Making Friends with AI

How AI is Evolving

A prediction for the next decade

Cassie Kozyrkov
3 min readApr 27, 2023

--

In the previous installment, I explained that because AI product designers of the past decade intentionally smoothed out the user experience (UX) to ensure that users wouldn’t notice when they interacted with AI components, it doesn’t make sense that users were the intended audience for all the buzz about AI back then (and there was a lot of buzz). All that AI talk wasn’t aimed at users or the general public.

It was aimed at the builders of AI systems.

That’s where you’d have found the trendiest AI frenzies of the past decade: with applied AI professionals and those who funded them.

Image created with Midjourney by the author.

Two decades ago: 2003–2013

Reading this, you’re in danger of thinking that AI buzz has its origins in the previous decade, so let’s wind the clock back another ten years. There was buzz in those days too. Except the audience was different. What kind of people would you have found nearest to the new millennium epicenter of AI excitement?

Researchers.

Two decades ago, most of the hollering in the AI space involved aspirational methodology — theory that hadn’t been put into practice yet — because two components were still missing: huge datasets and…

--

--

Cassie Kozyrkov

Chief Decision Scientist, Google. ❤️ Stats, ML/AI, data, puns, art, theatre, decision science. All views are my own. twitter.com/quaesita