A judge in California ruled Thursday that U.S. authorities cannot force people to unlock technology with fingerprint or facial recognition, even with a search warrant. A judge in California ruled Thursday that U.S. authorities cannot force people to unlock technology via fingerprint or facial recognition, even with a search warrant. Magistrate Judge Kandis Westmore, of the U.S. District Court for the Northern District of California, made the ruling as investigators tried to access someone's property in Oakland. Two people allegedly used Facebook messenger to threaten a victim with the release of an "embarrassing video" if they didn't hand over money. Authorities investigating the case requested a search and seizure warrant "to seize various items" believed to be at a home connected to the suspects.

The panel "Let's Chat About Bots" gathered representatives from four companies that have built robust bots for interacting with customers onFacebook's chat platform Messenger: Anastasia Sartan cofounder and CEO of the Russian e-commerce site Epytom Stylist; Hussein Fazal, CEO of the travel booking site SnapTravel; Felipe Bernal, head of product innovation at the Brazilian IT company Movile; and Chema Alonso, head of digital at the telecom giant Telefonica. Angelique Kamara, of Facebook Messenger Partnerships, moderated. "When we were analyzing how different types of customers were interacting with internet services, a lot of people preferred point and click, others preferred to chat with a bot," said Alonso. "…We didn't want to force one specific channel. We wanted to be in the channels our customers love. And Facebook Messenger is one of them."

These systems don't just require more information than humans to understand concepts or recognize features, they require hundreds of thousands times more, says Neil Lawrence, a professor of machine learning at the University of Sheffield and part of Amazon's AI team. Once they've been trained, they can be incredibly efficient at tasks like recognizing cats or playing Atari games, says Google DeepMind research scientist Raia Hadsell. A solution to this might be something called progressive neural networks -- this means connecting separate deep learning systems together so that they can pass on certain bits of information. One way of doing this is revisiting an old, unfashionable strand of artificial intelligence known as symbolic AI or Good Old-Fashioned Artificial Intelligence (GOFAI), says Murray Shanahan, a professor of cognitive robotics at Imperial College London (and also the scientific advisor on Ex Machina).

Speaking to attendees at a deep learning conference in London last month, there was one particularly noteworthy recurring theme: humility, or at least, the need for it. While companies like Google are confidently pronouncing that we live in an "AI-first age," with machine learning breaking new ground in areas like speech and image recognition, those at the front lines of AI research are keen to point out that there's still a lot of work to be done. Just because we have digital assistants that sound like the talking computers in movies doesn't mean we're much closer to creating true artificial intelligence. Problems include the need for vast amounts of data to power deep learning systems; our inability to create AI that is good at more than one task; and the lack of insight we have into how these systems work in the first place. Machine learning in 2016 is creating brilliant tools, but they can be hard to explain, costly to train, and often mysterious even to their creators.