Are we really committed to prohibiting chemical weapons, or only committed if no one is upset, only committed if no one complains, only committed if no one is put out?

By Jeremy Kahn Babylon Healthcare Services Ltd., the fast-growing mobile medical consultation service, said its artificial intelligence software, in tests, can assess common conditions more accurately than human doctors. London-based Babylon's AI correctly answered 81 percent of diagnostic questions designed to mimic those trainee doctors must answer as part of the Royal College of General Practitioner's exam that must be passed to become a qualified GP doctor in the U.K. The exam is graded on a curve, but over the past five years, the average score trainees needed to pass was 72 percent. Babylon demonstrated this technology publicly for the first time in a live test at an event at London's Royal College of Physicians Wednesday. The company said it would publish the results of its tests online. It's made the AI tool available for free through its app and website in some parts of the U.K., including London, as well as Rwanda. Because the technology has not been approved by regulators, Babylon calls the software's answers "health information" not "diagnoses."

One day the AIs are going to look back on us the same way we look at fossil skeletons on the plains of Africa. An upright ape living in dust with crude language and tools, all set for extinction.

People have never travelled as much as they do today. At the same time, travellers have never been so demanding. We want it all and we want it now. We expect speed, authenticity, personalization, seamlessness and security. To deliver on these high expectations, technology is a must.

Accurately predicting when and where ambulance call-outs occur can reduce response times and ensure the patient receives urgent care sooner. Here we present a novel method for ambulance demand prediction using Gaussian process regression (GPR) in time and geographic space. The method exhibits superior accuracy to MEDIC, a method which has been used in industry. The use of GPR has additional benefits such as the quantification of uncertainty with each prediction, the choice of kernel functions to encode prior knowledge and the ability to capture spatial correlation. Measures to increase the utility of GPR in the current context, with large training sets and a Poisson-distributed output, are outlined.

Given a classifier ensemble and a set of examples to be classified, many examples may be confidently and accurately classified after only a subset of the base models in the ensemble are evaluated. This can reduce both mean latency and CPU while maintaining the high accuracy of the original ensemble. To achieve such gains, we propose jointly optimizing a fixed evaluation order of the base models and early-stopping thresholds. Our proposed objective is a combinatorial optimization problem, but we provide a greedy algorithm that achieves a 4-approximation of the optimal solution for certain cases. For those cases, this is also the best achievable polynomial time approximation bound unless $P = NP$. Experiments on benchmark and real-world problems show that the proposed Quit When You Can (QWYC) algorithm can speed-up average evaluation time by $2$x--$4$x, and is around $1.5$x faster than prior work. QWYC's joint optimization of ordering and thresholds also performed better in experiments than various fixed orderings, including gradient boosted trees' ordering.

Standard approaches to probabilistic reasoning require that one possesses an explicit model of the distribution in question. But, the empirical learning of models of probability distributions from partial observations is a problem for which efficient algorithms are generally not known. In this work we consider the use of bounded-degree fragments of the "sum-of-squares" logic as a probability logic. Prior work has shown that we can decide refutability for such fragments in polynomial-time. We propose to use such fragments to answer queries about whether a given probability distribution satisfies a given system of constraints and bounds on expected values. We show that in answering such queries, such constraints and bounds can be implicitly learned from partial observations in polynomial-time as well. It is known that this logic is capable of deriving many bounds that are useful in probabilistic analysis. We show here that it furthermore captures useful polynomial-time fragments of resolution. Thus, these fragments are also quite expressive.

This, the 50th-anniversary summer of Arthur C. Clarke's and Stanley Kubrick's "2001: A Space Odyssey," is arguably the beginning of big-budget Hollywood science fiction as we now know it. Without "2001," the "Star Wars," "Alien," and the "Star Trek" film franchises might not be the force they are today. For this decidedly serious and sometimes ponderous 1968 film opus proved that science fiction could be both profitable and profound. "Space Odyssey: Stanley Kubrick, Arthur C. Clarke and the Making of a Masterpiece," is author Michael Benson's recently-published, fascinating and extraordinarily detailed take on the pair's four-year collaboration. The film's principal photography began on a U.K. sound stage at Shepperton Studios outside London on December 30, 1965. Benson writes "2001: A Space Odyssey" encompassed four million years of human evolution, from pre-human Australopithecine man-apes struggling to survive in southern Africa, through to twenty-first-century space-faring Homo sapiens, then on to the death and rebirth of their Odysseus astronaut, Dave Bowman, as an eerily posthuman "star child."

Claims that a chatbot can diagnose medical conditions as accurately as a GP have sparked a row between the software's creators and UK doctors. Babylon, the company behind the NHS GP at Hand app, says its follow-up software achieves medical exam scores that are on-par with human doctors. It revealed the artificial intelligence bot at an event held at the Royal College of Physicians. But another medical professional body said it doubted the AI's abilities. "No app or algorithm will be able to do what a GP does," said the Royal College of General Practitioners.

Microsoft has updated it's facial recognition technology in an attempt to make it less'racist'. It follows a study published in March that criticised the technology for being able to more accurately recognise the gender of people with lighter skin tones. The system was found to perform best on males with lighter skin and worst on females with darker skin. The problem largely comes down to the data being used to train the AI system not containing enough images of people with darker skin tones. Experts from the computing firm say their tweaks have significantly reduced these errors, by up to 20 times for people with darker faces.