Comment by onlyrealcuzzo
2 days ago
> It's insane to me that Andrew thinks this post will somehow exonerate Zig when it really just makes them look childish.
Antirez made a post equivalent to: you'd be a fool not to use AI to increase test coverage.
Zig on the other hand has embarrassingly low test coverage given its adoption and time in development.
Their stance on AI is completely childish. They could benefit massively from it, yet refuse to even consider any potential usage.
It's one thing to try to stop PR spam. It's another thing to tie your hands behind your back and not even use it internally for the lowest hanging fruit where it could have major benefits.
They could use AI to triage potential real bugs from PR spam... but instead they just let real bugs go unnoticed for longer than need be because they won't even use AI to help triage...
in what way is it chidlish to have a principled position like this? you may disagree, i def use llms. but andrew has clear reasons for why he doesnt.
Childish or no, anti-AI sentiment is ubiquitous and growing.
From a PR perspective there’s a lot to gain in the short term by picking the “anti-AI” lane. And you can always change your mind later.
Surveys show that 60% of US adults don't like it which means 40% either don't care or do. I'd advise not getting your ideas about sentiment towards it from places like HN which are very biased and unrepresentative places of anything.
I find the anti crowd increasingly to be hateful and close-minded and it is disappointing because I have a lot of friends in it. There's a moral puritanism which gives people feelings that they are on the "right" side and thus any level of rudeness or hatred is justified and it only hurts their side.
I don’t disagree with your characterization.
But from a purely Machiavellian perspective I don’t see a lot of downside in courting this group in the short term.
I'm not anti-AI at all but to call people who don't want to use it "moral puritans" is laughable. It's hard to read that as anything other than a self-report on your own extreme nonchalance when it comes to the many ethical thickets surrounding this technology. While the possible benefits of AI in fields like medicine are promising, the concerns that a lot of people have about LLMs and the supporting infrastructure, power shifts, and more that have come along with them are diverse, serious, and completely understandable, and I can't imagine them being "puritan" to anyone other than those ruthlessly driven by nothing but capitalistic self-interest.
1 reply →
Or perhaps it's https://en.wikipedia.org/wiki/Diffusion_of_innovations and the people with anti-AI sentiment are called "laggards". If so, their amount will likely be near zero soon.
This has happened thousands of times in human history.
Sure. That’s why I said “short-term” and “change your mind later”.
The thing about laggards is their money spends the same as everyone else’s.
Is popular sentiment a good measure for a decision? Anti lots of things have been growing at some points in history, including bad things.
My point is that it’s a non-decision. You can use AI later, when the models are better.
In the short term, you can address the anti-AI market.
It's ubiquitous among dinosaurs. Don't worry. They will be left behind.
[dead]
[flagged]
5 replies →
I am fine with people having principles and doing things their way. Not everything has to be a race to be the best. There are still plenty of people that appreciate traditional crafts. Anyway, if Zig+AI can be the next, greatest thing, can't someone just fork it and make it happen?
No one cares enough of zig to do it, people will just continue moving to Rust and similar.
Well exactly, "Zig + someone using AI" can never beat Zig + someone who cares. A developer that cares about the project comes first, no amount of AI can change that. AI could be a good tool for a developer though.
People are doing the simpler thing: moving ecosystems more suppotive of their goal.
> Zig on the other hand has embarrassingly low test coverage
And? Why would low test coverage matter. It’s not an indication of project quality nor does a high coverage mean an absence of bugs or errors.