Two observations:

1. AI amplifies existing competence. Current AI systems are not autonomous problem-solvers -- they are accelerants. Users who already understand their domain can spot errors, ask better follow-up questions, and integrate outputs efficiently. For less experienced or less analytical workers, AI often creates rework rather than savings, which is consistent with the “AI tax” described in the surveys.

It’s not strictly about being “smart,” but also about task structure and feedback loops. Some highly capable workers are in environments where AI cannot be safely or efficiently applied (compliance-heavy workflows, fragmented tooling, high-stakes accuracy requirements).

2. Impact varies strongly by role, as knowledge work is not homogeneous. Roles involving synthesis, drafting, ideation, coding, analysis, or decision support benefit far more than roles dominated by coordination, approvals, interpersonal judgment, or rigid process constraints. Executive workflows are especially well suited to AI assistance, which explains the perception gap between leadership and individual contributors.

Also, adoption maturity matters. Many organizations have introduced AI without training, workflow redesign, or incentives, which predictably limits upside regardless of worker capability. Would it be exactly surprising to see this outcome? Not to me.

As of today, “vibe coding” is mostly for non-coders who suddenly need to write code. It’s basically WordPress-on-steroids (the modern incarnation of which being Lovable) but for software instead of websites. Actual programmers are, of course, evaluated on how well they use whatever AI tools are all the rage this week, but they still have to fundamentally be, you know, programmers. :)

For context: I’m a software developer (wrote my first .COM virus in 1987), now a CTO of a mid-size regtech company, with an MBA for my sins.

I used Ruby for all my PoC experiments and web-facing MVPs. I mostly used Rails, though I’ve never been a fan of the “Rails way”; it has always reminded me of that “How to draw an owl in two steps” meme: draw two circles, then draw the rest of the owl. I follow the updates and lurk in the Telegram groups to keep up with what’s happening in Ruby and Rails.

Here’s the thing: Ruby is fantastic for a team that already knows what they’re doing. It’s concise, powerful, and expressive. I’ve been coding for 25 years before I got to the Silicon Valley during the early Rails boom, friends showed it to me, and I was instantly hooked. Everything I needed was exactly where I expected it to be, and when something wasn’t, Ruby’s metaprogramming made it very easy to hack something myself when I needed that.

But that is also Ruby’s biggest problem: it’s just not designed with new programmers in mind. It shines only when you have someone with years of accumulated knowledge. Yes, a couple of mythical 10x engineers can build a serious production system in Ruby, but they can’t sustain it alone, that’s what mid-level developers are for. And if there aren't any on the job what exactly are you supposed to do?

(Tangent: Is it just me, or has the quality of fresh university graduates dropped noticeably in the last decade? I used to meet young devs who had been programming since age 12–15 and were ready to build real things for reasonable pay. Now I interview graduates who can’t explain what a HAVING clause does in SQL, even though databases are proudly listed on their CV. I’m in the EU; maybe the situation is better across the pond.)

So is Ruby a serious programming language? Absolutely.
Can it support complex, production-grade systems? Definitely. Ask Shopify. Read DHH’s posts if you care.
Would I recommend Ruby for a brand-new production project today? It pains me to say this, but no way.

Programming languages live and die more by their market than their technical merits. Look at JavaScript: an architectural Jenga tower that somehow runs the world because we’re collectively trapped with it. Ruby doesn’t look great from that perspective -- there's no captive audience, and its niche remains PoCs and MVPs built by one-person teams. It needs something big to break out of that.

Unfortunately, Matz seems to have lost interest, and the recent PR disaster around the Rubygems takeover didn’t help. Sic transit.

these are probably the 50+ year olds who are too used to it. as that generation is naturally replaced, the sales aren't.

facebook became a monster after its algorithms figured out that people engage most with toxic and dumb content. that’s a reflection of reality, not something Zuckerberg invented -- the platform merely amplified it.

the same thing seems to be happening here: the algorithms learned how people like to interact with music and then amplified that behavior, which some of us may label dumb.

so, do we blame the algorithms now, or Spotify for that matter... or do we look in the mirror?

I certainly envy the faith of the editor using the past tense there. What are chances they changed it? :p

They’re trying to do something genuinely useful for everyone. It’s not really their fault that the markets are so eager for growth that expectations, and the money pouring in, are way over the top. At some point, there’ll be a correction and a lot of investors will probably be disappointed, but honestly, that’s just how capitalism works.

Many breakout successes in gaming happened due to a platform change. Think AJAX from 2005, then social networks from 2008, then mobile from 2010.
Anything tried since -- VR, blockchain, smartwatch -- failed to gain traction so far. So yes he is looking for a new platform, in part because his company is losing ground on the PC and consoles. And he's also right about the tiktoks of the world eating games' breakfast, especially in the mobile part.

Personally I have high hopes about AR gaming, but the hardware is not there yet.

What makes you think so?

still unsure what is so different from having an LLM de jour tab open and asking it my questions there.
okay sure Comet has a somewhat easier UI for answering questions about the page I'm visiting but I just don't seem to have a need to do that.
so far for me this falls into the "solution looking for a problem" category. I acknowledge I might be not the right audience for this though.

But the reason here is money. (Now who would've thunk!)

The more people are diagnosed with whatever, the more money there is to be made by producing remedies. Remember that if you exhibit symptoms of a disorder, doctors are obliged to treat that -- if they don't, you can sue them for negligence, even if they disagree with the diagnosis or with the whole approach.

Throw in the mix here the interests of the big pharma and medical insurance businesses, et voila, on a systemic level you have a recipe (pun not really intended) for a disaster. (Scientific integrity? No, haven't heard of that one, does it mean we make less money?)

the major reason enshittification happens is because of how software development is financed.

basically, as a company creating a product, you front the dev costs to build something that usually can’t justify those costs on its own. in other words, if you just put that money into an investment fund instead, you’d probably end up with a higher NPV.

but you still do it, because you expect the project to eventually consume itself through enshittification... and in the process, recoup the costs and hopefully make enough extra to keep the bean counters happy.

this dynamic has also created a warped job market for developers, who got too used to being paid a lot thanks to the speculative value of those future enshittified products they build. it does look like that market is now going through a pretty major correction.

oh, and let’s not forget that enshittification works because people still click on those ads and whatever else. it really DOES make more money for the company, at least in the short term. public corporations aren’t stupid, they’re merely myopic by design.

and backdoors for law enforcement and intelligence and whatnot. teh irony

My experience as a CTO hiring developers is very similar: there just aren’t many fresh graduates worth hiring anymore. I’m not sure exactly what changed: maybe universities got worse at teaching? maybe students got worse at learning? maybe a bit of both.

Ten years ago I had a steady supply of promising graduates to choose from. Now, most fall into two categories: a small handful with multiple offers (hard to compete for), and many others who are net-negative for at least two years. They require so much time and effort from more senior colleagues before becoming productive that the investment rarely pays off. Especially if you consider that often once they finally reach that point, they leave before they even break even. So what’s the point of hiring them in the first place? That's why we're only hiring folks with 5 or more of experience now.

(For context, I’m in the EU, so YWMV.)

I would surmise that advancements in AI have also filled much of the demand for simple tasks that can be done for a fiver.