Welp.
-
GitLab's CEO writes:
> Humans still own the judgment that matters most: architecture, deep understanding of the customer problem, the tradeoffs that require taste.I hold judgement over this decision.
I see it as a symptom of lack of understanding of how good software is actually built. And of greed, directed by the hype train and FOMO.
And I find this blogpost, and the whole idea of pushing slop generators tasteless.
And if you've read that far and you're in IT, I have one word for you: unionize.
-
And if you've read that far and you're in IT, I have one word for you: unionize.
-
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/ -
@toxy it is.
-
Software has been "built by machines, directed by people" for decades.
That's what compilers and linkers do, that's what uncountable lines of Bash and endless CI/CD pipelines are – machines building software, directed by people.
And for decades, the bottleneck has not been churning out code. It was code review, it was quality control, it was bug fixing. AI slop makes that *worse*, not better:
https://freakonometrics.hypotheses.org/89367GitLab, and the rest of the industry, is solving for the wrong problem.
they try to handle this by using multiple Agents now, and one gets the „personality“ of the reviewer, one for writing the concepts and requirements etc.
but if no human cares to thoroughly sit down and describe precisely what the real needs are that the software and its features must fulfill and how the fulfillment of these requirements can be verified (acceptance criteria) , a project will never fulfill the needs of humans or the company they work for
-
they try to handle this by using multiple Agents now, and one gets the „personality“ of the reviewer, one for writing the concepts and requirements etc.
but if no human cares to thoroughly sit down and describe precisely what the real needs are that the software and its features must fulfill and how the fulfillment of these requirements can be verified (acceptance criteria) , a project will never fulfill the needs of humans or the company they work for
so i also want to add proper requirements definition and product ownership to the list of things to be done by humans but missing in many projects, making them fail in terms of providing the needed functionality or explode in cost, most often both…
if you just tell an AI extremely unclear what you want it will not complain like a human engineer - it will just deliver something vaguely matching the vaguely defined requirements…
-
so i also want to add proper requirements definition and product ownership to the list of things to be done by humans but missing in many projects, making them fail in terms of providing the needed functionality or explode in cost, most often both…
if you just tell an AI extremely unclear what you want it will not complain like a human engineer - it will just deliver something vaguely matching the vaguely defined requirements…
that will feel more comfortable to some managers because they dont like to feel resistance or get criticism for unclear requirements… but also just create huge piles of unsuitable crap software
-
they try to handle this by using multiple Agents now, and one gets the „personality“ of the reviewer, one for writing the concepts and requirements etc.
but if no human cares to thoroughly sit down and describe precisely what the real needs are that the software and its features must fulfill and how the fulfillment of these requirements can be verified (acceptance criteria) , a project will never fulfill the needs of humans or the company they work for
@lazyb0y no amount of description, however "thorough" or "precise", will be enough. Dijkstra knew this decades ago:
https://www.cs.utexas.edu/~EWD/transcriptions/EWD06xx/EWD667.html -
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/ -
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/@rysiek "right-sizing" is such an euphemistic way to say that people will be fired. This decision to move that way was surely not driven by the development department but by their investors.
I guess that this currently happens in a lot of companies. And I really can't see that this will go out well. This whole "The AI will save us multiple people" doesn't work. Sure, it speeds up things, but it increases the review time a lot. And when I understood that post in the right way, they even dream of letting humans completely out of the loop - and that will fail.
-
Software has been "built by machines, directed by people" for decades.
That's what compilers and linkers do, that's what uncountable lines of Bash and endless CI/CD pipelines are – machines building software, directed by people.
And for decades, the bottleneck has not been churning out code. It was code review, it was quality control, it was bug fixing. AI slop makes that *worse*, not better:
https://freakonometrics.hypotheses.org/89367GitLab, and the rest of the industry, is solving for the wrong problem.
-
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/ -
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/ -
@rysiek There's a chance this is just some yapping about how great AI is to make investors happy.
But its best to remain cautious@Eichi whatever it is, it is a huge red flag and does not deserve the benefit of the doubt.
-
@rysiek "right-sizing" is such an euphemistic way to say that people will be fired. This decision to move that way was surely not driven by the development department but by their investors.
I guess that this currently happens in a lot of companies. And I really can't see that this will go out well. This whole "The AI will save us multiple people" doesn't work. Sure, it speeds up things, but it increases the review time a lot. And when I understood that post in the right way, they even dream of letting humans completely out of the loop - and that will fail.
@heluecht "right-sizing" is just a euphemism for "we, the management, fucked up long-term planning and over-hired; we are bad at our jobs, but since we are calling the shots, we are not the ones who will pay for them."
-
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/ -
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/ -
Welp. I've been using GitLab for over a decade and have been pretty happy with it. Deployed and maintained several instances, some personal, some for small hobby orgs, some for work.
But it looks like it is time to ditch GitLab for good:
> Software will be built by machines, directed by people. AI is the substrate on which future software gets built. Agents will plan, code, review, deploy, and repair.
https://about.gitlab.com/blog/gitlab-act-2/ -
@rbphotographic I am probably going to replace my GitLab instance with Forgejo.
️
️