Tag Archives: more
For first time artificial intelligence has been integrated into a MEMS device Continue reading
Making sure artificial intelligence does what we want and behaves in predictable ways will be crucial as the technology becomes increasingly ubiquitous. It’s an area frequently neglected in the race to develop products, but DeepMind has now outlined its research agenda to tackle the problem.
AI safety, as the field is known, has been gaining prominence in recent years. That’s probably at least partly down to the overzealous warnings of a coming AI apocalypse from well-meaning, but underqualified pundits like Elon Musk and Stephen Hawking. But it’s also recognition of the fact that AI technology is quickly pervading all aspects of our lives, making decisions on everything from what movies we watch to whether we get a mortgage.
That’s why DeepMind hired a bevy of researchers who specialize in foreseeing the unforeseen consequences of the way we built AI back in 2016. And now the team has spelled out the three key domains they think require research if we’re going to build autonomous machines that do what we want.
In a new blog designed to provide updates on the team’s work, they introduce the ideas of specification, robustness, and assurance, which they say will act as the cornerstones of their future research. Specification involves making sure AI systems do what their operator intends; robustness means a system can cope with changes to its environment and attempts to throw it off course; and assurance involves our ability to understand what systems are doing and how to control them.
A classic thought experiment designed to illustrate how we could lose control of an AI system can help illustrate the problem of specification. Philosopher Nick Bostrom’s posited a hypothetical machine charged with making as many paperclips as possible. Because the creators fail to add what they might assume are obvious additional goals like not harming people, the AI wipes out humanity so we can’t switch it off before turning all matter in the universe into paperclips.
Obviously the example is extreme, but it shows how a poorly-specified goal can lead to unexpected and disastrous outcomes. Properly codifying the desires of the designer is no easy feat, though; often there are not neat ways to encompass both the explicit and implicit goals in ways that are understandable to the machine and don’t leave room for ambiguities, meaning we often rely on incomplete approximations.
The researchers note recent research by OpenAI in which an AI was trained to play a boat-racing game called CoastRunners. The game rewards players for hitting targets laid out along the race route. The AI worked out that it could get a higher score by repeatedly knocking over regenerating targets rather than actually completing the course. The blog post includes a link to a spreadsheet detailing scores of such examples.
Another key concern for AI designers is making their creation robust to the unpredictability of the real world. Despite their superhuman abilities on certain tasks, most cutting-edge AI systems are remarkably brittle. They tend to be trained on highly-curated datasets and so can fail when faced with unfamiliar input. This can happen by accident or by design—researchers have come up with numerous ways to trick image recognition algorithms into misclassifying things, including thinking a 3D printed tortoise was actually a gun.
Building systems that can deal with every possible encounter may not be feasible, so a big part of making AIs more robust may be getting them to avoid risks and ensuring they can recover from errors, or that they have failsafes to ensure errors don’t lead to catastrophic failure.
And finally, we need to have ways to make sure we can tell whether an AI is performing the way we expect it to. A key part of assurance is being able to effectively monitor systems and interpret what they’re doing—if we’re basing medical treatments or sentencing decisions on the output of an AI, we’d like to see the reasoning. That’s a major outstanding problem for popular deep learning approaches, which are largely indecipherable black boxes.
The other half of assurance is the ability to intervene if a machine isn’t behaving the way we’d like. But designing a reliable off switch is tough, because most learning systems have a strong incentive to prevent anyone from interfering with their goals.
The authors don’t pretend to have all the answers, but they hope the framework they’ve come up with can help guide others working on AI safety. While it may be some time before AI is truly in a position to do us harm, hopefully early efforts like these will mean it’s built on a solid foundation that ensures it is aligned with our goals.
Image Credit: cono0430 / Shutterstock.com Continue reading
Shopping is becoming less and less of a consumer experience—or, for many, less of a chore—as the list of things that can be bought online and delivered to our homes grows to include, well, almost anything you can think of. An Israeli startup is working to make shopping and deliveries even faster and cheaper—and they’re succeeding.
Last week, CommonSense Robotics announced the launch of its first autonomous micro-fulfillment center in Tel Aviv. The company claims the facility is the smallest of its type in the world at 6,000 square feet. For comparison’s sake—most fulfillment hubs that incorporate robotics are at least 120,000 square feet. Amazon’s upcoming facility in Bessemer, Alabama will be a massive 855,000 square feet.
The thing about a building whose square footage is in the hundred-thousands is, you can fit a lot of stuff inside it, but there aren’t many places you can fit the building itself, especially not in major urban areas. So most fulfillment centers are outside cities, which means more time and more money to get your Moroccan oil shampoo, or your vegetable garden starter kit, or your 100-pack of organic protein bars from that fulfillment center to your front door.
CommonSense Robotics built the Tel Aviv center in an area that was previously thought too small for warehouse infrastructure. “In order to fit our site into small, tight urban spaces, we’ve designed every single element of it to optimize for space efficiency,” said Avital Sterngold, VP of operations. Using a robotic sorting system that includes hundreds of robots, plus AI software that assigns them specific tasks, the facility can prepare orders in less than five minutes end-to-end.
It’s not all automated, though—there’s still some human labor in the mix. The robots fetch goods and bring them to a team of people, who then pack the individual orders.
CommonSense raised $20 million this year in a funding round led by Palo Alto-based Playground Global. The company hopes to expand its operations to the US and UK in 2019. Its business model is to charge retailers a fee for each order fulfilled, while maintaining ownership and operation of the fulfillment centers. The first retailers to jump on the bandwagon were Super-Pharm, a drugstore chain, and Rami Levy, a retail supermarket chain.
“Staying competitive in today’s market is anchored by delivering orders quickly and determining how to fulfill and deliver orders efficiently, which are always the most complex aspects of any ecommerce operation. With robotics, we will be able to fulfill and deliver orders in under one hour, all while saving costs on said fulfillment and delivery,” said Super-Pharm VP Yossi Cohen. “Before CommonSense Robotics, we offered our customers next-day home delivery. With this partnership, we are now able to offer our customers same-day delivery and will very soon be offering them one-hour delivery.”
Long live the instant gratification economy—and the increasingly sophisticated technology that’s enabling it.
Image Credit: SasinTipchai / Shutterstock.com Continue reading