Category Archives: Artificial Intelligence

Context aware computing

Posted on by
1

Android eyeNever drunk dial again. That is the promise from Intel CTO, Justin Rattner, as he discussed context aware computing and the next generation of personal devices, at the annual Intel Develop Forum in San Francisco.

Combining GPS technology with data from microphones, cameras, heart monitors and brain scans, new apps could track and document your every move. Don’t worry about taking a picture of that landmark; your phone already did because it knows you never remember to. It also updated your Facebook profile and checked you in on Foursquare.

If you can get past the privacy concerns, this technology may have potential.

TV that recommends what to watch, based on who is holding the remote-because it can tell who you are by the way you hold it.

Apps that suggest where to eat in your area because the phone knows you just ran  5 miles and need those 40g of protein to continue your krebs cycle and it knows nobody wants to be around you when your blood sugar drops.

Context aware devices may have their day, and in certain situations certainly have their place, but with the serious implications to privacy and the flawed security structure of the internet, the zeitgeist will need to change significantly for this technology to be accepted.

Then again, I thought American Idol was rubbish, so what do I know?

Making computers that think like us

Posted on by
1

Photo of Tu-4 courtesy of Pavel Adzhigildaev. Accessed on Wikipedi (http://en.wikipedia.org/wiki/File:TU-4-MONIN0.jpg).In 1944, several of America’s most advanced  bomber planes, the B-29 Superfortress, landed in the Soviet Union. Years ahead of anything that the Soviets had in the way of strategic bombers at the time  (the B-29 was the plane used to drop the atomic bombs on Japan) Stalin hastily ordered that the design of the American airplane be understood. Over the protests of the Americans, the Soviets studied the construction of the planes, and, though they had no technical manuals of any kind, produced exact working replicas of the bombers under the name Tupolev Tu-4. These planes served in the Soviet Air Force for many years.

This process – studying a complicated system, frequently in order to replicate it – is called reverse engineering. And scientists, in a less hostilely plagiaristic form, are currently working to perform the same technique on the human brain.

An article titled “Reverse Engineering of Human Brain Likely by 2020” on Gizmodo.com talks about the potential for making computers think like humans do. There are two major challenges:

  1. Computational power – currently, as the article notes, even our best supercomputers lack the sheer power to calculate as much as the human brain. What is required is estimated, from the complexity of the genome base pairs that code the brain, as “a computational capacity of at least 36.8 petaflops and a memory capacity of 3.2 petabytes – a scale that supercomputer technology isn’t expected to hit for at least three years”.
  2. Complexity – the brain functions in ways that are very different from the linear way of computation that most computers utilize. Although some groups are hoping to move beyond the linear model and are exploring parallel and other forms of processing, this is undoubtedly the greatest challenge and will take much longer than three years to resolve.

Computers that can one day think like humans is a humbling thought. Just like the Industrial Revolution edged out the individual craftsman, so too may computers that think like us one day replace the “brain jobs” that have become so prevalent in the modern age. But that, thankfully for those of us who work with our brains, is many, many years away.