chx2d ago
What's special about this toot is that this is from an AI maxx person. They would regularly come to my LLM criticism posts and attack, quote toot others to question their LLM criticism. And yet, even they know it's fucking horrible because when they posted a beautiful photo from nature they felt the need to add "no AI here".
Show threadIaniv 🇨🇦 🇲🇽2d ago

@chx You can out me, it’s all public.

Be fair. My “attacks” were only challenges to “LLMs don’t work”, which you keep saying.

I never said anything about the real issues around the tech.

Show threadchx2d ago
@ianiv https://blog.avas.space/llm-circus/
our workplace LLM mass delusion

The in-house sessions showing off the tech are embarrassing; led by the wrong people, and should likely not even happen in the first place.

ava's blog
Show threadIaniv 🇨🇦 🇲🇽2d ago
@chx not sure what your point is. Yes, some people use a tool incorrectly. Many people I know and I use it and get incredible results and actual value out of this tool.
Show threadchx2d ago

@ianiv my point is that it doesn't work. It's that simple.

> Uber president and chief operating officer Andrew Macdonald said it’s hard to draw a connection between the company’s rising use of Claude Code and innovations meant to serve consumers.

> “That link is not there yet,” he said. “Maybe implicitly there’s more that is getting shipped, but it’s very hard to draw a line between one of those stats and ‘Okay now we’re actually producing like 25% more useful consumer features.’”

Show threadIaniv 🇨🇦 🇲🇽
@chx and my counterpoint is that I see it work every day. I’m not stupid and I work with very smart people who will say the same.
Jun 14 at 5:05pmIndigo
Show threadIaniv 🇨🇦 🇲🇽2d ago

@chx it did take some time to understand how to use it well, and we worked together and shared what we learned across our organization. And we keep doing that.

What makes your arguments weak is that all you do is point at things other people write that support your point of view, and
then reject outright what people with different experiences tell you.

And because you refuse (and I will respect that) to use them, you cannot form your own opinion based on real experiences

1/2

Show threadIaniv 🇨🇦 🇲🇽2d ago
It’s perfectly fine to talk about all the other concerns about this tech. But you seem biased about its usefulness, and unqualified to talk about that without personal experiences that go beyond a naive experiment.
Show threadIaniv 🇨🇦 🇲🇽2d ago
And I hope you do not see this as a personal attack. It’s only a criticism of how you approach communicating opposition against something you obviously feel passionate about.
Show threadchx2d ago
@ianiv yes, AI psychosis is real, I never doubted that. I have my own first hand experiences where the simplest of scripts go completely haywire, a three line go program crashes and so on. And I have the Uber COO saying it doesn't work. I am sure you perceive something as working -- but it doesn't. It likely makes you slower or the quality of output much worse. I can't tell.
Show threadIaniv 🇨🇦 🇲🇽2d ago
@chx You remain biased and when challenged you attack me personally. It is impossible to have a civil conversation like this.
Show threadIaniv 🇨🇦 🇲🇽2d ago

@chx I will leave you with an example. We have an internal library with an extensive test suite. But we realized a whole new class of tests would enhance confidence that it works correctly when it is used in a particular way.

It would’ve taken weeks to do this by hand. With Opus it took just a few hours to add full coverage and the new test suite uncovered 10 bugs and 2 architectural concerns.

This directly leads to a better product that uses this library and more confidence in new releases

Show threadchx2d ago
@ianiv why would I have a civil conversation about something that ... ah, here's a good list https://smallsheds.garden/blog/2026/on-the-acceptance-of-genai/
On the acceptance of GenAI

Ethics is the worst.

Joep Schuurkes
Show threadIaniv 🇨🇦 🇲🇽2d ago

@chx what does any of that have to do with what we are discussing: whether or not the tool is useful and works?

In any case, we all have to accept that almost everything we use in modern society has ethical problems. You choose this one to be your hill, that’s fine.

But since you very explicitly say you do not care to have a conversation, I will stop now. I won’t bother you again and I hope you can extend the same courtesy to me.

Show threadchx2d ago
@ianiv not sure what lead you to think I am discussing any of this. It doesn't work, it's extremely harmful, that's all there is.