Initially I was worried that AI will slash quality (in a sense of "more people would be able to generate something") but it looks like indeed it proves otherwise.
That's good news especially because it scales - millions of AI users will not even notice that they produce better quality (even they may not understand or even care about the code produced).
What's more above this "default standard" is that an experienced developer can rise the quality even further - "check every corner case", "validate everything was set correctly", "log everything according to the given log standards", etc. And all of that without [significant] intellectual effort - just result of discussions with AI models. And that's kind of human touch too. A place for experience and wisdom, i.e. it's good to have AI take care of "common standards", "most likely correct" approaches related to tiny implementation details but there will be experts capable of composing valuable solutions that are greater than ever.
TL;DR - I agree - craft for AI, wisdom for people :)
Love this insight, Andrzej. Your distinction between "default standard" and what experienced developers add above that is exactly what I mean by craft vs. wisdom. The question becomes: how do we intentionally develop that experience and judgment? Especially with junior developers.
In the past it took years and thousands of hours to develop the experience, the judgment capabilities.
Today, for both junior and senior developers AI brings instant entry level for any language or technology. IMHO the difference is in approach. As a senior, I request AI to deliver particular quality features from the beginning (especially additional something that checks the main code - unit tests, extensive verifications and double checks, perfect logging, etc). It is because I know what are these features and how important they are.
Junior may not ask for them and get something inferior.
However, if I were a junior today, I guess I would ask AI millions of questions "why this, why that, how to improve, refactor, restructure, do it differently, what are the key aspects, etc.". So maybe times are not so bad for juniors?!
It may be even more effective tools than what we had in the past:
(1) reading books and documentation,
(2) come to conclusions oneself through trial and error and/or deliberate in-depth planning and design,
Great perspective. I love this "we could design systems that guide junior professionals through problem-solving processes, asking Socratic questions that help them develop their own analytical frameworks." and also this "Experimentation Laboratories:". It may work this way!
I remember there were so many complaints in the past on SO, especially from juniors who felt treated not so well being forced to follow multiple rules.
But IMHO these rules were key to success - key to have questions asked well and answers to the point. Without the rules SO will be a mess like other Q/A platforms.
Now AI took all this work of millions for free. Don't know what to think about it. It's sad.
I'm also not sure how to think about it. On the one hand the democratization of knowledge carries a lot of good. On the other hand LLM companies monetizing years of work and contribution of SO community without any compensation feels unfair. Not even giving credit.
It's like "you must steal the first million". I hate tech companies for dealing with copyrighted contents by looking solely at their position in the industry, and future income and ignoring current authors needs. I would prefer world advance 10x slower but treating people well.
Initially I was worried that AI will slash quality (in a sense of "more people would be able to generate something") but it looks like indeed it proves otherwise.
That's good news especially because it scales - millions of AI users will not even notice that they produce better quality (even they may not understand or even care about the code produced).
What's more above this "default standard" is that an experienced developer can rise the quality even further - "check every corner case", "validate everything was set correctly", "log everything according to the given log standards", etc. And all of that without [significant] intellectual effort - just result of discussions with AI models. And that's kind of human touch too. A place for experience and wisdom, i.e. it's good to have AI take care of "common standards", "most likely correct" approaches related to tiny implementation details but there will be experts capable of composing valuable solutions that are greater than ever.
TL;DR - I agree - craft for AI, wisdom for people :)
Love this insight, Andrzej. Your distinction between "default standard" and what experienced developers add above that is exactly what I mean by craft vs. wisdom. The question becomes: how do we intentionally develop that experience and judgment? Especially with junior developers.
Just first thoughts.
In the past it took years and thousands of hours to develop the experience, the judgment capabilities.
Today, for both junior and senior developers AI brings instant entry level for any language or technology. IMHO the difference is in approach. As a senior, I request AI to deliver particular quality features from the beginning (especially additional something that checks the main code - unit tests, extensive verifications and double checks, perfect logging, etc). It is because I know what are these features and how important they are.
Junior may not ask for them and get something inferior.
However, if I were a junior today, I guess I would ask AI millions of questions "why this, why that, how to improve, refactor, restructure, do it differently, what are the key aspects, etc.". So maybe times are not so bad for juniors?!
It may be even more effective tools than what we had in the past:
(1) reading books and documentation,
(2) come to conclusions oneself through trial and error and/or deliberate in-depth planning and design,
(3) asking questions senior peers.
Andrzej I was thinking how to respond and a colleague forwarded me perfect post from Kristin Luck https://kristinluck.com/the-experience-gap-how-ai-could-disrupt-the-path-to-professional-mastery/
Great perspective. I love this "we could design systems that guide junior professionals through problem-solving processes, asking Socratic questions that help them develop their own analytical frameworks." and also this "Experimentation Laboratories:". It may work this way!
One more tool from almost the past: beloved SO.
I heard that usage of SO drops recently.
I remember there were so many complaints in the past on SO, especially from juniors who felt treated not so well being forced to follow multiple rules.
But IMHO these rules were key to success - key to have questions asked well and answers to the point. Without the rules SO will be a mess like other Q/A platforms.
Now AI took all this work of millions for free. Don't know what to think about it. It's sad.
You probably saw that news already https://www.reuters.com/legal/litigation/anthropic-wins-key-ruling-ai-authors-copyright-lawsuit-2025-06-24/.
I'm also not sure how to think about it. On the one hand the democratization of knowledge carries a lot of good. On the other hand LLM companies monetizing years of work and contribution of SO community without any compensation feels unfair. Not even giving credit.
It's like "you must steal the first million". I hate tech companies for dealing with copyrighted contents by looking solely at their position in the industry, and future income and ignoring current authors needs. I would prefer world advance 10x slower but treating people well.