Monthly Archives: March 2017

For an AI to understand thirst, you must first teach it about Drake Quartz What if Messenger Day automatically decided which groups received your photos based on an AI's understanding of what thirst looks like, combined with its knowledge of who should receive thirsty posts? More ambitious still: Could AI power a bot that ...

Read more from the original source:

For an AI to understand thirst, you must first teach it about Drake - Quartz

Charlie Osborne | ZDNet

Neural networking could pave the way for AI systems to be given a capability which we have, until now, considered a human trait: the possession of common sense.

While some of us may have less of this than others, the idea of "common sense" -- albeit a vague concept -- is the general idea of making fair and good decisions in what is a complex environment, drawing on our own experience and an understanding of the world, rather than relying on structured information -- something which artificial intelligence has trouble with.

This kind of intuition is a human concept, but according to Facebook AI research group director Yann LeCun, leaps forward in neural networking and machine vision could one day lead to software with common sense.

Speaking to MIT's Technology Review, LeCun said there is still "progress to be made" when it comes to neural networking which is required for machine vision.

Neural networks are artificial systems which mimic the structure of the human brain, and by combining this with more advanced machine vision -- which are ways to pull data from imagery for use in tasks and decision-making -- LeCun says common sense will be the result.

For example, if you have a dominant object in an image, and enough data in object categories, machines can recognize specific objects like dogs, plants, or cars. However, some AI systems can now also recognize more abstract groupings, such as weddings, sunsets, and landscapes.

LeCun says that just five years ago, this wasn't possible, but as machines are granted vision, machine expertise is growing.

AI is still limited to the specific areas that humans train them in. You could show an AI system an image of a dog at a wedding, but unless the AI has seen one before and understands the context of the image, the response is likely to be what the executive calls "garbage." As such, they lack common sense.

Facebook wants to change this. LeCun says that while you can interact with an intelligent system through language to recognize objects, "language is a very low-bandwidth channel" -- and humans have a wealth of background knowledge which helps them interpret language, something machines do not currently have the capability to draw on in real-time to make contextual connections in a way which mimics common sense.

One way to solve this problem could be through visual learning and media such as streamed images and video.

"If you tell a machine "This is a smartphone," "This is a steamroller," "There are certain things you can move by pushing and others you cannot," perhaps the machine will learn basic knowledge about how the world works," LeCun told the publication. "Kind of like how babies learn."

"One of the things we really want to do is get machines to acquire the very large number of facts that represent the constraints of the real world just by observing it through video or other channels," the executive added. "That's what would allow them to acquire common sense, in the end."

By giving intelligent machines the power to observe the world, contextual gaps will be filled and it may be that AI could make a serious leap from programmed algorithms and set answers. One area, for example, Facebook wants to explore is the idea of AI systems being able to predict future events by showing them a few frames.

"If we can train a system to do this we think we'll have developed techniques at the root of an unsupervised learning system," LeCun says. "That is where, in my opinion, a lot of interesting things are likely to happen. The applications for this are not necessarily in vision -- it's a big part of our effort in making progress in AI."

The next 2 Facebook moves that will disrupt the world:

Continue reading here:

Facebook AI chief: We can give machines common sense - ZDNet

Flippy the robot is starting its culinary career with one simple task, but just like any rookie, it is learning on the job. With some practice and training, Flippy will be able to do everything from chopping vegetables to plating meals like a pro. Miso Robotics created the robot, which debuted in a kitchen at the restaurant chain CaliBurger in Pasadena, Calif., this week.

"Flippy will initially only focus on flipping burgers and placing them on buns," David Zito, CEO of Miso Robotics tells ZDNet. He adds, "But since Flippy is powered by our own cooking AI software, it will continuously learn from its experiences to improve and adapt over time. This means Flippy will learn to take on additional tasks including grilling chicken, bacon, onions, and buns in addition to frying, prepping, and finishing plates. Eventually, Flippy will support CaliBurger's entire menu."

The robot can be installed in kitchens in less than five minutes, and it's designed to work alongside restaurant staff. Flippy will even politely move aside if it gets in someone's way. Computer vision and deep learning software make it much smarter than your average kitchen appliance.

"Flippy features a Sensor Bar allowing it to see in 3D, thermal, and regular vision for detecting the exact temperatures of the grill as well as readiness of each burger, which will expand to other menu items as Flippy continues to learn and adapt," says Zito.

Flippy uses computer vision and AI to cook burgers. (Image: Miso Robotics)

Flippy will be installed in more than 50 CaliBurger restaurants worldwide by the end of 2019.

"The application of artificial intelligence to robotic systems that work next to our employees in CaliBurger restaurants will allow us to make food faster, safer and with fewer errors," said John Miller, chairman of Cali Group, in a statement. "Our investment in Miso Robotics is part of our broader vision for creating a unified operating system that will control all aspects of a restaurant from in-store interactive gaming entertainment to automated ordering and cooking processes, 'intelligent' food delivery and real-time detection of operating errors and pathogens."

Automation is creeping into kitchens in many forms. This week, Chowbotics (formerly Casabots) announced that it raised $5 million of Series A funding for food service robots. Then there's also Grillbot pro, which is like a Roomba for your grill. Moley Robotics is developing a fully automated and intelligent robotic chef. Various robots sell pizza, cook it, deliver it, and can even print it in outer space .

VIDEO: FDNY uses drone to tame Bronx fire

More:

Flippy the robot uses AI to cook burgers - ZDNet

Forays by Google subsidiary DeepMind Health into the UKs medical institutions have been characterized by two major themes. First, amazing results powered by cutting-edge AI; and second, a lack of transparency over the handling of the UKs public-funded data. With the science going swimmingly, DeepMind Health is focusing more than ever on reassuring UK citizens that their medical records are in safe hands. Its latest plan is a public ledger that shows which bits of data its using; when; and for what purposes.

The initiative is called the Verifiable Data Audit, and was announced this week in a blogpost written by DeepMind co-founder Mustafa Suleyman and the companys head of security and transparency, Ben Laurie. The Audit technology is not yet in place, but would keep a publicly accessible record of every time DeepMind accesses hospital data, using technology related to the blockchain.

Each time theres any interaction with data, well begin to add an entry to a special digital ledger, write Suleyman and Laurie. That entry will record the fact that a particular piece of data has been used, and also the reason why for example, that blood test data was checked against the NHS national algorithm to detect possible acute kidney injury.

Like blockchain technologies, this information will be write-only it cant be edited after the fact or deleted. It will also make use of cryptographic proofs that will allow experts to verify the integrity of the data. Unlike most blockchain systems, though, the ledger wont be distributed among members of the public, but stored by a number of entities including data processors like DeepMind Health and health care providers. The company says this wont impede the verification process, and that the choice was made to make the ledger more efficient. Blockchain entities like Bitcoin are distributed among lots of different players require a lot of power (computing and literal) to compile and check as much as a small country, according to some estimates.

Speaking to The Guardian, Nicola Perrin of the Wellcome Trust said the technology should create a robust audit trail for public health data managed by DeepMind. One of the main criticisms about DeepMinds collaboration with the Royal Free [Hospital Trust] was the difficulty of distinguishing between uses of data for care and for research, said Perrin. This type of approach could help address that challenge, and suggests they are trying to respond to the concerns. DeepMind Health says it wants implement the first pieces of the audit later this year.

More:

Google's AI subsidiary turns to blockchain technology to track UK ... - The Verge

The line for the SXSW panel Eyes in the Sky: The Future of AI and Satellites snaked around many corners in Austins JW Marriot Hotel understandably, AI coupled with space shit, bring it on.

Spaceknow Incs CEO Pavel Machalek did most of the talking during this session. Spaceknow is a San Francisco based company building an AI system that can process the petabytes of data from the hundreds of commercial satellites circling us up above.

Run an early-stage company? We're inviting 250 to exhibit at TNW Conference and pitch on stage!

We are digitizing the physical world, so we can build apps on top it, Machalek stated. According to the Czech CEO, were currently going through a sea of change in how we use satellite data.

Everything from camera technology to actual satellites to launching those satellites to space is getting cheaper. Couple that with the abundance of computer power and the development of more robust machine learning system, and it follows that we can start extracting actionable information about the world, Machalek says

His company works for lots of industrial clients, who want to know how many ships visit a certain harbor, or how many trucks pull up to a refinery to move oil. But some of the information theyre extracting is also coupled to the Bloomberg terminal, informing investors about the growth of industrial areas in China.

By counting and classifying things you get an as objective grip on reality as possible, he says, after telling a story on how the information they collect contradicted the official numbers the Chinese government put out. In a world like this, in which people make up statistics, our numbers offer an objective look.

In a similar way, Spaceknow also distributes the Africa Night Lights Index, an index that is based on the light intensity measured by satellites and then aggregated according to individual countries as a more reliable economic indicator for developing countries in Africa.

In the end, Machalek says that hed like to cover the whole world with Spaceknows system, allowing anyone with a smartphone to do real-time queries about real-world data meaning you could check from space how long the line is for the bar you want to go to, I guess.

Read next: Future voice interfaces could turn us all into geniuses -- or idiots

The rest is here:

Satellites and AI will bring real-time, real-world data to your phone - TNW