AI Superpowers!

What values do you think AI will bring to society? And how do you think they will change human life?

Rise of Artificial Intelligence is eminent. Thanks to advances in algorithms, computing power and explosion of digital data, it is becoming possible for machines to learn patterns across various types of data and carry out higher order tasks. Traditional machine learning has been around for quite some time, but it is the advent of deep learning, is beginning to revolutionize the world of computing as we know. With traditional machine learning, a set of rules (features) are needed to be defined by humans, in order for machines to infer patterns and predictions. Whereas, using deep learning, machines can infer a set of features associated with a certain objective, ultimately leading to inferring the objective itself. In other words, AI is now embarking on unstructured data analysis and autonomous decision making.

AI is forcing us to rethink existing norms with a new mindset. The impact is groundbreaking. Experts are comparing the emergence of AI to the time when steam or electricity was invented, leading to disruption in all aspects of the society today, starting with technology, businesses, labor markets, governments and so on.

When it comes to impact on businesses, one can speculate a range of outcomes based on the learnings we gathered from similar transformations in the history.

Businesses which upgrade themselves with the new paradigm of AI, will survive and will mostly like thrive.
Businesses which will ignore this paradigm will undoubtedly perish.
New businesses and new markets, which were never possible, will be created. E.g. Past “Mobile” paradigm created several new business giants such as Uber, Twitter, Instagram and so on. These were never possible without the rise of smartphones.

Human lives will get impacted in several ways.

  • Favorable: AI will enable new type of products and services, and that will bring in change in lifestyle in an unprecedented way. Just think about self-driving cars.
  • Not-so-favorable: AI will bring in disruption in to the job market and to the workforce. We have passed beyond debating “if”, but the question in front of us is, “how” it will impact labor market in the new economy powered by AI.
  • Unknown: And there is the third possible way AI will impact our society. We just can’t name it yet. Let’s just call it “unknown” impact. For example, AI getting consciousness, new weapons created by AI, and other random possibilities. For example, government and world orders got changed due to the invention of Nuclear technology. Any one of the above mentioned possibilities can lead to fundamentally changing human lives and beyond.
The more technology advances, the more complex society becomes. Ironically, I think that the solution to this complexity is, after all, technology. How can your field of expertise be used to solve complex problems?

As technology progresses, so do human consumption patterns. Progress leads to advance lifestyle, resulting in the “first world problems”. These are the problems we humans have created for ourselves while we were making progress with help of technology.

It is exciting that AI can solve our first world problems. At the same time, I am equally excited about the opportunity for AI to solve fundamentally impacting

unsolved problems in the world. Here are some of the examples:

  • Solving water crisis: AI-powered image analysis augmented with IoT technologies can help optimize existing water usage in a way that was not possible before.
  • Fight against diseases: AI researchers are able to generate novel molecules using recurrent neural networks. A specific type of creative AI should bring more diversity to the lead generation process, and ultimately, AI could substantially improve the work of human medicinal chemists.
  • Animal/forest protection: A new AI system that enables surveillance drones to automatically detect both humans and animals could help conservation experts and rangers protect endangered wildlife starting in 2018.

As the advances in AI continues, we will see more and more use cases where AI will help solve fundamental problems.

And then there are opportunities for AI to influence advanced “first world problems” that will ultimately simplify our day-to-day life. It can range from our shopping behavior, to our transportation, to our entertainment. Some of the industries we will see impact

  • Commerce: A new type of commerce will emerge thanks to AI’s ability to comprehend unstructured data.
  • Transportation: Transportation will get a new face based on self-driving technology.
  • Lifestyle: AI assistants will get wiser and wiser. It has already began with Alexa, Siri and Google Home.
  • Finance: AI can not only help with better financial systems and options, it will also help us make it secure with advanced ability to detect fraud.
  • Art & Entertainment: A new type of Generative AI is already on the verge of producing art, music and videos which are beyond distinguishable level by average consumers.
Diverse technologies and services including your field of expertise have emerged as next-generation growth engines. Among these, what technologies and services will change society the most rapidly? Also, what will have the biggest impact when they enter our life – big enough to cause the existing order to collapse?

I am particularly excited about impact of AI on Commerce. Traditional machine learning has been in use in the world of commerce since quite some time now. It allowed companies to build linear systems operating on structured data, enabling inventory analysis, ad targeting and so on. But still needed human intervention, in one way or the other, to carry out these tasks.

What AI/Deep learning is enabling us are the following fundamental capabilities:

Credit to Frank Chen from A16Z for the original thoughts.

These new capabilities are the missing piece in technology stack that will lead to what we call “self-driving commerce”.  

  • AI can help listen to consumer opinions at unprecedented scale disrupting the way products are made and marketed. And this has been the first mission for our company, Commerce.AI
  • AI will bring in computer vision at higher order enabling experiences such as Amazon Go.
  • Rapid advances in IoT/combined with AI will redefine logistics and supply chain
  • Self-driving robots/vehicles will finally solve last-mile delivery automation challenges
  • AI/Chatbots will assist us not only for customer service but also for our purchasing decisions
  • In-store robots powered by AI vision, can do quick inventory and assist customers
  • Traditional ecommerce will be disrupted by AI powered search and recommendations.
  • Product manufacturing will change too, thanks to AI powered smart factories.
You have read complex environments and have made decisions. In this process, have you ever run into a failure in terms of decision-making? If you have, what are the causes of decision-making failures, and how have you gotten over these failures? What were the biggest challenges that you had to overcome in the process of your company’s growth? Among these, what can be said to be failures, and how did you overcome them?

Before founding the current startup, Commerce.AI, I had founded another startup called Fitiquette. That was my first time as a tech entrepreneur. During the course of building that startup, I have gone through several small successes and failures, which I refer as “learning opportunities”. Eventually, Fitiquette had a successful exit. I learned some of the important lessons throughout our journey. I have been sharing my learnings in various conferences and panel discussions. Here are some of the thoughts associated with decision making:

  • As an entrepreneur, you have to take many decisions. Right from registering the company, to partnering with right co-founders, to building the right product for the right market. It continues, as to raising funding, to exiting the company.
  • Key is to identify which decisions lead to binary identifiable failures and which ones lead to slow degradation. Let me explain, Some of the decisions can not be undone. Some can be changed but with high cost and some can be changed without significant impact.
  • Type A decision: Decisions that have binary outcomes are great. You can find out what went wrong and take corrective measurements. So something like product failures, raising capital etc are somewhat binary in nature. Either your product works or it doesn’t. Either you raise funding or you don’t. Even if the failures are not obvious initially, they become obvious relatively quickly.
  • Type B decisions: The decisions that lead to low impact outcome are the ones that concerns me most. These are the decisions which result into mediocre results. In these cases, it’s hard to pinpoint that something is going wrong and associate that to a particular decision.
  • For us, a Type B example leading to a slow failure was experienced right in the beginning. It was about choosing the name for the company. It sounded like a trivial decision while registering the company. But it was one of those with long lasting impact. Initially we had named our company Zmorph, primarily because of the domain name availability. But the name was geeky and was not rhyming with our fashion-savvy target customers. It took long time for us to realize that. It was not at all obvious that the ‘branding’ was the problem. In retrospect, early lack of interest in our product was largely due to our name and the brand. When realized that and decided to rename our company, we had to go through an expensive and time consuming rebranding process, almost like pressing a reset switch. As soon as we relaunched it, we could immediately see the positive impact. It was the same product offering underneath and the same target customers. We started receiving significant PR coverage and our name started spreading.
How would you define failure? And what do you think are key drivers to overcoming failure?

To summarize my experience as an entrepreneur in tech startup, it is an interesting reflection to think about “Failures”. More I think about it, more glaringly it becomes obvious that failure is not defined by an immediate outcome after a certain decision is taken. It is defined by our ability to identify what’s wrong in a given situation, our ability to associate that with a particular decision and courage to take corrective action.

It is like navigating a large ship. There are several decisions a captain takes throughout the journey. Even though the ship goes through several ups and downs due to external circumstances, a successful captain always has her/his aim set on the final destination and has ability to measure and assess deviations from the ideal path, course-correct and navigate it to the ultimate destination.

We as a society, tend to get attracted to stories about overcoming failures and personal heroism associated with it. But if we dig deeper in each case, they all have similar patterns underneath.

  • Successful teams in such situations have great tools in their arsenal.
  • They all have great mechanisms to listen to advisors (customers in the case of businesses).
  • A combination of these has helped the people in such situations identify failures and track them down to a specific decision or a set of decisions.

Which brings us back to our earlier discussion about AI. I truly believe that AI is enabling such tools to help identify patterns about successes and failures and help humans to make decisions leading to measurable successes. Our goal at Commerce.AI, is to build the AI that can listen to consumer opinions in various forms, at massive scale, and in a truly real-time fashion. We believe a new era is beginning in which AI will help businesses with all sorts of decisions through their journey, and has a true potential to be the force that will shape the society in the coming century.