Did the statement invoke a sense of Déjà vu? Maybe a decade ago, it was “there’s an app for that”. The app store (iOS) launched 11 years ago with just 500 apps. The app store though was revolutionary and enabled a completely different set of business models, enabling entrepreneurs to build apps for everything – transport, food delivery, email, home services, expense tracking, navigation, music, photos, banking…you name it. Uber and Airbnb probably wouldn’t even exist without smartphones and apps. Starting from meagre 500 apps in 2008, there are now over 100,000 apps in just the iOS app store – now there really is an app for everything.
There’s now another technology growing exponentially, and I think it’s far more disruptive – Artificial Intelligence (AI). I have been tracking improvements and use cases scenarios in AI for a couple of years now – and the growth rate is exponential. In the last year or so, I have seen AI being deployed in writing fiction (Open AI’s Text Generator), medical diagnosis (India hospital uses AI for medical diagnosis), recruitment (AI for Recruitment), predictive retail analytics (How AI will change retail analytics), music creation (Open AI’s Muse Net), art (Indians using AI to create Art), Call center software (Google Duplex), modulating AI voice audio (Dasha AI), (deep-fake) video creation(!Samsung’s AI for Video), video editing (Change what people say), drug discovery(Deepmind’s protein folding algorithm), learning (Third Space Learning,), Knitting (MIT Research on AI driven Knitting Software), creating whiskey blends (AI Whiskey), facial recognition for law enforcement (Amazon’s facial recognition for law enforcement). Apparently, AI from Palantir (Mark Bowden’s the Finish) even played an important role in locating Osama Bin Laden by analyzing tons of disparate intelligence inputs. And I haven’t even mentioned the obvious AI applications like voice assistants, self-driving cars, image recognition in photos (Google Photos) or Tik Tok (AI is the Product by Connie Chan)
There is however a critical difference between growth of the app economy and AI. The first is that while apps (or other technology interventions in general) made executing a task easier or faster or cheaper, AI may just eliminate need for a human to do that task. It’s similar to how Steve Jobs thought about computers as the ‘bicycle for the mind’. Bicycles made it easier and faster to get from point A to point B, but it was still the same person who, instead of walking, pedaled from point A to point B. Similarly, one could take a taxi to work before Uber or Ola– but Uber/Ola made it significantly easier to find one (and pay for it). But there is still a taxi and there is still a taxi driver. With an AI driven self-driving car however, there will not be a driver anymore. The justification provided for AI is similar to apps– it will make the task easier, faster or more convenient and in a lot of cases that’s true. The Madurai hospital which uses AI to diagnose diabetic blindness will do the diagnosis faster and more accurately thus enabling the hospital to serve more patients. A self-driving car will surely have a far less accidents. Also, the official explanation from companies building, deploying or investing in AI is that it will free up people to focus on more creative and fulfilling jobs. But honestly, I am having a hard time imagining how thousands or hundreds of thousands of people can in a short time (AI improvement is exponential) start moving to ‘creative’ jobs.
So, while the statement ‘there is an app for that’ had happy consequences, ‘there is AI for that’ is scary.