Sramana Mitra: Where do you think are the opportunities of building really interesting ventures right now? What’s your investment thesis?
Shomit Ghose: We can talk a long time about this. Hardware, software, and bandwidth are all commoditized. You can’t make money by selling any of those. For the past past 10 years, our investment thesis is all about the data. We continue to focus in areas that are based on extracting the semantics of the data and particularly in using data for business.
The future that lies before us is really vast. The companies that are most interesting to us are companies that are focused on Big Data. One company deals with Big Data for security purposes – for continuous authentication. The other one is in medical informatics to help physicians in not only things like prescription but also tests. They employ data for prescribing the correct tests. It’s the same thing for pharmaceuticals.
There’re about 15,000 prescribe-able pharmaceuticals. Physicians, being human, have no way of knowing the specifics of those 15,000. They end up doing the same thing you and I do – reading the back of the bottle. With pharmaceutical prescriptions, one size doesn’t fit all. There are teams out there that are employing Big Data for pharmacology. There is no end to opportunities here.
Sramana Mitra: Do you watch Fareed Zakaria’s Sunday morning GPS show?
Shomit Ghose: I’ve seen it in passing.
Sramana Mitra: It’s a very good show. Last week, he did the whole show on innovation. One of his guests was Vinod Khosla. Vinod made a comment which I actually agree with which segues into medical informatics and even how doctors diagnose and decide on treatment.
If you apply Big Data on patient scenarios, at any given scenario, the software is much more capable of coming up with better solutions, prognosis, and forecast than a human doctor. Vinod’s thesis was that doctors will be replaced by artificial intelligence. I happen to agree with that.
Shomit Ghose: I happen to agree with that as well. Perhaps where physicians will still play a key role is human beings react quite well to the moral and emotional support that another individual gives them. It might be that the physician is delivering the diagnosis and the therapeutic treatment plan, but the diagnosis is formed, to a large part, by something like Watson or an AI-based approach.
Sramana Mitra: Which then leads to the next big leap into futuristic discussions. How far is Big Data, expert systems, and AI in general going to go? Where is it taking us? This week, the media is abuzz about Stephen Hawking’s comment that if we let AI go to its full potential, the species is going to go extinct.
He’s probably not talking about the next 20-year to 40-year timeframe but we are probably within striking distance of extraordinary levels of automation. This renders human beings with no vocation. Maybe there’s 0.1% of human beings who have something to do and then the rest of the people have nothing to do. Where are we going?
Shomit Ghose: That is a great technological and sociological question. I don’t pretend to be wise enough to know the answers. I find the topic totally engaging but I haven’t read enough about it. I think there’s a real danger there. Similar to the Terminator movie, you certainly could have a malevolent artificial intelligence. Maybe that’s a threat at some point.
The bigger threat is artificial intelligence inadvertently harming us. Off the top of my head, if you had an AI system and asked it to do something as benign as optimizing the produce of a wheat field, You teach the system what to do to make sure that the wheat field is as productive as possible.
One morning, the AI system detects that there’s a family having a picnic on the wheat field. What is it going to do? The family is no different than an invading locust as far as the AI is concerned while human beings wouldn’t do anything like that. That might be the bigger threat – the unintended consequences.