The Aye and Why of AI

Pratima H

INDIA: When he advises to think of an ocean with a prudent dimension and not think of every wave as ‘the’ one to ride and run away from; you seriously doubt whether you are conversing with a surfing pro or a product-engineering veteran. But that’s how Hari Haran, President - Global Sales, Marketing and Business Lines, Sasken Technologies, absorbs trends and disruptions.

Speaking of which, there’s a lot that is defining new currents with the wind of Artificial Intelligence or AI. Many industries, willingly or otherwise, are changing tracks in a major way. Autonomous cars could be disrupting the global economy by 2035 and from what Forrester augurs, this will happen in many industries like automotive, shipping and logistics, insurance, government, media and entertainment and safety/security.

Gartner, too, feels that by 2020, AI will be a top five investment priority for more than 30 per cent of CIOs but at the same time analysts also warn of 'AI washing' wherein the AI label can be used a little too indiscriminately and which is already having real consequences for investment in the technology.

Let’s see if we can get a grip on AI despite these greasy corners. Haran deconstructs many such fears, many what-ifs, all that dystopia, the possibility of auto-makers getting blindsided with software’s onslaught, along with recent software-related grey area incidents, and some other issues in this ‘intelligent but not artificial’ chat.

Automation has acquired a new gravity altogether in the last two-three years. What’s new, unusual and exciting about it from where you are watching it?

There are a lot of interesting things happening in the industry. Some verticals, like automotives, in particular, have seen a lot of disruption. The value of a car has shifted from the mechanical aspects to user experience and cognitive elements.

That’s putting many traditional automotives under a lot of pressure who were so far good when it came to building brakes, transmissions, machines etc. Now the customer may not pay a premium for the machine however, since that’s the hygiene thing going forward. Sasken Technologies is all about connected experience so this disruption is a considerable opportunity. Also, when we look at factories, the machines so far have only been innovative in mechanical terms. Now so much software and electronics are coming to that realm too.

Software is driving the connected experience, and we, who have been enablers of products and product-makers, and have the entire stack from chips to services, are excited and observant of what’s changing all around us.

But what they are good at is their ‘core’. Why would auto-makers move over to software as the new ‘core’ and then let technology-experts take care of it? What remains for them? Even when they outsourced something, the pithy part stayed with them, so, aren’t the traditional folks disoriented and disturbed as they tackle these huge shifts?

There used to be a time when a lot would be spent inside the factory and that reflected in the price and the product. But today connectedness penetrates every point in the experience. Cutting costs is the only way to move on and survive. Things have changed a lot, specially in a global-sourcing market.

Automation needs a lot of connected-ness between things so that human is almost out of the picture. Automotives have been realizing the new dynamics but industrial segment is slightly slow in adopting change.

So automakers are not paranoid?

They are, but now car companies have started talking about software instead of the car-that-was. Like how Tesla is. It’s not easy to think like a software company though. The need for change is fast and doing it organically can be difficult.

Software and metal or the machine – are they like oil and water? Are car-companies able to weld in software in an organic and fluid way at all?

It is hard to do this in an organic manner. But these aspects are not like oil and water fundamentally, albeit they tend to appear that way in terms of the kind of people and work involved. Once you make the machine smart; you can keep adding layers, sensors, platforms etc. to make it more cognitive and sharp. Then things become as smooth and seamless as water.

Hari Haran, Sasken Technologies

How do they come back as leaders in a game that is now about software and not hardware?

Accept that you are not a software company to start with so you have to rely on partners. Audi, BMW etc are already partnering with so many players and companies like ours or Wipro, Tata Elxi etc are offering capabilities in their ways. The software has to be so much smarter than the next car company to make sure you have some lead as an automaker in the next age.

Cars and their adventures with software have also led us to goof-ups like software-cheats and emission-scandals that we have seen off and on. What’s your view?

It is one of those things where technology can be used in different ways depending on the intent therein. My take is that even if there is a cheating-software, there would be other piece of software that can catch this con-software. Cars can stop relying on software alone and use sensors and other credible, hard-to-tweak ways to ensure frauds do not happen.

The consequence of these incidents is that the industry is coming up with more smart answers and better-coping tactics.

That means that there is no room for dystopia with all this accelerated automation? Redundancy and the Fourth Revolution – should we be worried now or later?

These topics that we are discussing are tectonic shifts. The concept of ‘value migration’ is shifting to industries like ours. Many jobs might be lost in one industry with automation’s impact, but many jobs will be created in other industries simultaneously. The magnitude of this migration is in millions. Three new industries gain from the loss of one industry, which has to change fast to arrest losses. The ocean is too big to grasp for anyone.

Let’s talk about AI here from a hype-angle. How overrated is it and what’s the real formula to using it smartly?

Unfortunately the industry operates a certain way. You can sometime feel overwhelmed at the sight of a wave coming from far away but it may not have much momentum by the time it reaches you. You can only watch waves and ride well when you get a good one. But not all waves are scary or disruptive.

The ocean, as I said, is too big to time well. Some of the things we see (ex-Superphones) already have a bit of AI inside. Some things will happen with time. That said, AI is not all hype because it will be significant in a few years. Yet, not all opportunities about AI may materialise as the big waves we see them to be. There is a lot of guesswork at the moment.