The rapid pace of artificial intelligence (AI) has raised fears about whether robots could act unethically or soon choose to harm humans. Some are calling for bans on robotics research; others are calling for more research to understand how AI might be constrained. But how can robots learn ethical behavior if there is no “user manual” for being human?

Continue Reading…

by JOSHUA OGAWA, Nikkei staff writer

The game of GO is considered more complex than either chess or shogi.

Google has developed a computer program capable of beating professional go players, opening the door to new applications — and new ethical concerns — as artificial intelligence draws closer to matching human thinking.

The U.S. tech giant presented its results in the Jan. 27 issue of Nature. Though a computer managed to topple the world chess champion in 1997 and defeat the top women’s shogi (Japanese chess) player in 2010, Google’s AlphaGo program (official blog site) is the first to triumph over professional go players under official rules.

TOUGHER THAN CHESS The go board is larger than that used in chess or shogi, with total possible gameplay scenarios numbering 10 to the 360th power. Anticipating and solving all possible board combinations is impossible even for today’s most advanced computers, and many researchers had predicted that a program capable of besting pro players was at least 10 years away.

Google’s go AI bypassed that problem with deep learning, a technology that mimics human neural pathways and learning processes. Rather than working through the possible scenarios by brute force, the program considers the board as a whole and draws on accumulated experience to choose its next move. The company has used such technology before. Last year, it presented its deep Q-network algorithm, which let computers master electronic games by analyzing pixel and score data over repeated plays.

For its go project, Google collaborated with pro players to teach the computer 30 million plays, eventually enabling it to predict humans’ moves with 57% accuracy. The AI then was put through several million matches against itself, forcing it to work out winning strategies by experience. Its ability to select the best move by analyzing the state of the board is now nearly equal to a human’s ability to do the same based on skill and intuition.

The program can beat existing go software 99.8% of the time, and it won all five games against reigning European champion Fan Hui in October. The AI will face another five-game challenge in March, when it goes up against Lee Se-dol, one of the world’s top players.

WAY FORWARD The question now is where Google will direct its AI efforts next. Games are an excellent arena in which to develop and test AI, but the goal is to turn such technology toward solving real-world problems, said Demis Hassabis, Google’s AI chief. The priority, he indicated, is on developing robust multipurpose AI tech.

Deep learning is particularly promising, given its ability to process visual and audio information in a manner resembling human perception byGoogle Deep Mind LOGO finding patterns in large data sets. Recent research has sought to apply the technology in diverse fields, such as using it to control robots’ movements or to analyze medical image data to help diagnose patients. Simple forms are already at work in today’s tech, such as voice recognition software on smartphones. The dawn of AI capable of replicating the intuition of human professionals would drastically expand the current slate of applications.

Yet researchers also are growing more cautious as AI advances. Though Google is pleased to have overcome a major challenge for AI, Hassabis said, the company is aware of the ethical issues surrounding the technology. Academics and other public figures have warned that unchecked development could lead to AI programs that are hostile to society as a whole.

As with “any powerful new technology,” developers of AI must “take seriously our responsibilities” and “have ethical concerns at the top of our minds,” Hassabis told the BBC last year. Google has established an AI ethics board to address those concerns. Now that technology has won out against humanity in one of the ultimate game-based challenges, the focus of research should turn to cooperation between man and machine.

MIT-deep learning chip Eyeriss
Chip could bring deep learning to mobile devices

via MASSACHUSETTS INSTITUTE OF TECHNOLOGY

In recent years, some of the most exciting advances in artificial intelligence have come courtesy of convolutional neural networks, large virtual networks of simple information-processing units, which are loosely modeled on the anatomy of the human brain.

Neural networks are typically implemented using graphics processing units (GPUs), special-purpose graphics chips found in all computing devices with screens. A mobile GPU, of the type found in a cell phone, might have almost 200 cores, or processing units, making it well suited to simulating a network of distributed processors.

At the International Solid State Circuits Conference in San Francisco this week, MIT researchers presented a new chip designed specifically to implement neural networks. It is 10 times as efficient as a mobile GPU, so it could enable mobile devices to run powerful artificial-intelligence algorithms locally, rather than uploading data to the Internet for processing.

Neural nets were widely studied in the early days of artificial-intelligence research, but by the 1970s, they’d fallen out of favor. In the past decade, however, they’ve enjoyed a revival, under the name “deep learning.”

“Deep learning is useful for many applications, such as object recognition, speech, face detection,” says Vivienne Sze, an assistant professor of electrical engineering at MIT whose group developed the new chip. “Right now, the networks are pretty complex and are mostly run on high-power GPUs. You can imagine that if you can bring that functionality to your cell phone or embedded devices, you could still operate even if you don’t have a Wi-Fi connection. You might also want to process locally for privacy reasons. Processing it on your phone also avoids any transmission latency, so that you can react much faster for certain applications.”

The new chip, which the researchers dubbed “Eyeriss,” could also help usher in the “Internet of things” — the idea that vehicles, appliances, civil-engineering structures, manufacturing equipment, and even livestock would have sensors that report information directly to networked servers, aiding with maintenance and task coordination. With powerful artificial-intelligence algorithms on board, networked devices could make important decisions locally, entrusting only their conclusions, rather than raw personal data, to the Internet. And, of course, onboard neural networks would be useful to battery-powered autonomous robots.

Division of labor

A neural network is typically organized into layers, and each layer contains a large number of processing nodes. Data come in and are divided up among the nodes in the bottom layer. Each node manipulates the data it receives and passes the results on to nodes in the next layer, which manipulate the data they receive and pass on the results, and so on. The output of the final layer yields the solution to some computational problem.

In a convolutional neural net, many nodes in each layer process the same data in different ways. The networks can thus swell to enormous proportions. Although they outperform more conventional algorithms on many visual-processing tasks, they require much greater computational resources.

The particular manipulations performed by each node in a neural net are the result of a training process, in which the network tries to find correlations between raw data and labels applied to it by human annotators. With a chip like the one developed by the MIT researchers, a trained network could simply be exported to a mobile device.

This application imposes design constraints on the researchers. On one hand, the way to lower the chip’s power consumption and increase its efficiency is to make each processing unit as simple as possible; on the other hand, the chip has to be flexible enough to implement different types of networks tailored to different tasks.

Sze and her colleagues — Yu-Hsin Chen, a graduate student in electrical engineering and computer science and first author on the conference paper; Joel Emer, a professor of the practice in MIT’s Department of Electrical Engineering and Computer Science, and a senior distinguished research scientist at the chip manufacturer NVidia, and, with Sze, one of the project’s two principal investigators; and Tushar Krishna, who was a postdoc with the Singapore-MIT Alliance for Research and Technology when the work was done and is now an assistant professor of computer and electrical engineering at Georgia Tech — settled on a chip with 168 cores, roughly as many as a mobile GPU has.

Act locally

The key to Eyeriss’s efficiency is to minimize the frequency with which cores need to exchange data with distant memory banks, an operation that consumes a good deal of time and energy. Whereas many of the cores in a GPU share a single, large memory bank, each of the Eyeriss cores has its own memory. Moreover, the chip has a circuit that compresses data before sending it to individual cores.

Each core is also able to communicate directly with its immediate neighbors, so that if they need to share data, they don’t have to route it through main memory. This is essential in a convolutional neural network, in which so many nodes are processing the same data.

The final key to the chip’s efficiency is special-purpose circuitry that allocates tasks across cores. In its local memory, a core needs to store not only the data manipulated by the nodes it’s simulating but data describing the nodes themselves. The allocation circuit can be reconfigured for different types of networks, automatically distributing both types of data across cores in a way that maximizes the amount of work that each of them can do before fetching more data from main memory.

At the conference, the MIT researchers used Eyeriss to implement a neural network that performs an image-recognition task, the first time that a state-of-the-art neural network has been demonstrated on a custom chip.

READ MORE: http://news.mit.edu/2016/neural-chip-artificial-intelligence-mobile-devices-0203

ADDITIONAL: http://web.mit.edu/

I came across artificial intelligence and was just enthralled. I went home the next day and bought a programming book and decided that was what I was going to teach myself to do.

Brittany Wenger , as a high school senior, brilliant young scientist and Grand Prize Winner in the 2012 Google Science Fair, for her project “Global Neural Network Cloud Service for Breast Cancer” talks about how she came to science in Research and Inspiration.

Ms. Wenger is currently at Duke University