Europe's Human Brain Project, the team behind the program was trying to figure out how to incentivize researchers to share data. Now they face a more fundamental problem: getting researchers to take part in any capacity. At the time of writing, 280 scientists have signed an open letter threatening to boycott the project.
The letter lays bare tensions that have existed since the early days of the initiative. Some of the complaints relate to how the project is being managed--notably the way in which one subproject was dropped--but the scientists who signed the letter have more fundamental concerns about exactly what Europe is trying to achieve with its €1 billion ($1.4 billion) budget. The headline objective is to create a computer simulation of the human brain, a goal University College London's computational neuroscience director Peter Dayan called "radically premature" in an interview with The Guardian.
With some scientists thinking the goal is ambitious at best, they have looked to subprojects to ensure that the project yields tangible benefits even if it falls short of its big objective. But the open letter criticizes the project for being too narrow from the start and becoming more so in a recent refocusing. The team behind the project argues that it will deliver a range of benefits, particularly by giving neuroscientists the tools to manage and analyze the ever-larger data sets being generated by researchers.
Henry Markram, head of the Human Brain Project
"The rationale of the Human Brain Project is a plan for data: what do we do with all this data? This is a very exciting [information computer technology] project that will bring completely new tools and capabilities to all of neuroscience. It is not a general neuroscience funding source for more of the same research," Henry Markram, head of the Human Brain Project, said. In calling the initiative an ICT project, Markram hit upon one of the problems some people have with the project.
"There is a danger that Europe thinks it is investing in a big neuroscience project here, but it's not. It's an IT project. They need to widen the scope and take advantage of the expertise we have in neuroscience," Geneva University's Alexandre Pouget said.
While researchers in Europe were drafting a letter objecting to aspects of the Human Brain Project (HBP), a company across the Atlantic was showcasing the fruits of 9 years of work into mimicking neurology. And while the company, Numenta, has made progress in creating apps that reproduce the brain, it thinks a fully functioning model is impossible without rethinking the underlying hardware.
Numenta set up shop in 2005 with the objective of replicating the processing power of the brain and has just introduced its first product, VentureBeat reports. Numenta began by building biological and computer science models of parts of the brain, leading to the creation of an algorithm that replicates a small, 1,000- to 5,000-nerve-cell piece of the cortex. This algorithm underpins its first product, a tool for detecting unusual patterns in IT systems.
Best of 2014: How Google Cracked House Number Identification in Street View
Google can identify and transcribe all the views it has of street numbers in France in less than an hour, thanks to a neural network that’s just as good as human operators. In January, its engineers revealed how they developed it.
Their method turns out to rely on a neural network that contains 11 levels of neurons that they have trained to spot numbers in images.
Google’s Intelligence Designer
The man behind a startup acquired by Google for $628 million plans to build a revolutionary new artificial intelligence.
Demis Hassabis started playing chess at age four and soon blossomed into a child prodigy. At age eight, success on the chessboard led him to ponder two questions that have obsessed him ever since: first, how does the brain learn to master complex tasks; and second, could computers ever do the same?
Now 38, Hassabis puzzles over those questions for Google, having sold his little-known London-based startup, DeepMind, to the search company earlier this year for a reported 400 million pounds ($650 million at the time).
Google snapped up DeepMind shortly after it demonstrated software capable of teaching itself to play classic video games to a super-human level (see “Is Google Cornering the Market on Deep Learning?”). At the TED conference in Vancouver this year, Google CEO Larry Page gushed about Hassabis and called his company’s technology “one of the most exciting things I’ve seen in a long time.”
Researchers are already looking for ways that DeepMind technology could improve some of Google’s existing products, such as search. But if the technology progresses as Hassabis hopes, it could change the role that computers play in many fields.
DeepMind seeks to build artificial intelligence software that can learn when faced with almost any problem. This could help address some of the world’s most intractable problems, says Hassabis. “AI has huge potential to be amazing for humanity,” he says. “It will really accelerate progress in solving disease and all these things we’re making relatively slow progress on at the moment.”
At Harrah’s Casino on the shores of Lake Tahoe, DeepMind researchers showed off software that had learned to play three classic Atari games - Pong, Breakout and Enduro - better than an expert human. The software wasn’t programmed with any information on how to play; it was equipped only with access to the controls and the display, knowledge of the score, and an instinct to make that score as high as possible. The program became an expert gamer through trial and error.
No one had ever demonstrated software that could learn to master such a complex task from scratch. DeepMind had made use of a newly fashionable machine learning technique called deep learning, which involves processing data through networks of crudely simulated neurons (see “10 Breakthrough Technologies 2013: Deep Learning”). But it had combined deep learning with other tricks to make something with an unexpected level of intelligence.
DeepMind had combined deep learning with a technique called reinforcement learning, which is inspired by the work of animal psychologists such as B.F. Skinner. This led to software that learns by taking actions and receiving feedback on their effects, as humans or animals often do.
Artificial intelligence researchers have been tinkering with reinforcement learning for decades. But until DeepMind’s Atari demo, no one had built a system capable of learning anything nearly as complex as how to play a computer game, says Hassabis. One reason it was possible was a trick borrowed from his favorite area of the brain. Part of the Atari-playing software’s learning process involved replaying its past experiences over and over to try and extract the most accurate hints on what it should do in the future. “That’s something that we know the brain does,” says Hassabis. “When you go to sleep your hippocampus replays the memory of the day back to your cortex.”
But Hassabis sounds more excited when he talks about going beyond just tweaking the algorithms behind today’s products. He dreams of creating “AI scientists” that could do things like generate and test new hypotheses about disease in the lab. When prodded, he also says that DeepMind’s software could also be useful to robotics, an area in which Google has recently invested heavily (see “The Robots Running This Way”). “One reason we don’t have more robots doing more helpful things is that they’re usually preprogrammed,” he says. “They’re very bad at dealing with the unexpected or learning new things.”
Brain-Training for Baseball Robot
In other advanced computing news, researchers from the University of Electro-Communications in Japan have developed a robotic system with powerful new learning capability and motor control, mimicking how the cerebellum in the human brain functions. In their tests, the researchers were able to “teach” a robotic system to adjust to the trajectory of a ball and hit it with a bat, similar to how a baseball player must process complex visual information quickly in order to make contact. The graphics-processing unit built for this exercise comprised over 100,000 “neurons,” which allowed the robot to learn over time and then process reactions in real time.
In One Aspect of Vision, Computers Catch Up to Primate Brain
MIT scientists, writing in PLoS Computational Biology, reveal the creation of computer technology, which is roughly on par with the human brain for visual recognition of objects. For over 40 years, researchers have worked to develop “neural networks” based on what is known about to brain to attempt to create computers capable of recognizing an object in real time with the same speed and efficiency as the primate brain. One possible future application of this new technology could be the ability to use computer recognition systems to help repair the human eye after an injury.
By Tom Simonite on December 30, 2014
A phase-change memory chip that learns to recognize handwritten numbers by simulating a network of neurons is tested in IBM’s Almaden Research Center near San Jose, California.
Researchers at IBM used what’s known as phase-change memory to build a device that processes data in a way inspired by the workings of a biological brain. Using a prototype phase-change memory chip, the researchers configured the system to act like a network of 913 neurons with 165,000 connections, or synapses, between them. The strength of those connections change as the chip processes incoming data, altering how the virtual neurons influence one another. By exploiting that property, the researchers got the system to learn to recognize handwritten numbers.
A Startup’s Neural Network Can Understand Video
Software that understands what it sees in video could lead to new forms of advertising, or make video editing easier.
The company says its software can rapidly analyze video clips to recognize 10,000 different objects or types of scene. In a demo given last week at a conference on deep learning, Clarifai’s cofounder and CEO Matthew Zeiler uploaded a clip that included footage of a varied alpine landscape. The software created a timeline with graph lines summarizing when different objects or types of scene were detected. It showed exactly when “snow” and “mountains” occurred individually and together. The software can analyze video faster than a human could watch it; in the demonstration, the 3.5 minute clip was processed in just 10 seconds.
Researchers unveiled a software system Wednesday which had taught itself to play 49 different video games and proceeded to defeat human professionals -- a major step in the fast-developing Artificial Intelligence realm.