The traditional definition of artificial intelligence is the ability of machines to execute tasks and solve problems in ways normally attributed to humans. Some tasks that we consider simple--recognizing an object in a photo, driving a car--are incredibly complex for AI. Machines can surpass us when it comes to things like playing chess, but those machines are limited by the manual nature of their programming; a 30 gadget can beat us at a board game, but it can't do--or learn to do--anything else. This is where machine learning comes in. Show millions of cat photos to a machine, and it will hone its algorithms to improve at recognizing pictures of cats.

The FBI maintains a huge database of more than 411m photos culled from sources including driver's licenses, passport applications and visa applications, which it cross-references with photos of criminal suspects using largely untested and questionably accurate facial recognition software. A study from the Government Accountability Office (GAO) released on Wednesday for the first time revealed the extent of the program, which had been queried several years before through a Freedom of Information Act request from the Electronic Frontier Foundation (EFF). The GAO, a watchdog office internal to the US federal government, found that the FBI did not appropriately disclose the database's impact on public privacy until it audited the bureau in May. The office recommended that the attorney general determine why the FBI did not obey the disclosure requirements, and that it conduct accuracy tests to determine whether the software is correctly cross-referencing driver's licenses and passport photos with images of criminal suspects. The Department of Justice "disagreed" with three of the GAO's six recommendations, according to the office, which affirmed their validity.

The FBI has amassed more than 411 million photos as part of its vast facial recognition database, according to a government watchdog -- which in a new report criticized the system for its lack of safeguards and protections. That includes millions of individuals' driving license photos, as well as photos of foreigners applying for visas, and criminal mugshots. Where the millions of facial recognition sets of data came from. But the Government Accountability Office (GAO) criticized the FBI in a new report, published Wednesday, saying that the agency had not properly tested the system or balanced civil liberties and privacy. The system kicked into operations in early-2015 after what appeared to be a relatively successful pilot.

The Singularity is a term you'll find in science and in science fiction. It was coined by mathematician John von Neumann to define a theoretical moment when the artificial intelligence of computers surpasses the capacity of the human brain. The term is borrowed from physics and quantum mechanics, where the term gravitational singularity is used in the study of black holes. These events are all considered singular because we are unable to predict what happens next; the disruptive degree of change associated with the event is simply too great for our current body of knowledge. While we are far from attaining the goal of artificial intelligence, there was a brief flurry of excitement recently when a computer passed the Turing Test, to mixed reviews.

The FBI has fallen short on assessing the privacy risks and accuracy of a huge facial recognition database used by several law enforcement agencies, a government auditor has said. A new report, released by the U.S. Government Accountability Office Wednesday, shows the FBI's use of facial recognition technology is "far greater" than previously understood, said Senator Al Franken, the Minnesota Democrat who requested the GAO report. The FBI's Next Generation Identification-Interstate Photo System (NGI-IPS), which allows law enforcement agencies to search a database of more than 30 million photos of 16.9 million people, raises serious privacy concerns, Franken added in a press release. "Facial recognition technology is a new and powerful tool that holds great promise for law enforcement," he said. "But if we don't ensure its accuracy and guard against misuse, I am concerned about the risk of innocent Americans being inadvertently swept up in criminal investigations."

The traditional definition of artificial intelligence is the ability of machines to execute tasks and solve problems in ways normally attributed to humans. Some tasks that we consider simple--recognizing an object in a photo, driving a car--are incredibly complex for AI. Machines can surpass us when it comes to things like playing chess, but those machines are limited by the manual nature of their programming; a 30 gadget can beat us at a board game, but it can't do--or learn to do--anything else. This is where machine learning comes in. Show millions of cat photos to a machine, and it will hone its algorithms to improve at recognizing pictures of cats.

After almost 10 years of working incognito, former National Aeronautics and Space Administration head Daniel Goldin is finally ready to formally present KnuEdge to the world. KnuEdge is a "neural technology innovation company," an outfit that builds hardware and software based on neural technology, with a main focus on human-machine interaction. While newly revealed publicly, it has been in stealth mode for a decade now, and has already raised 100 million in funding to build its neural chips. The company has revealed its two primary products: KnuVerse, which is a voice authentication technology, and KnuPath, its state-of-the-art neural chip. It has also unveiled Knurld.io, a software development kit with a cloud-based voice recognition and authentication service. Foremost of these offerings is KnuPath.

Before becoming a reality, AlphaBounce had to be shaped by ARAMIS, a motion-capture system that helps determine the amount of strain and tension caused by different materials. ARAMIS, which NASA has used to inspect the outer hull of space shuttles, combines high-speed cameras with flexion sensors to gather information at up to 500 frames per second. Since the software maps skin, bone and muscle, it can give Adidas engineers insight into how they should go about designing a more comfortable running sneaker. ARAMIS can visualize the level of comfort on every area of an individual's foot, from the heel to the toes, as well as indicate when the fabric may be getting in the way of performance. "It's a really versatile tool," said George Robusti, senior design director of global running at Adidas, of the ARAMIS system.

Digital disruption has the potential to threaten 40 per cent of jobs over the next 10 to 15 years as automation and machine learning shake up the economy, according to a Productivity Commission report out today. In research entitled Digital Disruption: What do governments need to do?, the Commission warned that governments and regulators need to prepare for changing times as "disruption" moves beyond Uber and Air BnB. Productivity Commission chairman Peter Harris said developing disruptive technologies of machine intelligence and automation will gradually change economies. "There's little doubt that in some sectors there will be dislocation of labour and dislocation of capital. "It's not just a cost to employees, it will be a cost to certain businesses as well," Mr Harris told The World Today. "Things like 3D printing are going to have an impact.

What if you could get anything you wanted just by asking for it? It sounds like a fantastical prospect, but this kind of interaction now appears to be the goal of a category of software tools called "bots." Bots have long been used to automate repetitive computing tasks, but are now increasingly used to communicate directly with users. These artificial intelligence programs mimic human speech to answer questions, order products, and complete many other tasks. Several technology companies are working on integrating bots into their own products and competing to add functionality.