To really understand the impact of artificial intelligence in the modern world, it's best to think beyond the mega-research projects like those that helped Google recognize cats in photos. According to professor Ajay Agrawal of the University of Toronto, humanity should be pondering how the ability of cutting edge A.I. techniques like deep learning--which has boosted the ability for computers to recognize patterns in enormous loads of data--could reshape the global economy. Making his comments at the Machine Learning and the Market for Intelligence conference this week by the Rotman School of Management at the University of Toronto, Agrawal likened the current boom of A.I. to 1995, when the Internet went mainstream. Gaining enough mainstream traction, the Internet ceased to be seen as a new technology. Instead, it was a new economy where businesses could emerge online.

Adversarial input can fool a machine-learning algorithm into misperceiving images. Over the past five years, machine learning has blossomed from a promising but immature technology into one that can achieve close to human-level performance on a wide array of tasks. In the near future, it is likely to be incorporated into an increasing number of technologies that directly impact society, from self-driving cars to virtual assistants to facial-recognition software. Yet machine learning also offers brand-new opportunities for hackers. Malicious inputs specially crafted by an adversary can "poison" a machine learning algorithm during its training period, or dupe it after it has been trained.

Machine Learning (ML) tasks are becoming pervasive in a broad range of applications, and in a broad range of systems (from embedded systems to data centers). As computer architectures evolve toward heterogeneous multi-cores composed of a mix of cores and hardware accelerators, designing hardware accelerators for ML techniques can simultaneously achieve high efficiency and broad application scope. While efficient computational primitives are important for a hardware accelerator, inefficient memory transfers can potentially void the throughput, energy, or cost advantages of accelerators, that is, an Amdahl's law effect, and thus, they should become a first-order concern, just like in processors, rather than an element factored in accelerator design on a second step. In this article, we introduce a series of hardware accelerators (i.e., the DianNao family) designed for ML (especially neural networks), with a special emphasis on the impact of memory on accelerator design, performance, and energy. We show that, on a number of representative neural network layers, it is possible to achieve a speedup of 450.65x over a GPU, and reduce the energy by 150.31x on average for a 64-chip DaDianNao system (a member of the DianNao family).

Video Friday is your weekly selection of awesome robotics videos, collected by your Automaton bloggers. We'll also be posting a weekly calendar of upcoming robotics events for the next two months; here's what we have so far (send us your events!): Let us know if you have suggestions for next week, and enjoy today's videos. I really don't know much about this beyond what's in the video, but we don't see a lot of Russian robots around here, so: It looks to be a project from the Russian equivalent of DARPA, designed to go into space by 2021. This video of Agile Justin feeling up different kinds of rods was a finalist for both "IROS Best Paper on Cognitive Robotics" and "IROS Best Student Paper": This is cool because until this point, building pneumatic robots required making molds and casting custom parts.

Department of Transportation Secretary Anthony Foxx recently released a 116-page policy document that aims to guide automakers and technologists on best-practices when it comes to the manufacturing and deployment of autonomous vehicle features. Apple, which has been rumored to be building a car, recently laid off employees of its automotive project and pivoted from making a car to creating autonomous software, according to reports. Another aftermarket self-driving tech company recently completed a successful 120-mile beer delivery without anyone at the wheel. A big rig cab equipped with sensors made by Otto, a startup bought by Uber recently for $670 million, made the delivery of Budweiser beer while its driver rested in the sleeper berth during most of the trip down Colorado's Interstate 25.

The new MacBook Pro was always going to be expensive. But it is very expensive indeed in the UK, after the prices were shifted because of the falling pound. The computers are as much as £500 more expensive than previous models, because of a combination of the currency changes and the upgrades introduced by Apple. Buying the cheapest version of the MacBook Pro with Touch Bar that was released yesterday will cost customers £1,749. Customers can spend as much as £2,699 on stock models of the computer, if they buy the 15-inch version.

The New York businessman with a penchant for celebrity television may suddenly find himself in love with artificial intelligence developed in India. The polls and simulations that involve the skills and insight of human beings suggest Donald Trump could be heading for something of a pasting. But an artificial intelligence (AI) system developed in Mumbai, and which correctly predicted the last three US presidential elections, puts the Republican nominee ahead of his rival Hillary Clinton in the battle to secure the keys to the White House. MogIA was developed by Sanjiv Rai, the founder of Indian start-up Genic.ai. It has taken 20 million data points from public platforms such as Google, Facebook and Twitter and analysed the information to create predictions, CNBC reported.

European privacy experts have sent letters to WhatsApp telling it to stop sharing people's data with Facebook. WhatsApp announced in recent weeks that it would start handing over information about the people who use it to Facebook, so that its parent company could use that data to better target ads. But the company didn't give a very easy way of opting out of it, and the deal has drawn the attention of customers and regulators. Now EU privacy watchdogs have told WhatsApp and Facebook that the deal must be stopped until it can be checked whether it is legal or not. Presumably if the deal is found to be illegal then it will be forced to stop.

Adult website Pornhub says that it can save Vine, the video sharing company, by buying it. The company has sent an official letter to Jack Dorsey, the CEO of Vine's owner Twitter, offering to buy the company. In doing so it hopes to help avoid it being shut down entirely, as Twitter and Vine have announced it will do. The company does make clear that the site would turn into a porn site. The letter sent to Mr Dorsey by Pornhub CEO Corey Price says that Vine would be restored "back to its NSFW glory".

Apple has launched its latest products in its headquarters in Cupertino, California. In the 24 hours since, I've been trying out the new MacBook Pro for size. First of all, I should point out that this is the entry-level MacBook Pro, not the more expensive MacBook Pro with TouchBar. TouchBar is the ground-breaking innovation where the top row of function keys is replaced with a long, touch-sensitive display strip. I tried this briefly and it's frankly pretty amazing. But I'll be reviewing that and the sapphire-crystal power and Touch ID button also found on the MacBook Pro with TouchBar in due course.