Use BBC.com or the new BBC App to listen to BBC podcasts, Radio 4 and the World Service outside the UK. Is China Replacing Workers With Robots? Today, we discuss China's £14.7bn investment into AI and robotics in the hopes of overtaking the US as the world leader in innovation. China has two million robots already working in factories, and is investing heavily in the hopes that one day they could bridge gaps in an already sluggish economy. What can these robots do?

Use BBC.com or the new BBC App to listen to BBC podcasts, Radio 4 and the World Service outside the UK. The world's largest data centre (62sq miles) has been approved in Utah, but there is growing opposition towards the project. At twice the size of Manhattan with promises to create thousands of jobs, we look at the bi partisan opposition against it. In this episode, Justin and Anthony discuss the enormous buildings being built across rural America, to house the huge amounts of data that A.I companies work with. Tech bosses say the centres are essential to the growth of Artificial Intelligence.

During a recent episode of the January 6 Commission hearings -- this summer's most-talked-about new show -- comedy goddess Merrill Markoe took to Twitter to call attention to what she was seeing at the bottom of her screen: The Jan 6 hearings were once again amazing, stunning and magnificent. In a related story, because it is my responsibility and my job, I have recorded the valiant attempts made by the closed-captioning software to spell Pat Cippolone. "Patsy Baloney" was soon trending on Twitter, but that's not the really interesting part (though it was hilarious, which was Markoe's intent). For the first time I alone am responsible for change in the permanent political record. In the closed captioning for the Jan 6 hearings (where i got all the screen grabs above) mid-way in Patsy Baloney and many of her friends have been replaced w/correct spelling.

We propose a gossip-based distributed algorithm for Gaussian mixture learning, Newscast EM. The algorithm operates on network topologies where each node observes a local quantity and can communicate with other nodes in an arbitrary point-to-point fashion. The main difference between Newscast EM and the standard EM algorithm is that the M-step in our case is implemented in a decentralized manner: (random) pairs of nodes repeatedly exchange their local parameter estimates and combine them by (weighted) averaging. We provide theoretical evidence and demonstrate experimentally that, under this protocol, nodes converge exponentially fast to the correct estimates in each M-step of the EM algorithm.

We propose a gossip-based distributed algorithm for Gaussian mixture learning, Newscast EM. The algorithm operates on network topologies where each node observes a local quantity and can communicate with other nodes in an arbitrary point-to-point fashion. The main difference between Newscast EM and the standard EM algorithm is that the M-step in our case is implemented in a decentralized manner: (random) pairs of nodes repeatedly exchange their local parameter estimates and combine them by (weighted) averaging. We provide theoretical evidence and demonstrate experimentally that, under this protocol, nodes converge exponentially fast to the correct estimates in each M-step of the EM algorithm.

We propose a gossip-based distributed algorithm for Gaussian mixture learning, Newscast EM. The algorithm operates on network topologies where each node observes a local quantity and can communicate with other nodes in an arbitrary point-to-point fashion. The main difference between Newscast EM and the standard EM algorithm is that the M-step in our case is implemented in a decentralized manner: (random) pairs of nodes repeatedly exchange their local parameter estimates and combine themby (weighted) averaging. We provide theoretical evidence and demonstrate experimentally that, under this protocol, nodes converge exponentially fastto the correct estimates in each M-step of the EM algorithm.