The creeping growth of machine intelligence
An invisible revolution is occurring. While at the grocery store recently, I paid with a debit card. Being too lazy to use the pen, I pressed the on-screen ‘OK’ button on the device with my finger. The machine haughtily told me to please use the pen provided, not my appendage. I thought to myself, “I’m the human here, how can a computer tell me what to do?”
For several years now, companies have realised the value of embedding computers into every device, and the ability to feed back sensor information to ‘HQ’. Have you noticed how everything is getting ‘smarter’? Cable companies monitor channels watched, cellphones report their usage and diagnostic information back to network providers, and power companies are investing in smartgrid technologies that can transmit energy usage information back to the generating facility for dynamic and efficient electricity production. The effect of this is that the location of where decisions are being made is being pushed closer to the consumer.
This is not a bad thing, but needs to be done in a secure fashion. When we delegate more and more complicated decisions to machines, security, trust and integrity become more important. As I’ve hinted in a previous column, in the rush to bring these technologies to market, manufacturers often push security to the back burner.
To get an idea of just how many ‘internet connected’ devices there are, browse over to www.shodan.com. Shodan is a search engine specifically for finding devices that are web connected. It stores information on webcams, routers, power plants, iPhones, wind turbines, refrigerators, VoIP (voice over internet protocol) phones and more, and it lets you search for specific computer types. Security professionals use it to find devices vulnerable to attacks.
By 2020, there will be approximately eight billion people in the world and approximately 40-50 billion devices. What kind of problems will this cause, security or otherwise, for humanity increasingly dependent on the internet? We should take steps now to ensure systemic failure doesn’t occur.
Fortunately, there are people out there who are already keeping an eye on this type of issue. For example, the MOBISEC group of conferences brings together companies and security professionals to consider these exact problems and make decisions that will affect the future of mobile technology.
One concern is that the sheer volume of connected devices may ‘crowd out’ people using the internet. When the internet was first designed, the engineers never foresaw the speed at which the technology would be adopted. Already, the internet addressing mechanism is reaching its limit and there is a worldwide project to move to a new, more plentiful one. Adding thousands of devices simply makes the problem worse.
Another concern, recently reported by the BBC, regards the security of devices and the vulnerability to manipulation for nefarious purposes. Researchers recently discovered that medical implant devices can be reprogrammed wirelessly to the detriment of the patient without major effort, because of lax security. The lead researcher was quoted as saying “future devices will be . more connected to the internet and will have much more use of wireless technology. There is no silver bullet, its not that these problems are easy to address, but there is technology available to reduce these risks significantly.”
Technology can work wonders, but with anything there is risk, and these risks need to be properly identified and managed.