The pitch is, for those that have never coded it’s a gateway into a skill they never wanted to learn

I’m simply whelmed by it

You never hear someone ask “how can a socket wrench help us with this problem”, why do we do it with LLMs?

Spitting bars.

I’m working as a TLM for a company that’s only recently become AI maximalist… I see myself in your writing 🫠

So which is it? Are developers 55% more productive, or are they losing 20% of their time to inefficiencies and burning out at record rates?

The answer: executives are measuring—and reporting—what makes their stock price rise, not what’s actually happening on the ground.

Or if you want to get slightly more conspiratorial: the execs are all buying shares in OpenAI, Nvidia, and the like - so now they're more interested in ordering people to use LLM tools so that these stocks rise in price, even if it means sabotaging their own company.

The average corporate executive has no idea just how much of their organizations resources are dedicated to "churn." By "churn" I mean the flurry of activity that provides the illusion of productivity while actually producing very little.

I worked for one company that liked to move dev teams around to different platforms to [theoretically] increase productivity around major feature releases.

Pretty much anyone who has been in software engineering for more than a week knows that reassigning devs to a project that they're unfamiliar with and only plan to work on for a short time will actually slow development down, not speed it up. To make matters worse, some of these teams had very poor practices. I'm trying to be charitable with my words.

The immediate result was that the number of bugs and defects skyrocketed, dramatically increasing the workload for QA and the core dev team. Management's response was to pile a bunch of unnecessary requirements that had to be met before a PR could be merged. This failed to produce a meaningful reduction in defects while bringing development almost to a halt.

One of the directions I gave to the core team was that when they worked a defect ticket, in addition to fixing the defect, they needed to perform a root cause analysis, link the defect ticket in Jira to the ticket that introduced the defect, and write the root cause analysis in the comments.

If anyone from the product team had bothered to review the defect tickets, they would have found that a small number of individuals produced 90% of the defects. Instead, they carried on with the unfortunately common notion that software engineers are just overpaid monkeys and that pretty much anyone can bang out web apps.

The current views on "AI" are often a continuation along that line of thought. The value of Institutional knowledge is basically impossible to quantify so it gets ignored altogether. I honestly don't know what impact this will all have on the average organization. I do think that if someone has staked their entire organizations future on an LLM, they're in for a bad time.

Aside from all the financial and anti-social considerations, I think AI is "vindication" for all the middle and upper managers who felt engineers were either lazy or precious about their work. "See, I shouted at the computer and it did what I wanted in seconds instead of months! Why can't you do that, nerds?"

Unfortunately, voicing concerns about quality will only reinforce this dynamic.

And while I agree broadly with what you wrote… AI writing unit tests? God help us all.

P.S. neologism alert: morged.

Amen.