It's way more than code. Sure AI can crank out code at prodigious rates. Gary Tan, Y Combinator's CEO says he ships 37,000 lines of AI code per day [0]
And so can I. (oops)
"In the Beginning" (I was there) people wrote accounting packages in BASIC. It worked, the language allowed rapid prototyping, and out the door quickly, but BASIC lent itself to spaghetti code, and for anything really serious, the programs were too lightweight, and were very difficult to document and maintain, so that bugs could be fixed and esoteric features added (for $$$) without the fix breaking something else. Every damn line of code had to be commented so that someone else could pick it up when you left and maintain and upgrade it.
So, AI's got a mind of its own, and from what I hear, every time you get a solution (code) it's different from the previous. At this point, no solid libraries, such as mathematicians, physicists, medical researchers and yes, rocket scientists can rely on as 100% solid and "bet your life on it"
In addition, the hype has extended AI into more general areas, including "bet your life on it" situations where people are using it for therapy, with fatal consequences at times [1] "Nearly 1 in 5 U.S. Adolescents and Young Adults Use AI Chatbots for Mental Health Advice" (RAND) and it's so flawed.
And also, it leads to cognitive surrender. [2] "AI and the Psychology of Cognitive Surrender" (Psychology Today)
Key points:
• AI subtly erodes our cognitive strength by making delegation seem like self-generated thought.
• After repeatedly turning to AI for answers, the first thing that erodes is tolerance for not knowing.
• True judgment is built by wrestling with uncertainty, not outsourcing discomfort to machines.
In a very brief thread about Siri becoming AltSiri [3] my comments regarding the wide use of a tool that is IMO overextended and using the general population as guinea pigs:
---
I view and use computers as tools. They (mostly) do what I command.
That's because I am by nature a problem solver, and so are others. In fact, if knowledge consists of understanding a particular domain, and wisdom consists of applying knowledge across different domains, creativity of a sort, one of them being that unknown called the future then "button pusher" answers kill my ability to deal with future situations which are not recorded in "The Book of Common Knowledge" (a SNL reference).
When "computers" wrestle control of the situation and solve everything, then, as someone said in the early 20th century "Everything that can be invented has already been invented" then there's now no need for computers at all, since "Every problem can be solved by a chatbot" and no need for creative (genius) things like the famous "Wordless Workshop" that ran in Popular Science and Family Handyman magazines.
Just answer machines. No need to learn anything, nor to create.
Creativity and genius move us forward. That's why we have Hacker News as opposed to those "answer forums"
---
And YES, code that you have to reverse engineer in order to maintain must be understandable and well-architected. If that's "Elegant" then So be it.
I rapidly prototyped in-house apps, quickly and well, and they had a limited life span.
But "enterprise" software isn't going away. And whom [4] do you call when some CTO calls you at 1 a.m. when their business takes a header? Claude?
[0] https://news.ycombinator.com/item?id=48414607
[1] https://www.rand.org/news/press/2026/06/nearly-1-in-5-us-ado...
[2] https://www.psychologytoday.com/us/blog/the-digital-self/202...
[3] https://news.ycombinator.com/item?id=48413555
[4] I was born in Boston. Cheers!