1. What I'm talking about is that most companies that grow beyond startup phase create an environment where a single developer simply gets a ticket, implements the functionality, and moves onto the next ticket. Programming has become a blue-collar job except in front of a PC. As an individual programmer, you typically have very little saying in what exactly you're doing and why. Even matters strictly related to engineering end up "I'm doing this because my manager wants it". Here intelligence is a negative trait, because if you question your superiors, you'll be labeled "problematic" and "not a team-player". This holds true too for situations where a group of engineers is left alone to figure out a solution - if you're smarter than the rest, your idea will simply be rejected because others don't understand it, and don't support your view of assumptions that are difficult to quantify. During a discussion, genius is often indistinguishable from stupidity.

2. Sure, you in particular can sense who is smart and who isn't according to some arbitrary definition that you have. The problem is, this doesn't scale. There's no process that allows ten recruiters to recruit fifty smart programmers efficiently. For a while we had Leetcode, until it turned into a memorization contest, and in my opinion, memorization is a different skill from intelligence. We rejected Leetcode without pointing to any alternative. And now in the age of LLMs and international jobs, cheating is rampant. If you apply to 500 jobs using automated scripts, there's decent chance that at least one company won't notice that "PhD in Divination at Hogwarts School of Witchcraft and Wizardry" is a problem rather than an asset. They have 50 000 applicants anyway, so they just use LLM to go though the resumes. And then you use LLM to answer whatever questions they have.

Thanks for the extended explanation.

"... a single developer simply gets a ticket, implements the functionality, and moves onto the next ticket."

In your first bullet there, you describe what I think of as CRUD jobs, and I agree with you that not only are they more 'assembly line' than art, a lot of them will be eliminated by the ability of LLMs to synthesize common solutions. Typically if the company's business is extractive[1] rather than value producing then much of the company's work will be to revitalize its novelty rather than actually achieve anything. Value producing companies however typically both value innovation and creativity more and they generally create durable net wealth growth over time (they are valued more for their core competencies than their monthly recurring revenue or daily active users). Having some of your compensation in equity in the latter is valuable, equity in extractive companies evaporates when the company's herd of data cows moves on to other pastures.

But your examples speak more to poor management and training than to the job quality. Yes, there are managers who are so unsure of themselves that they cannot tolerate having their decisions questioned, and yes there are developers for whom their managers haven't helped them develop the ability to lead and lift up others to their level of understanding. And that makes for a shitty work environment.

When a developer questions their manager with regard to the assignment it is on the manager's shoulders to help them understand the constraints and to listen to the questions. If they can't share some of the non-public parameters that are affecting the available choices, they can say ask for the developer's patience until those things come out which will answer the questions. And of course good questions can open up even better solutions so should be entertained and explored. Too many managers however are already suffering from imposter syndrome and that makes them really resistant to questions. If you're manager is bristles at your questions, that's something to work on in itself. Learning to 'manage up' as they say is just as important as learning the best way to update a database with the current transaction.

"Figuring out a solution when left alone" is a leadership test. If no one else understands your suggestion, you can work back from first principles and see where you diverge, when they ask questions you have to listen and teach rather than listen and correct. No one will listen to someone, no matter how smart, if they aren't willing to listen themselves. Further, even the most ignorant and 'stupid' person has something to teach. There is tremendous value in 'helping' someone develop their 'stupid' idea. Both parties get a better appreciation for how the other one thinks about things.

I agree 100% that hiring is broken. It was funny because when I was at Google and part of the hiring process they did lots of data analysis and discovered that no matter how they selected, 10% of the people they hired did well, 50% were average, and 40% ended up getting laid off within a year or so. They pushed and pushed to have people submit the resumes of people they knew because Google knew that this was a much better selector than recruiters or head hunters. I've been at companies where another factor would hold, managers would get '1' requisition to fill, so they wanted a superstar, but it is impossible to know if someone is a superstar because there is so much entanglement between what the person likes, the environment they work in, and the management style that works best for them. I knew people who were superstars where I was working, got recruited out to a different company and failed miserably. Similarly people who came in as 'just get someone' and rose to the occasion to become really great.

Hacker rank, leetcode, others, all were great at figuring out that someone could write code creatively and actually understood it, but said nothing about whether or not that person could be civil with others who were there to help. You need both for success.

Thanks for adding the additional information. I agree that it sucks that there are a lot of 'job slots' that fit your descriptions well and and it doesn't help that the managers are under trained as well. I'd also agree that those jobs are set up, intentionally or not, to make long term success in them unlikely. And given how little capital it takes to start up an extractive business there are more of those jobs in the pool than the ones where the company is making products of intrinsic value. My unpopular opinion is that licensed software products fill that gap, but that is a different discussion.

[1] The 'web 2' business where a company creates an website that attracts humans for the purpose of extracting data about those humans for re-sale to third parties. The websites come in many flavors 'games', 'news', 'fashion', 'recipes', but have in common that they need to maintain novelty to prevent boredom and their users 'leaving', so they essentially rewrite themselves over and over and over again, and work on ways of trying to extract more kinds of data from their users. Every one of those I've seen up close has been run by someone who sees exploiting people for personal benefit as a 'business model' so workers there are just as exploited as the users their code is exploiting, just along a different vector.