What's worth learning now (in the era of LLMs)?

The topic of whether it’s too late to become a programmer, given the rapid development of AI, is very popular right now (and, I have to say, rather speculative). During one of our mentoring sessions, we also touched on this topic, but from a slightly broader angle: “What’s worth learning now in the era of LLMs?”

First off, it’s worth noting that learning anything new at any age is fundamentally a great idea — it, at the very least, creates new neural connections and supports some level of neuroplasticity (and let me just remind you that on average, we hit a neuro-plateau around age 21–24). So if you like something and feel like learning it, go for it, and don’t listen to any so-called experts.

But back to the question. From a strategic point of view, in my opinion, given the current trends, studying development makes sense either at the level of general concepts or very deeply — if you’re planning to become a high-level specialist in programming/deployment. In all other cases, I dare say you won’t bring any added value compared to a much more skilled AI.

But what is definitely worth learning is sales and marketing. And here’s why.

1. The way we present a product critically shapes how it’s perceived. It can be amazing on the inside, but packaging makes the first impression. Have you noticed that in some companies, there are employees with titles like “partnership manager”? Do you know what they do? Sales. So why not call them “sales managers”? Because “partner” sounds much more pleasant to clients than “salesperson,” and this approach can sometimes increase the response rate to cold emails by dozens of percentage points. And for exactly the same reason, food products are more likely to say “80% fat-free” than “contains 20% fat.”
2. Most of our work is sales, even if we don’t think of it that way and believe it doesn’t apply to us. Product managers sell the idea of the next cool feature, the tech lead sells the importance of refactoring and addressing tech debt instead of allocating 100% of resources to those very features, designers sell their vision of a great user experience. And so on — you get the idea. Everything around us is an attempt to negotiate and sell something.

In neither of these areas can AI (at least for now) not only compete with humans but even deliver a decent autonomous result. Yes, modern models can generate good creatives, even run scripted calls (or send emails) using synthesized speech. But inventing meaning, handling objections, reading emotions and using them in negotiations, skillfully manoeuvring through risky situations and adapting — these are all still the domain of human expertise.

So, if we want to remain useful to AI (🥲) in the tech industry, I’d strategically be investing in these very skills.