A technology called "predictive algorithms" can learn what your face looks like and then predict what your face would look like in other situations. Unsurprisingly, some are using it to map celebrities' faces onto the bodies of porn stars having sex. In the latest episode of Today, Explained, Vox's Aja Romano tells host Sean Rameswaram how "deepfakes" are spreading across the internet, what it means for consent, and what platforms like Reddit are doing -- or not doing -- to stop it. Later in the episode, Peter Eckersley, chief computer scientist at the Electronic Frontier Foundation, explores how the same technology could tear our society apart in bigger ways -- namely by encouraging the spread of fake news. How do I get even more Today, Explained?

January 25 & 26 in San Francisco will see the sixteenth global Deep Learning Summit and the fifth global AI Assistant Summit joined by the first ever Deep Learning for Enterprise Summit. Use code KDNUGGETS to save 20% on Early Bird passes!

A while back, I was reading an article posted on Facebook, about Clovis people found alive and well living in Florida, with a picture featuring tribesmen (see below.) The quality of the picture was poor, and the URL was very suspicious: baynews9.com.ddwg.clonezone.link, as to make it appear that it was from Baynews9.com. It turned out that the picture (and thus the whole story) was fake: these people are real people living in Peru, see here for a Youtube video about them. My question is how to detect that a story is fake? The picture might have metadata embedded in it, allowing the data scientist to find the real source, unless it is a screenshot.

Communication is an essential pillar of society. Humanity's progression over the past millennium was largely driven by the development and evolution of communication as a tool for distributing siloed thoughts from one individual to others. Communication is naively defined as content and the mode of transmission -- symbols manifested as images, language transmitted through speech and writing, digital files sent through the internet. These are methods through which we communicate thoughts, ideas, facts, and opinions. New forms of communication emerge to expand the lexicon of thought and reduce the friction required to create and transmit content.

This is what our readers loved most in February. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. However, our picks and opinions are independent from USA Today's newsroom and any business incentives. Despite being a short month, February was full of deals on some great products. We saw sales on high-end OLED and 4K smart TVs, mattresses infused with green tea, Instant Pots, robot vacuums, and tons of kitchen and tech accessories.

We've been running into performance issues when training an LSTM/RNN based Seq2Seq to create a multivariate, online trainable forecasting model - Does anyone here have experience with much harder (1000, n-dimensional) problems? From the paper, I feel like this might lead to a pretty dramatic improvement in performance.

Strictly Come Dancing and Dancing with the Stars judges watch out. Artificial intelligence is taking on its next challenge: dance talent appraisal. It sounds unlikely, but Abu Zaher Faridee and colleagues at the University of Maryland, Baltimore County, revealed their AI-assisted plan to help dance teachers at the HotMobile conference in Arizona, last month. Dancers often get in each other's way as they make their moves and normally significantly outnumber the teacher. This makes it hard for them to assess if everyone is dancing correctly.

Internet giants Google, Facebook and Twitter are facing renewed pressure to tackle the problem of terrorist propaganda online after the European Commission (EC) gave them just a one hour deadline to remove offensive content from their pages or face penalties. The EC's demand comes at a time when the major search and social media companies are being urged to do more to censor inappropriate or illegal material posted by users and hosted on their domains. "Considering that terrorist content is most harmful in the first hours of its appearance online, all companies should remove such content within one hour from its referral as a general rule," the EC said in a statement. The commission will also ask companies to report back on the degree of co-operation they receive from other organisations in order to determine whether stricter legislation is necessary. Most online media companies have clear rules in place warning users against publishing hate speech and routinely investigate and remove troubling content as soon as it is reported by users.

Google's new Clips camera takes short looping videos of your kids and your pets, and it does it all using AI. Sure, that sounds like some weird technophobic nightmare--Google, one of the most data-thirsty companies on the planet, pointing a camera at your kids? But the way the company designed the Clips keeps all of your visual memories private until you decide to share them, if you decide to share them at all. Mike and Arielle talk about how Google's innovative approach to computer vision in consumer products could lead to other types of digital cameras that are as private as a Polaroid. Send the hosts feedback on their personal Twitter feeds.

Technology and culture writer Navneet Alang, based in Toronto, Canada, outlined his concerns over the gadgets in an in-depth commentary published by The Week. The pocket-sized digital camera, which decides on its own whether an image is interesting enough to shoot, was launched on Tuesday. Reviews of its performance have been mixed, but some experts are more worried about what it might lead to. Even devices with seemingly innocent purposes can go on to become more sinister, he says. Mr Alang believes Clips is a test-bed for broader applications of'smart camera' technology, including for spying on us.