In an interview with Axios, Anthropic CEO Dario Amodei said that AI could “wipe out” as much as half of all entry-level white-collar jobs. Of course, I wouldn’t expect any other statement from an AI company’s CEO, but still this immediately brings up a question — where the experienced people will come from, if the entry-level jobs are getting destroyed?

There are plenty more questions that are triggered by such a statement — if we remove these people from the economy, who’s going to consume all those AI resources? In the past, during all kinds of industrial revolutions, people would just shift to do some similar “simpler” tasks, but will all white-collar jobs now shift to prompt engineering? Having all these people unemployed or low-income is going to do wonders to the economy that is based on consumption.

But I’m honestly concerned about something else entirely. I’m worried that with AI everywhere we are entering the “age of mediocrity”, or “the age of ‘good enough’”, if you prefer a more PC term. AI is really good as a modifier or a replacement for a lot of tasks. AI is really good at churning out endless volumes of content, or decisions, or designs that are interesting or “good enough” — but where the breakthroughs will come from? As someone with extensive product management experience, I already can see AI replacing me in a lot of things — quickly drafting a user story, mocking up some UI concepts (and even producing code for it), or even generating a PRD document. And all of this is going to be generated based on already existing content, regurgitating the same old concepts, endlessly mixing and remixing them.

I’m not saying that AI itself lacks brilliance, but it seems to me that we are training it to be good enough, to deliver something satisfactory, something “right in the middle”. Something that will immediately get consensus, won’t trigger arguments or discussions. I still vividly remember some heated arguments with UX designers regarding some button placement or app navigation concepts, when trying to find something that we thought would be great for our users. (Whether it was or wasn’t — it’s another question, and a subject for A/B tests, lol). But making some bold decisions, taking some risks, producing something creative or original — I feel like AI does not go there, because it’s not trained to do so.

I hope people at OpenAI, Anthropic and other AI companies are asking themselves — are we building a future that will bring innovation, or just automating ourselves into an endless sea of beige?