After months of planning, Intel officially announced this week the opening of its Collaborative Research institute for Computational Intelligence. Working in cooperation with both the Technion and Hebrew University, Intel will be spending $15 million in the coming years to develop projects that will take computers to the point where they can begin to think for themselves, or at least anticipate problems before they happen and suggest solutions.
At a gala event in Tel Aviv, top Intel executives presented details of the plan, which combines advancing work in the areas of machine learning, which enables the combining of massive amounts of data from a range of sources (the cloud, the sensors, the network and the environment) with historical and current private and public data. The aim is to create solutions to everyday needs, along with moving forward research in advanced computer architecture, to produce better, faster, and smaller processors to provide the computational power needed for the advanced apps and devices the institute hopes to produce.
Attending the event was Science and Technology Minister Daniel Hershkowitz, who praised the cooperation between industry and academia as “a wonderful approach to research. It’s not surprising that Intel is at the forefront of this welcome trend, and it makes sense that computational intelligence would be the subject of this cooperation.”
Also appearing at the event were two of Intel’s top executives, Corporate Vice President and Chief Technology Officer Jason Rattner, and Intel Israel head and senior VP of world Intel, Mooly Eden. At a press conference after the event, Rattner described details of the program, what kind of technology the Institute plans to produce, and when it’s expected to come to market. In a word, said Rattner, “it’s about the future of computing.”
The computer, formerly just a dumb box, is about to get a slew of new capabilities, thanks to the Computational Center. Life will get more convenient, safer, easier — or less private, depending on your point of view. One project Rattner described involves a device (or an app) that keeps tabs on what we do, and when and how we do it. The technologies and devices will take all the data gathered by sensors — GPS, accelerometers, cameras, and others yet to be developed — and mashup the information, with advanced algorithms learning the relationship between the data and making “decisions,” suggestions, and choices for us, all in the interest of making our lives less complicated.
For example, the device (or a series of devices collaborating in the cloud) could videotape us as we go about our daily activities. It we forget where we left the car keys (the system knows we need them because we are standing by the car door) it will tell or message us that we left the keys upstairs on the dresser. If it determines that we are serial forgetters, the system will take the liberty of reminding us not to forget the keys when we approach the door to go outside.
Once we’re in the car, we’ll find even more smart apps, devices, and technologies. “We’re working on this with all the major car companies, including Nissan, Toyota, BMW, and others. ‘Intel Inside’ is a logo we’ll see not only on computers, but in all sorts of intelligent car apps and products,” Rattner said. Working together, Technion and Hebrew University researchers will combine their advances in machine learning and architecture to produce those technologies, and the products that will make them useful and accessible.
Already, more rudimentary versions of these smart machine learning technologies and form factors are out there, Rattner said, describing a smart shoe sales system, components of which are already in use in the US. “You walk up to a pair of signs in the shoe store and it sizes you up, deciding if you are a man, woman, adult, or child. Then it runs a program which shows you the shoes appropriate to your needs — men’s styles for men, etc. With a few more gestures you can call up a specific model and size, and a salesperson will bring the shoes you gestured for.
If all this sounds a bit big brother-ish, Rattner sympathizes. “A few years ago we made security one of the pillars of Intel technology,” a commitment the company backed up with cash when it acquired McAfee, the security software company. “When developing new technologies, especially smart technologies like these where you have the ability to aggregate lots of information, it is incumbent to make that the data remains remains private and secure, and is only disclosed in ways with which the individual is comfortable.”
In other words, we should feel safe using whatever the Computational Institute comes up with, Rattner said — because the security is going to be built-in. “Security is difficult to retrofit on devices and technologies that were built in an insecure manner,” he said, citing email as an example. “If we had known that security was going to be an issue then, we would have designed in differently, in a more secure manner.” Now, though, designers have no excuse, and they will be designing smart technologies with built-in (and scalable, to ensure that they can keep up with the hackers) security systems.
And although he is helping to helm a project that some would consider a bit “invasive,” Rattner himself believes in keeping a low (online) profile. “Personally I think there is too much information about us on the Internet,” he said. “I think that society in general should embrace the view that our personal information is personal. We, not others, should be setting the agenda on how it is to be used.”