Humans - The new programmable Robot

What does this even mean? How can humans be programmed? And how can we be robots? To answer this question, we need to look into the major the building block of this - Artificial Intelligence.

A recent TCS Global trend study on AI reveals that investments in AI is increasing and will continue to increase across the industries and across geographies. So apart from being smarter in order to give better user experience and also push the revenue/margins graph up, could there be other implications to this increased investments into AI?

Human minds are programmable - of course! We develop our thoughts and views based on what we read, see, hear and experience. Companies already use techniques to influence us to buy their products and services. They track our internet browsing history, our preferences, shopping history, social media activities, etc. to reach out to their targeted prospects. Others like politicians also program us but probably at a more macro level with very powerful stories that touch the chords of voters. So this is not new. But what if we can be programmed individually to a level which we have never imagined. What if the machines around us start to know things about us that probably no other human knows about us? With such data, we could be an open book with little privacy - moving us closer to a robot - who is also programmable and also have no private data.

"Hello, Human World!"

Rowan Trollope (SVP, Cisco) in his amazing keynote at Cisco Live 2018 Barcelona referred Google's head futurist, Ray Kurzweil's prediction that 2045 will be the year of the point of singularity - which means this is when AI will become smarter than humans. Looking closer to the trends around, it is getting obvious that there are several developments leading to the point of singularity.

There is little debate that machines are getting smarter by the day. Though there have been a period of romance with AI multiple times since 1970s, experts say this time it is for real. The reason being the sheer large amount of computing power and data availability as well as advances in algorithms. The AI techniques such as Deep learning and Neural networks have created a variety of new-age applications such as NLP based digital assistants, Self Driving cars, Self healing systems, Smart Chatbots, Selfie drones, Advanced robotic systems etc. A decade ago for a common person this was science fiction. Today it is science reality.

But for AI to be able to program us in the true sense, it needs two things - Data and Method. It needs to have massive and invasive data points about individual humans and also need a method of programming humans. Fortunately or Unfortunately we are moving towards enablement of both of these.

Data

A 2017 study from Stanford university reveals that using deep neural networks on facial images, they were able to predict the sexual orientation of a human with up to 91% accuracy! The professor leading the study has also reportedly stated that AI will be able to judge the IQ or Political inclination of humans by studying the human face.

It is not just the face, but technologies are being developed to read thoughts from our mind directly. In a recent research by Carnegie Mellon University scientists, they were able to read complex human thoughts by using brain activation patterns. Though the focus of this kind of research is usually to make Telepathy an everyday reality, like any other technology it can always be used adversely. Similar research of mind reading is being done for healthcare purpose by a startup called OPENWATER led by Dr. Mary Lou Jepsen, a former executive at Facebook, Google and Intel. Apart from solving certain healthcare issues, they also claim that their technology will help to stop AI from taking our jobs! But again, there are always two side to any technology.

One thing is clear - whether it is reading our face or our mind, it will lead to reduced privacy. Similar to what social media has done to us by making us addicted to it and in turn making us give away a lot of our private lives to the world. Now with such invasive data in hand, the only thing required to make us Robots is a method of programming.

Method

One innovative method of Human programming is described by Mikko Alasaarela as "The Human API". Many companies already use this approach, but he has very creatively described it as API to Human brains. He talks about how in a very scientific way AI can use methods to engage us, extract data from us and make us do tasks that it wants us to do.

Another interesting story here is the announcement from Elon Musk last year about his new venture called Nearalink. Neuralink is into developing ultra high bandwidth brain-machine interfaces to connect humans and computers. Though their stated aim is to have a "safe interface", it is an interface between machine and humans - A possible method?

In any case if AI does get smarter than us as mentioned earlier in the article, I am sure it can find new ways to program us as we have found ways to create AI.

Final note....

As we move towards being a robot and an AI robot starts to get as smart as humans, it is possible that we are moving towards a convergence. Elon's Neuralink is possibly the one of the closest attempt to this convergence.