TikTok, YouTube, Netflix: Recommender Engines

Making choices in the AI age

First published on my Substack.

In 2017, I had two choices: AI or Crypto.

There were 2 job offers in front of me. One was a software company that specialized in AI-driven decisioning, and one was a startup in the emerging field of cryptocurrency and was building something on the blockchain protocol.

I had yet to hear of one single enterprise company that made money in blockchain. As a wag put it, blockchain is “a very, very slow database”. That turned me off the opportunity.

On the other hand, AI was booming. IBM started plastering “Watson” on everything. (The irony that Watson was Sherlock’s sidekick and while a great documenter, was hardly the brains in 56 short stories and 4 novels, I will ignore for now!)

It was clear that AI was here to stay in every field. But what caught my attention was one thing in particular with practical implications: The Recommendation Engine, or Recommender systems.

What is a Recommendation Engine?

[Note: One caveat for all of these explanations is that you are hearing from a business guy, who grew up in tech. I’m not a data scientist.]