I am often asked to articulate some problems that people can start working on immediately and get some real world industry experience in artificial intelligence. The hard part for these situations is that the problem or a part of it should be meaningful to people with every level of technological sophistication. So here is an attempt to articulate such a problem for Natural Language Processing.

Consider this Wikipedia based data set (40MB), which is built by the SQUAD team at Stanford. Try to solve the following problems for the text contained in these articles. …

Ada Lovelace spent more than a year thinking about the world’s first program. She published it in 1843. Since then, there have been many definitions of software and many characteristics that industry has generally accepted for good software — it should complete a specific task, it should not be fallible to small errors, it should be debuggable and maintainable, it should have clear metrics that track its success, among others.

So far, no one has bothered to say that in addition to all other such characteristics, a good software should exhibit deterministic behavior. Determinism has evolved as a philosophical concept…

This is not a political writeup, but an enumeration of the factors that may entrepreneurs, businesses and individuals may find useful as they think about the next decade.

Joe Biden will take over as the President of United States in 2021 after 80 million people voted for him. This is unprecedented in US history, so are the 74 million votes that his opponent received. A central issue in the divisive election was the K-shaped recovery from pandemic-induced economic hardships in 2020. …

AI has been top of mind of the startup world for at least five years. Hundreds of billions have been invested, thousands of entrepreneurs and intrapreneurs — some of the best minds of our generation — have tried their hands with varying degrees of success, and none of the pure AI companies has made it to IPO.

The Problem

We all have learnt, some the hard way, that artificial intelligence cannot be packaged into a product. The factors that scale a typical venture backed company, say in SaaS, don’t apply to AI.

1. First, to be useful an AI must train on data…

The world of real estate is often thought to be a slow adopter of technology, relying more on handshakes and glib negotiations. Increasingly, this view is turning out to be a myth. Artificial intelligence is making the world of real estate to be more nimble and more efficient. Consider these five applications below:

1. Lower Vacancies and Costs

Netflix uses AI to suggest what you should watch next. It predicts your tastes and lines up a collection that it thinks you will like. Amazon uses similar technology, called a recommendation engine, to suggest that you may like to buy an ottoman after you bought a…

No data science team is the same, which, by itself is not a problem. The trouble is that there is no real way to judge whether the role you are joining is going to catapult you for success or turn out to be the biggest mistake of your career. A data science job can really surprise you when you join it.

In this article, I have suggested some strategies to avoid these surprises, but let’s set the context and the extremes first.

The Four Roles

The biggest misconception out there is that there is one, homogenous, data science role. In truth, four completely…

One of the key tasks of a product manager is to keep track of multiple workstreams related to their product or feature. A project manager, a division manager and many other business managers find themselves in similar situations from time to time.

The Old Way

A common tool to track progress of all these workstreams is to use status indicators like the traffic signals. Green means everything is good. Yellow typically means that things are on the borderline and Red means things are alarmingly behind. Different teams and even different individuals have different thresholds about when to change the color.

In my opinion…

While there are so many new advances in Natural Language Processing (NLP) making headlines these days, in truth artificial intelligence is far from the point it can carry rich human conversations. There are many challenges, but a big one is the ability to use lies effectively. Consider these three scenarios:

1. Understanding Lies

When humans communicate using language the literal meaning of the text is often different from the intended meaning. If someone says “Praful is not the brightest bulb,” you understand that I am not an electric appliance and that the speaker does not hold my intellectual prowess in high regard. …