With machine learning algorithms evolving at an incredibly fast pace, concerns are mounting whether artificial intelligence (AI) is the logical continuation of human history or its demise. RT talked to three experts in the field about the benefits and dangers of AI.
Three AI experts engaged in a debate on RT about the benefits and dangers of rapidly-developing technology and AI.
Stanford researchers have developed an algorithm that offers diagnoses based off chest X-ray images. It can diagnose up to 14 types of medical conditions and is able to diagnose pneumonia better than expert radiologists working alone.
A paper about the algorithm, called CheXNet, was published Nov. 14 on the open-access, scientific preprint website arXiv.
“Interpreting X-ray images to diagnose pathologies like pneumonia is very challenging, and we know that there’s a lot of variability in the diagnoses radiologists arrive at,” said Pranav Rajpurkar, a graduate student in the Machine Learning Group at Stanford and co-lead author of the paper. “We became interested in developing machine learning algorithms that could learn from hundreds of thousands of chest X-ray diagnoses and make accurate diagnoses.”
In the mid-1900s, art historian Maurits Michel van Dantzig developed a system to identify artists by their brush or pen strokes, which he called Pictology. Dantzig found shape, length, direction, and pressure all contributed to a kind of stroke signature, unique to each artist.
New research with contributions from The Hague suggests that Pictology might be the key to helping machines understand art, allowing systems to quickly verify whether brushstrokes were from an original painter or a forger.
After analyzing 80,000 brushstrokes from 297 digitized sketches and drawings, an AI system was able to spot forged drawings in the style of Pablo Picasso, Henri Matisse, and Egon Schiele with 100% accuracy. The “fakes” were commissioned recreations of specific drawings, which the algorithms had not been shown previously.
Of course, it’s not actually all that doom and gloom, the child AI is really only capable of a specific task – image recognition. Using its AutoML AI, Google’s AI-building AI created its child AI using a technique called reinforcement learning. This works just like machine learning, except it’s entirely automated where AutoML acts as the neural network for its task-driven AI child.
Known as NASNet, the child AI was tasked with recognising objects in a video, in real time. AutoML would then evaluate how good NASNet was at its task and then improve its algorithms using the data to create a superior version of NASNet.
READ NEXT: Watching an AI create fake celebrity faces is nightmare fuel.
And it is on Singles’ Day, automation, robots, AI and machine learning will be widely applied to all aspects of the annual shopping ritual, right from product selection to delivery.
This year’s November 11 shopping ritual will engage a recommendation algorithm, robots, and chatbots capable of understanding human emotion.
Thank a new approach to spoofing image recognition AIs, developed by a team from Kyushu University in Japan, for that joke.
Trying to catch out AIs is a popular pastime for many researchers, and we’ve reported machine-learning spoofs in the past. The general approach is to add features to images that will incorrectly trigger a neural network and have it identify what it sees as something else entirely.
The new research, published on the arXiv, describes an algorithm that can efficiently identify the best pixels to alter in order to confuse an AI into mislabeling a picture. By changing just one pixel in a 1,024-pixel image, the software can trick an AI about 74 percent of the time. That figure rises to around 87 percent if five pixels are tweaked.
Google unveiled software aimed at making it easier for scientists to use the quantum computers in a move designed to give a boost to the nascent industry.
The software, which is open-source and free to use, could be used by chemists and material scientists to adapt algorithms and equations to run on quantum computers. It comes at a time when Google, IBM, Intel Corp., Microsoft Corp. and D-Wave Systems Inc. are all pushing to create quantum computers that can be used for commercial applications.
A group of astronomers from the universities of Groningen, Naples and Bonn has developed a method that finds gravitational lenses in enormous piles of observations. The method is based on the same artificial intelligence algorithm that Google, Facebook and Tesla have been using in the last years. The researchers published their method and 56 new gravitational lens candidates in the November issue of Monthly Notices of the Royal Astronomical Society.
When a galaxy is hidden behind another galaxy, we can sometimes see the hidden one around the front system. This phenomenon is called a gravitational lens, because it emerges from Einstein’s general relativity theory which says that mass can bend light. Astronomers search for gravitational lenses because they help in the research of dark matter.
The hunt for gravitational lenses is painstaking. Astronomers have to sort thousands of images. They are assisted by enthusiastic volunteers around the world. So far, the search was more or less in line with the availability of new images. But thanks to new observations with special telescopes that reflect large sections of the sky, millions of images are added. Humans cannot keep up with that pace.
One of the most defining scientific discoveries in recent decades is the development of induced pluripotent stem cells, which lets scientists revert adult cells back into an embryonic-like blank state and then manipulating them to become a particular kind of tissue.
But now a new model could do away with this time-consuming process, taking out the middle step and directly programming cells to become whatever we want them to be.
“Cells in our body always self-specialise,” explains bioinformatics researcher Indika Rajapakse from the University of Michigan.
