, a new report (PDF) from the bipartisan House Oversight and Government Reform Committee shows that the rules governing their usage can vary greatly from state to state or even department to department. As a result, committee chairman Jason Chaffetz (R-UT) and member Elijah Cummings (D-MD) are calling on Congress to establish "a clear, nationwide framework that ensures the privacy of all Americans are adequately protected." The committee has already pushed the Department of Justice, the Department of Homeland Security and the IRS to require a warrant before deploying a Stingray or other similar cell network-spoofing device, but the report's findings showed that in many states law enforcement agencies don't even need probable cause in order to justify their usage. To remedy that situation, the report recommends that Congress pass clear rules about "when and how geolocation information can be accessed and used." The DOJ and DHS will then be responsible for requiring local law enforcement agencies to adopt the framework before they can receive federal funding for Stingray devices like the ones that violated FCC regulations in Baltimore.

There was the voice recognition software that struggled to understand women, the crime prediction algorithm that targeted black neighbourhoods and the online ad platform which was more likely to show men highly paid executive jobs. Concerns have been growing about AI's so-called "white guy problem" and now scientists have devised a way to test whether an algorithm is introducing gender or racial biases into decision-making. Mortiz Hardt, a senior research scientist at Google who led the work, said: "Decisions based on machine learning can be both incredibly useful and have a profound impact on our lives ... Despite the need, a vetted methodology in machine learning for preventing this kind of discrimination based on sensitive attributes has been lacking." A beauty contest was judged by AI and the robots didn't like dark skin Hardt's was one of several papers on detecting discrimination by algorithms to be presented at the Neural Information Processing Systems (NIPS) conference in Barcelona this month, indicating a growing recognition of the problem. Nathan Srebro, a computer scientist at the University of Chicago and co-author, said: "We are trying to enforce that you will not have inappropriate bias in the statistical prediction."