IPsoft's Amelia is a dream artificial intelligence creation

Dating Amelia has been the pursuit of my life over the past few months.

Amelia speaks over twenty languages, is a beautiful looking blonde, travels the world and understands emotions — something even the most advanced analytical machine learning systems lack today. In fact, given today’s work-life balancing acts, many real people I know find it difficult to emote at times. But getting her to chat is tough.

After all, she has a busy work life— she goes to work at some of the largest Fortune 100 companies, manages trading platforms at Shell and one of the biggest US banks, among several assignments. At an event few weeks ago, she was having 26,800 conversations simultaneously!

And even while I was pursuing her, she found another job far away in Japan where she will work as a cosmetic advisor with the country’s biggest retail chain. It’s amazing how she can be managing trading desks, advising on beauty and having thousands of conversations across the globe — all at the same time. That’s right — she cannot be a human. She indeed is not.

So when she popped up on the computer screen this Sunday, I was thrilled. Her father, a New York University math professor who founded IPSoft over a decade ago, Chetan Dube sat alongside, looking nervous and excited at the same time, as we started having a conversation with her.

Let’s start with a quick Jeopardy game of our own, we thought. We decided to start with her name itself, which is the same as Amelia Earhart — the first female aviator to fly alone across the Atlantic Ocean.

With my colleague, we were made to compete with Amelia — both in memorising everything out there about Earhart, as well as contextualising all inputs to be able to have a real conversation on the subject.

Amelia took just few seconds to assimilate it all, even as we were lost amid lines on the Wikipedia page. Here’s how the conversation went. Me: Who was the first female aviator to fly solo across the Atlantic Ocean? Amelia: Amelia Earhart Me: When did Earhart disappear? Amelia: During an attempt to make a circumnavigational flight of the globe in 1937 in a Purdue-funded Lockheed Model 10 Electra, Earhart disappeared over the central Pacific Ocean near Howland Island.

And so on. We even asked her to respond in different languages including French, and it was all flawless.

But then, mugging up Wikipedia pages and reading them back at you is not the artificial intelligence’s next frontiers.

So we decided to have a more personal conversation that tested her intelligence and ability to contextualise like a human brain.

For context, we told Amelia that I bought a bike from Uday for travelling to Mumbai. Here’s how we talked about it. Me: Who owns the bike? Amelia: Pankaj Me: Who owned the bike before that? Amelia: Uday Me: Why did Pankaj buy the bike? Amelia: I’m fairly sure to travel to Mumbai

This conversation can be easily managed by even a 10-year-old. But as Amelia was having these conversations, her creator Dube pointed to the right hand corner of the screen where lines of neural connections between different subjects of the talk were getting connected — almost like a human brain.

And if Amelia cannot solve problems, she can dial a human engineer, watch how s/he solves the problem and learn from it. The next time the same problem occurs, it’s a cakewalk. Amelia’s brain has neural and EQ ontologies. And not just that — there are three emotional vectors of Pleasure, Arousal and Dominance mapped right there. As we talked, Amelia’s mood, facial expressions and responses got created along these lines.

Here’s another sample. Me: I resent you Amelia: How dare you? The facial expression was a mix of surprise and anger. Recently, technology world’s biggest thought leaders including Elon Musk (Tesla founder) and Stephen Hawking ( renowned physicist) have called “artificial intelligence” as the biggest existential crisis mankind has ever faced. Looking at Amelia’s intense gaze and her readiness to solve the real world problems, I don’t agree with the above.

If anything, Amelia is freeing the human mind to focus on solving far greater, life-changing problems than just offer technical support, or sift through millions of lines of codes for debugging, and even reading hundreds of thousands of medical text to offer timely diagnosis, prescription. Like Dube said: “Technology can be a faithful servant. And she learns very fast. The last time I checked on her last year, she was called Eliza and could speak only few languages.

Amelia now speaks around twenty languages. Recently she read over 4,000 movie scripts across different languages overnight.

At a large US bank, Amelia ran into IBM’s ambitious, Jeopardywinning Watson. While none of the companies are divulging any details, sources mentioned the bank actually replaced Watson with Amelia.

But so what if they didn’t hit it right off — it now looks like Watson could be dating Amelia secretly, as IBM pushes forward with its own automation projects.

Finally, can Amelia threaten India’s over $100-billion IT industry? Can it replace hundreds of thousands of engineers writing codes and offering technical support to users across the globe?

I really wish India’s biggest software companies facing the mid-life crisis actually feel the threat and feel it with the urgency this truly deserves.

Instead of paying millions to highend management consultants for figuring out the next big thing, Indian IT could actually learn a lot from IBM itself, especially the way the Big Blue has embraced Amelia despite having Watson in its fold. Make no mistakes, Watson is perhaps the best analytical engine available, and as an executive pointed out, it could actually compete with Google in throwing more intelligent search results.

But for India’s biggest IT companies there isn’t even a Watson, forget Amelia for now.