Distinguishing among the 50 shades of artificial intelligence

Just a few years ago, the craze was to have a global delivery centre in India, either through an outsourcing relationship with an IT service provider or by tapping directly into the technology labour pool in India. This was starkly obvious to me when we hosted more than one Western client at an industry or service provider event—the first metrics they gauged each other by while sizing up where each stood on the totem pole were usually: “number of people in India" and “number of trips taken to India".

Today, this chatter has moved on to topics around automation, artificial intelligence (AI), machine learning (ML), deep learning (DL) and blockchain. The irony is that many of these disciplines are actually quite old, but they have only now become fashionable catchphrases. Automation and AI have been around for decades, and the concepts behind blockchain, ML and DL were defined at least 10 years ago.

The more these catchphrases become prevalent, the more difficult it is to distinguish one contender from another. This provides a problem both for buyers of this technology as well as for those who would like to invest in start-ups or other firms that profess to have made great strides in these technology sets. No one really has a way of distinguishing the truth from the noise, and very soon a sub-genre of consultants and analysts spring up to capitalize on this “information asymmetry", where buyers know less about a subject than sellers.

Information asymmetry is manna from heaven for people like me, who make a living out of being slightly better informed than the people we are advising and express a strong opinion about technology trends and issues, backed with some facts of course, and tried and tested analytical and due-diligence methods.

I spoke recently with Kashyap Kompella, chief executive of a firm called rpa2ai. He works with Prabhash Thakur, an adviser to the firm. Kompella and Thakur were both colleagues of mine when we were trying to capitalize on the information asymmetry around offshore outsourcing that existed in the mid-2000s. They have now come together to create a model to assess firms who profess to have skills in automation or AI, or who are building products in these realms. Their clientele mainly comes from non-technical investors who are making bets on technology, since such investment bets aren’t being made by technology companies alone. Automation and AI talent is scarce even inside the industry for those who are actually building solutions, let alone among non-technical investors and acquirers.

According to rpa2ai, the rebranding of old products and existing solutions as AI and ML is rampant. It finds that more than two dozen tangential technologies are being bandied about as AI and in many instances these AI components are not core to the product. The duo at rpa2ai says that they have seen products going back two decades being marketed as “deep learning" when, according to them, the field is itself not even a decade old. The duo has created a 25-point analysis of AI products which they use to look under the hood to assess how effective they might eventually turn out to be. Among these, they look for:

• The quality of data assets and data processing methods

• The ML models used

• The appropriateness to the business problem at hand

• Current approaches being used and how ML might change them

• The security, technical and regulatory risks

• And last, the knowledge, skills and vision of the team managing the firm that is looking for investment

Analysts and consultants will keep their jobs in the digital age. As humans, we will always need other humans to guide us in a structured fashion when we are on the wrong side of an information divide.

Siddharth Pai is a world-renowned technology consultant who has personally led over $20 billion in complex first-of-a-kind outsourcing transactions.