Image generated with ChatGPT We have been reading pessimistic articles and expressions about AI for a long time. It will take our jobs, make us less human, eradicate us, etc.
The other day I came across this reflection:
Once you see Claude or some other LLM perform a task that used to take you three hours in just three minutes, all the work starts to lose its purpose.
All the years of experience you accumulated in your field suddenly seem useless.
Maybe they can completely replace you. Maybe they can hire someone at 50% of the cost to do the same task with an AI.
What struck me most while reading it is the dependency relationship between knowledge and effort. Throughout my career, one of the most rewarding feelings has always been to notice that something that initially took me ages now flows smoothly, almost automatically. That’s what I call expertise.
And although AI is one of the most differentiating tools I have used, it is not the first time that a technology has significantly reduced the effort of my routine tasks and, consequently, the time I dedicated to them. Code editors, syntax checkers, frameworks, third-party libraries, search engines, Stack Overflow… all of that has saved us thousands of hours and allowed us to truly focus on tasks that added much more value.
That doesn’t mean there isn’t value destruction along the way. If someone was getting paid for a task that needed three hours and now can be reasonably well solved in three minutes, there is a real economic problem. Many skills will be worth less and less. But I’m not sure that means all the accumulated knowledge to acquire them will also lose its value.
A programmer may lose some of their value for being able to produce five hundred lines of code a day. But twenty years of programming can still be differential in knowing what needs to be built, what seemingly innocuous decision will penalize us in six months, what requirement makes no sense, or when the AI is generating garbage.
What if what is becoming cheaper is not knowledge but some of its execution forms?
Now, with AI, all that time I’m freeing up is allowing me to tackle projects that once seemed unattainable. And I’m not just talking about professional projects. For example, I’m giving free rein to something that has always concerned me and that I hadn’t managed to channel until now: civic action. Monitoring institutions, gathering information, cross-referencing it, and building tools that allow for systematic action.
I’m also managing to revive literary projects that had been shelved for years, not just as a writing assistant—I enjoy writing on my own too much—but as a market analyst and as a rudimentary literary agent.
These are projects I always envisioned but didn’t have the time to execute. Perhaps we are falling short when we talk about productivity. We tend to think about doing more of the same in less time: more code, more reports (that only the AI reads now), more product units. More, more, more, but without a why.
But freeing up time also allows for something more interesting: experimenting, trying everything we had left on the back burner because it didn’t seem profitable or simply because we couldn’t tackle it.
I’ve said it before: I’m quite bad at anticipating the future. Which doesn’t mean I don’t have, from time to time, a good idea that comes at the right moment. But the merit, I believe, is the combination of tenacity and a hyperactive mind that is always scheming. Once in a while, the flute plays.
I don’t know if these cyberpunk predictions will materialize in the medium or long term, although it’s normal that they end up not being exactly what was imagined. If we look back, we can find dozens of examples where disasters were predicted, and in the end, the scope was not so drastic. Also, of course, technologies whose negative effects were underestimated.
In any case, it doesn’t seem very likely that we will all become Luddites and destroy technology. Renouncing it individually won’t achieve much either. I prefer to try to make the best use of it, that is, using technology to build a better society.
My feeling is that as long as AI doesn’t have its own initiative, we have a huge window of opportunity to transform the world.
And by initiative, I mean something quite specific. Yes, I can give it a goal, ask it to break it down into tasks, and let it work for hours without bothering me. But AI still doesn’t look out the window one morning and think: “This city is getting dirtier. There’s a problem here. I’m going to create something to try to solve it.”
That still occurs to us. At least for now. And as long as that remains the case, we have an execution capacity that just a few years ago would have seemed absurd, but we are still the ones who decide where to point it. The limit is our ambition… and the tokens.
Of course, I do not undervalue the critical and denunciatory work that many people and activists carry out against the almost monopolistic dominance of Big Tech, and I do not deny that we have concerning challenges, such as environmental impact, biases, and privacy, among others. But from this publication, I prefer to advocate for pragmatic optimism because it is the only thing that can yield interesting results.
Everything else, the IApocalypse or losing our jobs, are issues over which we have very little or no capacity for influence. So, as I always say, just as our ancestors faced an unfathomable and threatening nature seeking a crack of survival, we should consider those threats as relentless phenomena that, in any case, will offer some crack, some fissure, that will allow us to continue thriving.
❓ Do you have questions? Ask me in the comments.
🗣️ Did you find it useful? Share it so it reaches others.
Thanks for reading Charcos Tecnológicos! Subscribe for free to receive new posts and support my work.