A mattress might be the last thing you'd dream of applying machine learning to, but your bed is where you spend a third of your life. And if it can help you sleep better, that could improve the hours of the day when you're not sleeping, as well. Now, there's a bed that promises to do some of the thinking for us to streamline our sleep time. At CES 2017, as I plopped myself down on the Sleep Number 360 smart bed, I wanted to find out: Can AI really us sleep better? Sure, it was conformable, and, yes, I looked ridiculous, but it's all for science.

Semiconductor Engineering sat down to discuss the increasing reliance on architectural choices for improvements in power, performance and area, with Sundari Mitra, CEO of NetSpeed Systems; Charlie Janac, chairman and CEO of Arteris; Simon Davidmann CEO of Imperas; John Koeter, vice president of marketing for IP and prototyping at Synopsys; and Chris Rowen, a consultant at Cadence. What follows are excerpts of that conversation. SE: Can IP be designed for an entire system, and does that change what has to be done architecturally? Janac: If you are using layers and stacks, you can go all the way from layout into architecture for a particular piece of a chip. It gets used by the architect, by the RTL developer, by the layout person, by the verification engineer, for what is essentially a vertical slice of the chip.

In 2017 Artificial Intelligence (AI) will redefine the possibilities of research and academia. Some doomsayers assume that most human employees will soon be replaced by smart machines. But I think that AI techniques will make many jobs better and more innovative. Bringing AI to the world of academic research, for example, will be akin to how Galileo's creation of the astronomical telescope in 1609 radically advanced the study of cosmology. For the Failure Institute (the research arm of the Fuckup Nights movement) AI will improve the speed and level of complexity at which we study how businesses and ideas fail.

Continuing their push to jointly develop self-driving vehicles, Intel (INTC), Mobileye (MBLY) and BMW (BMWYY) say they will put a fleet of about 40 autonomous vehicles on the road by the second half of 2017. The announcement, made at the CES show in Las Vegas on Wednesday, follows the partnership established by BMW Group, Intel and Mobileye last July. BMW plans to use Mobileye and Intel technology for "highly and fully automated driving" in the BMW iNext, an all-electric vehicle. The 40 autonomous vehicles will demonstrate "the significant advancements made by the three companies toward fully autonomous driving," according to a joint press release. They said the BMW 7 Series will employ cutting-edge Intel and Mobileye technologies during global trials that will start in the U.S. and Europe.

P-values for coefficients of cylinders, horsepower and acceleration are all greater than 0.05. This means that the relationship between the dependent and these independent variables is not significant at the 95% certainty level. I'll drop 2 of these variables and try again. High p-values for these independent variables do not mean that they definitely should not be used in the model. It could be that some other variables are correlated with these variables and making these variables less useful for prediction (check Multicollinearity).

Colonial First State data scientists are working with a team of computer engineering PhDs from the University of Technology, Sydney, to develop deep learning algorithms to predict investor responses to market shocks and tailor the communication of financial advice. A five-year partnership between Commonwealth Bank of Australia-owned CFS and UTS has resulted in the asset manager providing 20 years of investment and behavioural data for 1 million customers to machine learning researchers at the university, who are using its cutting-edge super-computers to forecast investor reaction. Peter Chun, the general manager of product and investment at Colonial First State, said artificial intelligence and big data analytics will also help the asset manager predict which customers might be more receptive to investment opportunities. He points to the example of the government's non-concessional contribution rules for superannuation; customers have a window of opportunity before June 30 to invest more than the caps.

In order to understand trends in the field, I find it helpful to think of developments in deep learning as being driven by three major frontiers that limit the success of artificial intelligence in general and deep learning in particular. Firstly, there is the available computing power and infrastructure, such as fast GPUs, cloud services providers (have you checked out Amazon's new EC2 P2 instance?) and tools (Tensorflow, Torch, Keras etc), secondly, there is the amount and quality of the training data and thirdly, the algorithms (CNN, LSTM, SGD) using the training data and running on the hardware. Invariably behind every new development or advancement, lies an expansion of one of these frontiers. Much of the progress we have seen this year is driven by an expansion of the former two frontiers; we now have systems that are able to recognize images and speech with an accuracy that rivals that of humans and there is an abundance of data and tools to develop them. However, almost all of these systems rely on supervised learning and thereby on the ready availability of labelled data sets.

LG Electronics will use machine learning technology to offer customized after sales (AS) services for its smartphone users, the company announced. The South Korean tech giant said it will adopt machine learning, big data analytics, and other, latest artificial intelligence technologies sequentially, starting in the first quarter of this year, to offer remote AS services via its apps. Thanks to machine learning, the app's diagnosis will improve over time, data will be processed faster, and LG will be able to offer services customized for the individual customer. The South Korean tech giant said over 80 percent of visits to service centres from its smartphone users were for simple requests or software problems. The remote service will drastically reduce its customers' needs to visit physical service centres, it said Artificial intelligence technology will be put into its "Smart Doctor" app.

Professor Einstein rolls his eyes, sticks out his tongue, and can give a simple explanation of the theory of relativity. With his lifelike rubbery "skin" and bushy mustache, he can almost make you forget he's a robot. The Einstein robot is among dozens roaming the Consumer Electronics Show in Las Vegas that can be your companion, educator or babysit your children. While robots have been around for years, advances in technology and artificial intelligence have allowed developers to give them traits that enable the devices to be seen as members of the family. "We make robots that have personality and come to life," said Andy Rifkin, chief technology officer of Hanson Robotics, the Hong Kong-based firm which is bringing the $299 Einstein robot to consumers this year.

No billion-dollar machine could replace a doctor. But a $25 machine can tell you when you need one. In 1996, the ER at Cook County Hospital of Chicago used an algorithm to determine when a patient with chest pain was in danger of having a heart attack and was thus worth one of its scarce hospital beds. Using a systematic, flowchart-based approach of basic tests, the algorithm proved not only to be quick and efficient, but accurate: It sorted 70 percent more patients into the low-risk category, but caught a higher percentage of heart attacks (95 percent) than human doctors (75-89 percent). And this was before any deep computing was involved. Now consider that there are around 6.4 billion IoT devices in use this year -- nearly one for every living human.