AI – An experiment gone wrong?

The world seems to be in quite a rush to let AI takeover and jazz up life the way we have seen and admired in movies for decades. But have we intentionally turned a blind eye to its dark side? The side that can swallow the human race in a matter of time? It is really a thought to ponder, when at this point in time, issues like hacking, virus and malware  are our major concerns, are we even prepared for the devil that tags along with this advanced technology? What happens when it outsmarts us? Are we digging our own grave?

For now AI, is referred to as narrow AI because of its basic functions that it performs. Which in itself can pose a major threat even now! The most relevant example is Siri, who leaves us awestruck with her quick and sometimes even witty responses. Well! for now she only picks data from the Internet and shares it with us. The Strong AI, which we would like to term as the dormant AI, would have cognitive functions making it capable to strike conversations, learn and even reason. Pause! And retrospect what happens when it reaches this stage?

Flipside of the Coin

Primary concern with the the advancement in this technology is, it falling into the wrong hands. When weapons were first discovered the probable intent was to defend oneself from external danger like animals. However, we have a come a long way when we have turned these weapons each other. Likewise, what happens when the AI is used by our people against us?

Artificial Intelligence can be programmed to kill. Like it or not,in all possibility this technology would get into the wrong hands and can cause mass slaughter. According to Future of Life, such tech weapons are possible in narrow AI as well. What happens when we have strong AI? In all possibility, AI may treat human as a blockage in meeting a task allotted to it.

For now we lurch somewhere between strong and weak AI. The perfect example being, IBM’s supercomputer, Watson which has been designed to rummage through a lot of data to arrive at evidence based conclusions. In Fact, in 2011 it won $1 million on the quiz show Jeopardy against its former winners. Admirable? May be! But imagine the amount of data that it has access to within moments and what happens when it has access to all the data?

What you and I see today is Applied AI, that enables machines to perform specific tasks. That it is on the basis of data it can make medical diagnosis or recognise faces etc. General AI is yet to come into existence, which will give the machine the ability not to just deduce results from a specific type of data but also will allow it to perform any task that we humans perform.Which raises the next question.What happens to our jobs?

Time to Change the Focus?

We don’t want to be prophets of doom but, it isn’t for no reason that big names like Bill Gates and Elon Musk worry about the threat AI poses. We are progressing way ahead in AI advancement as compared to what was predicted a few years back. More than progressing towards super intelligence, it is important we figure out hacks to secure it.

With its growing intelligence, it is possible that with time, its goals may be aligned  on a different from ours. And we don’t know what this misaligned intelligence would eventually lead to. While it may take some decades for it to happen, what is actually worrying is the fact that it will take much more time to figure out a safety hack.