F2F #77: AI-natives will eat us alive

I have spent almost the entire week at Mobile World Congress / 4 Years From Now / Talent Arena. The three conferences that took place in Barcelona, each one centered around a specific subset of sectors and technologies, but the three of them converged around AI.

AI all the things ✨, much like in the last two years, with little to no added substance to stuff that's been stagnating for years: SaaS, VC, robotics and such. To avoid all the fluff, I've spent my time talking to great people and in sessions about coding.

The main consensus around the AI world is that juniors/entry level profiles are doomed. AI - they say - replaces the need for interns and lesser-experienced profiles because that's essentially what ChatGPT does. In terms of one of my best friends, ChatGPT is like "an army of interns with infinite time".

Up until now, we have used these AI tools for mostly mechanical and somewhat low-level tasks because it lacked depth (until Deep Research arrived) and taste. Now, with the advent of certain models that do research for you, reason (sort of) and go beyond the basic google queries, we're sort of there in the depth department.

As for the taste, that's something you acquire with seniority... or so they say. I won't agree with this, but I will concede that it's true in a vast majority of cases. As it's true that a hell of a lot of senior people do not have any kind of taste whatsoever.

For me, it boils down to training. I was part of the "digital native" or even "internet native" generation that was going to replace everyone in the job market. We didn't. The generation after mine, the "mobile native" didn't do it either.

We did have certain advantages because we knew our way around technology but we were limited to the gadgets that we grew up with. You had to also put effort into it when you wanted to hack things, and spend countless hours stuck because little to no documentation was available, and the internet back then was scarce and slow. Whenever I got stuck mounting my first desktop pc's, I had to ask for help to actual people, go back to the store or read the manual in languages I didn't comprehend.

If we built resilience back then it was because there was fuck all else to do, nothing worked on-demand and life had another pace, generally speaking. We could put in the 10000 hours to master something, but the reality is that very little people actually did it.

The mobile native generation had lower barriers of entry: cheaper devices, more widespread adoption of technology and trends, instant access to documentation & help but also infinite garbage to consume. Still, it produced lots of very valid people and lots of new markets and jobs were created: social media manager, youtuber, influencer, vlogger, etc. Still, if you really wanted to make it big, it required you to put in a lot of effort.

Enter the AI native generation. They will crush the market because of two reasons.

First, the rest of us are figuring out AI as we go and we will never use it in full. We will learn it and use it to the extend that it solves (some of) our problems and we're satisfied with the ration investment vs. return. We will acquire vices and very few of us will re-wire our brain to use AI like a native would.

Second, because when they will join the market, they will already have spent a considerable amount of time building their own agents, subagents and tooling to make their life easier. And the fact that it'll cost almost nothing to delegate any kind of task to an AI leaves no room for laziness. Even the laziest ones will want to automate their stuff - provided costs are negligible and this technology crosses the chasm.

But what about the taste? Will taste become finally irrelevant?

Nope. Probably that won't ever happen.

Taste changes its form and shape, but it does not disappear. When information is abundant and production gets cheaper, technical execution matters less on its own, so taste matters more as a filter: what to make, what to ignore, what to combine, what to leave unfinished.

What can become irrelevant is a certain kind of taste: inherited prestige, empty aesthetic snobbery, or gatekeeping disguised as discernment. But judgment under uncertainty remains valuable. In fact, the noisier the world gets, the more useful taste becomes. We won't be able to judge the new juniors with the same concept of taste we've always had.

While the rest of us will be stuck sending articles to chatgpt for a quick summary or asking it to translate stuff for us, they'll be replacing us in the blink of an eye.