Hey, it’s a typical Wednesday, and today we’ll talk about quite interesting research that focuses on how generative AI transforms your work. And in theory, you should work faster, do more, or work less; but as human nature says, that’s not the case. And instead of acceleration, you have the intensification, actually. So you work more intensively, and you’re more exhausted as a result.

The Problem of Expectations and Boundaries

And it’s a problem that as we get the more powerful tools, we get some false expectations and we don’t know how to work with them properly, how to build the proper boundaries. And the result of it could be just the burnout, or it could accelerate your burnout.

The Economic Divide and Access to Tools

Yet another separate topic of the discussion that we don’t maybe want to cover in this video: that we start segregation of the people based on the access to the GenAI. And for example, the Claude-Anthropic is super powerful. But for example, if you want to have the co-working mode, or if you want to use it seriously with the Opus, you definitely need to pay something like this more advanced subscription for a couple of hundred of dollars. And some people use it so actively and so proactively that they pay thousands per day. And for sure, even for me, a top engineer, maybe it’s a bit too much money, and I haven’t decided yet that it’s worth of giving it a try.

The “False Feeling” and Verification Loops

But another problem with the intensification of your work is not the burnout — it’s the false feeling that you don’t need other people anymore. And the GenAI closes the gaps, and it gives you the possibility to work in areas that you don’t fully understand and deliver some results that you could not fully validate.

And Anthropic even accelerates this problem with the statement that this “read-write-verify loop” with the thinking is not needed anymore because the Anthropic agents could replace the intellectual workers. And as far as we have a lot of people that are quite far from the areas where they have the critical thinking and could evaluate the work properly, they really believe that they do the high-quality work in this space.

The Evolution of Engineering

And I’m not talking only about engineering — don’t get me wrong. I’m not trying to protect the fact that engineers will disappear or will transform into something completely new because, just believe me, if you take a guy from the 60s or 70s that was working as a software engineer, he will be quite impressed with what you do, but quite disappointed of the level of your knowledge about the systems that you try to build with the modern tools. So it’s just the evolutionary process and we could not stop it anymore. So I’m already quite keen to see the future where we don’t need to write the code and all of these things, for sure.

The “Snake Oil” Paradox

But I’m talking about a different paradox: this paradox, actually, you know, it’s a perfect time for the people that were selling the “snake oil,” and right now they have the power to build the factories that create these snake oils on an amazing scale.

So we should be a bit more critically thinking about the things that AI companies try to sell to us. Because if Anthropic says that their products are fully built by Anthropic:

• First of all, we need to understand that they have the exclusive access to the models that maybe not all the people have.

• They’re not counting the cost of using the models because it’s their own product.

• This changes heavily the landscape, and they have the personal interest to convince you that it’s the way of doing the things in the future because their future depends on your workflow right now.

Lessons from Bertrand Russell

And recently I posted the interview with Bertrand Russell, where he explained why Marxism failed, and he has a lot of deep thoughts about the intention of the philosophy. But his own message to the people was that we should treat any of the philosophers with a bit of critical thinking and a bit of skepticism. And it’s the vaccine from the totalitarianism and all of the things, and from turning the philosophy to the religion and the false beliefs.

And I guess that Bertrand Russell’s message nowadays, actually, is crucial. Because I see the people that really make the religion out of the agents; they make the religion out of the “vibe coding”. They really believe that they could vibe code everything — that they don’t need engineers, they don’t need architects, they don’t need products — because right now, actually, everything that you need is just a product idea. And be the product architect and product engineer in a single place.

The Rise of the Solo-Entrepreneur

And right now it’s the rise of the solo-entrepreneurs that could just spend thousands and thousands of dollars per day for the AI and get something real, finally. But we’re not talking if this something real could scale, is it secured, is it really on the quality that will not fail and all other things, because we don’t have this ability to validate. We just could validate it from the user perspective, not from the engineering perspective.

Smart Hallucinations

And it’s the same for some research in a field of mathematics or in a field of some other science that, for example, I have the weaknesses. Sometimes I really get the shiny ideas from the cloud that look like the insights, but when I try to validate it with current mathematical theories and theorems, I just see that it’s a super natural and really smart hallucination. And but you know, I have the background in mathematics; I could feel some “smell”. But when the people simply don’t have any background in a particular field, they really see it as a pure magic, and it’s the problem.

Summary of the Crisis

Because, yeah, we have a couple of problems:

• We believe in a magic.

• We turn the things to the religion.

• But instead of acceleration, we get the intensification that actually leads us to the faster burnout.

And another big problem is that a lot of business leaders — that have no context in many areas that their teams actually have (and it’s the main reason why they hired them in some point of time) — start believing on this magic. And they’re constantly raising the bars; they constantly right now expect the more. That right now the single engineer needs to do the work of entire team, or maybe three teams, because right now in three days we could do something that we usually spent couple of weeks.

And it’s completely true if you’re building yet another microservice, and actually even from the 80s we was fighting for the CASE systems that just generate the code from the diagram. And right now we have super powerful and smart debugging duck that start talking to us and CASE system on steroids, but we forgot the sad truth of the CASE systems that failed completely.

And if you do something that’s already well-defined and it’s a lot of training materials for the models, yeah, they producing something usable, let’s say. But quite often it’s still quite far from something that we could use in a high-load environments. And the main problem of a majority of the business leaders and thought leaders, and especially solo-entrepreneurs, is that right now they’re raising the bar, they start relying on the people — they start believing on the AI magic — and as a result, they start burning faster themselves and they start killing the people around.

Closing Thoughts

And I see it as a pandemic, and it’s quite a dangerous for all of us. I’m not against AI — I use it daily — but the problem is that we need to apply a bit of critical thinking to everything and avoid turning anything to the religion and gods.

Reference