The curious case of the boiling frog and rise of artificial intelligence
There is a furious debate raging in the technology world right now. On one side of the debate are luminaries like Elon Musk and Bill Gates. “If you’re not concerned about AI (artificial intelligence) safety, you should be. Vastly more risk than North Korea”, tweeted Musk in August, with a picture that had the caption: “In the end the machines will win.” A month earlier, he had warned that AI represents “a fundamental risk to the existence of human civilization”.
On the other side are younger luminaries like the Google founders, and Facebook co-founder Mark Zuckerberg, with the latter waxing eloquent on how AI would “bring so many improvements to our quality of life” and calling Musk’s assertions “pretty irresponsible”. The last word belonged to Musk, who tweeted: “his (Zuckerberg’s) understanding of the subject is pretty limited.”
The real question is, “Who will have the last word—us humans, or our AI creations?”
Most of us have heard about doomsday scenarios before. The advent of the automobile was supposed to put the horse-and-buggy drivers out of work. Long-distance aircraft were supposed to decimate ships and ocean liners. When computers came in, for instance, millions of people were supposed to lose their jobs. Instead, millions of new jobs were generated, and entirely new industries created.
In the AI world, there is a concept called Singularity. Ray Kurzweil, the well-known futurist who is claimed to have been 86% accurate in the 147 predictions he made, defines Singularity as the time when machine intelligence will be infinitely more powerful than all human intelligence combined. Singularity is also the point at which machine intelligence and humans would merge. 2029 will be the year when computers will have human-level intelligence, he says, and 2045 will be the year when Singularity will be achieved.
However, my belief is that we do not need to wait until 2045, or even 2029—AI has already started taking over our world, but in a gradual and insidious way. It is not that humankind wakes up one morning in 2045, and finds that machines have taken over us. The way it is going to happen is the way it happened to the mythical frog in a vessel of slowly boiling water. The water heated up gradually, and the frog was lulled into warm comfort and then at another moment, the water and the frog were both boiling.
Ten years back, we remembered everyone’s phone numbers. Now, our phones remember them for us, and it is a rare human who remembers his spouse’s or child’s number. Even three years back, we used to remember routes, and how to get from Place A to Place B. If we did not, we opened a map and figured that out, using our intelligence. Or, if we were in India, we interacted with other human beings who told us how to get there. No longer.
Google and its ilk tell us exactly where to go, how to reach there, what time it will take and how much traffic we will encounter on the way. For some reason if the global positioning system (GPS) or the map does not work, we get frustrated and literally do not know how to proceed.
Google Now sits on the super-intelligent Android devices we call phones, reads our calendar and our emails and our texts, and then tells us where we are supposed to go next, and when we should leave. Every evening at 6, my phone pops up the directions and the time to leave for my badminton courts. If I am not in my home town that day, it knows, and tells me to go elsewhere instead. And I do.
Next, we will lose our very human ability to drive, to control our car-machine as per our whims. My phone will summon my autonomous car around 6pm, depending on traffic, and the car will take me to the courts. It will make sure that I am hydrated (my wristband will tell it so), the rear-view mirror will ensure I am dressed for playing, and this human vegetable will be deposited at exactly the right time to start his game. I would not be surprised that when I reach there, my AI robot clone will be the one that actually plays, but I digress.
Then there are intelligent homes, powered by AI and IoT (Internet of Things). The mixer has made my protein shake when I come home, the AC is on and the shower heated. My refrigerator and microwave have conferred with each other, and prepared my meal…you get the picture.
What if the machines we create are more intelligent than us, and they start creating other things that ensure their welfare, not ours?
Machines defeated us at chess—ask Garry Kasparov. Recently, the world Go champion was defeated by a machine; and Go is a very ‘human’ game, with more permutations and patterns than chess. The 2016 American elections, it is now widely believed, were won more by intelligent networks and bots on Facebook and Twitter, rather than by Donald Trump and the Republicans.
This is why Musk and Gates are worried, and someone like even Russian President Vladimir Putin has weighed in, saying that the country which controls AI will own the world. Musk has gone so far as to say that AI should be one thing that should be regulated by government and countries.
It is also quite clear I belong to the Musk camp; my faith in human beings making the right choices is currently at an all-time low. We need to be more educated and aware, and if required, we need regulation to come in. Otherwise, I fear, we will be the human-shaped frogs in the vessels we made, slowly getting cooked to death, by the forces we ourselves created to make us live forever.
Jaspreet Bindra is senior vice-president of digital transformation at Mahindra Group.
With artificial intelligence gradually taking over human life, it may not be too long before machine intelligence becomes more powerful than all human intelligence combined