Ask HN: What would happen if your company stopped using all AI tomorrow?
This is something I’ve been turning over in my head lately. I’m not denying AI is being used everywhere, but I don’t know if it’s actually helping businesses overall. If we shut it all down it seems that businesses would be fine (and maybe even better since there’d be no token burning budgets).
104 comments
[ 0.23 ms ] story [ 24.8 ms ] threadBut if everyone else got to keep their AI, we'd lose customers very quickly.
Most just use Gemini or ChatGPT.
What does Anthropic bring to the table? Nothing, just a “me too” company
Related to code quality, I've noticed developers picking up tricks from AI. Like using feature/xxxxx as the branch name. Last year exactly zero developers used this convention. If any used paths in their branches, it was almost certainly their name. Now everyone uses this convention. Though, they do use correct capitalization instead of lowercase freaking _everything_ that AI does.
But for branches in addition to Feature/, we're seeing Hotfix/, Modernization/, you name it. If you're parking branches for a long time without completing them, I'm 100% appreciative they're categorized.
(Personally, for long term parked branches I prefer `{username/dev name}/` branches because then you know who the owner is without having to look at the commit log and guess based on recent commit authors. But that's the fun thing about branching conventions, there are so many to choose from, many of which have been around forever.)
[0] https://nvie.com/posts/a-successful-git-branching-model/
I really don't care where this convention originated because my point has nothing to do with the origin, my point has to do with the popularity of the convention. And I'm seeing it more and more lately.
Do you really think the phrase "load bearing" hasn't existed for years?
The AI learned it from somewhere though, it didn't come up with it on its own
TF at least gave a nice folder view on the branches screen, which we've lost now we've moved to GitHub, but it has made it easier to enforce.
People have been doing this for years.
Tell me something I don't know. My point is that it's usage has seen a _measurable_ increase where I work over the last two years.
I don't think I have a single "feature/*" branch before 2024. Developer names, feature names, sure. But specifically feature/, no.
> I don't think I have a single "feature/*" branch before 2024.
You might not, but other people have. Just because you haven't seen something doesn't mean it doesn't exist. Don't assume that your personal experience is representative of the whole world of software.
For me, this popularity is graphable. Telling me it's a convention doesn't change that fact. I'm sorry.
But holy heck, I really have to argue with their literal first point. All lowercase, for what reason? It's never explained why they insist on a dated convention instead of proper style.
And in what world is "feat/" abbreviation considered allowable?
I lump this in with advice from people who put their source in a src/ folder instead of Source/.
But that doesn't change my argument. For me, this popularity is graphable. Telling me it's a convention doesn't change that fact. I'm sorry.
Huh.
I've been using that convention for ages, but I have to remind Claude on new projects to actually use git in the first place, and then also specifically tell it to use branches.
Then again, I was using em-dashes before they got un-cool.
Regex the pattern over time if you want to make an argument against my supporting example.
'twas an expression of surprise, not an argument.
Surprise in both directions, 'cause what you like about Claude is one of the ways it fails for me.
> Now do an analysis of the people at your company.
It was mandatory at several of them.
> Regex the pattern over time if you want to make an argument against my supporting example.
I wonder which of the various employers I've had in the last decade would consider me regexing commits to be an NDA violation?
I choose not to discover.
Is this a subtle joke?
Proper weight distribution is critical in determining where the seams are.
I hate it if the assumption is that I'm communicating with a human, and I then get hit with a wall of text that I have to decipher the context behind. I'd rather read the prompt then.
That said, if everyone (not just devs) came in and there was no AI, the org would come to a halt. AI is load-bearing in my company. Everything from dev, code reviews, planning, doc writing, ops heavily uses AI. I mean heavily, there are dev that would not know where to start triaging an alarm without the robots. Same with customer service. So much of the human capital has leaked out of peoples' brains that is fair to compare it to brain drain.
Realistically, it would be painful for 1-3 months. People would go to teams with AI. A year from now, leadership is gonna have a hard time justifying keeping the lights out.
It is akin to asking "what would happen if your company stopped using the internet(to look things up)".
I've got all kinds of useful tools that do helpful things, like check to make sure Amazon hasn't lost a bunch of my inventory or suddenly deindexed my product from search. These sorts of things happen without warning from time to time and are trivially fixable if you catch them, but you have to be constantly monitoring 10 different things to catch anything that could go wrong in a timely fashion.
I've similarly got an inventory/cashflow management system that's deterministic that would be fine, but some of my other tools, like one that scrapes my Alibaba/WeChat/Whatsapp/Gmail supplier convos to keep my order status board updated, is finicky. I wouldn't be able to repair that once it broke.
For some tasks, I'd have to go back to freelancers. At this point, when I take over a new brand and want to refresh images or want to A/B test new ones with an existing brand, I can do that entirely with image gen. No AI would just send me back to Upwork.
For other tasks, I'd just have to do them myself. Right now when I acquire a brand, I have a Claude skill that runs through all of the Amazon search and ads reports to find the key search terms, analyzes the text of the listing and proposes updates to make sure we're hitting the ones that matter. I used to do that myself, but it takes somewhere between 2-10 hours depending how many products we're talking about.
The biggest loss would be the inability to expand. I started out just buying Amazon-only brands because Amazon provides a crazy amount of infrastructure that enables me to run a lot of them by myself. My most recent acquisition came with a Shopify site that had minimal sales plus a Meta ad account the guy who sold me the brand had done a little experimentation with. A year ago, I would've just discarded those, since my time is the biggest bottleneck. It takes a lot of time to run your own site and Meta ads, so it makes more sense for me not to bother and instead just make more Amazon-only acquisitions.
But now I'm having a great time really pushing Claude and ChatGPT to do most of the work. I redesigned the site more or less by just having ChatGPT do research on competitors and best practices, then having Claude run with that. That took maybe eight total hours of work.
I'm spinning up an enormous amount of creative to test on Meta, and I had ChatGPT research and spec all of the kinds of landing pages that are worth trying. Claude implemented them and set up the ads using Meta's ads APIs. I got all of that done yesterday.
So yeah, basically if I couldn't use AI any more, I'd get back to the original thesis of my business, which is buying small brands at reasonable prices and operating them as leanly as possible. It still works, but it'd be a lot less fun and profitable.
LLMs are a cool tool but we’re over-encouraging use in every domain, even tasks they’re not good at. Every tech company in the world seems to be getting rich though how can I possibly say things would be better without it?
Not my experience at all. The quality has gone up dramatically. Because we also use LLMs to make large scale refactorings that cleans up the code and use it to find lots of bugs that would otherwise have exploded in production at some point in time.
I went through each one through the code path (a couple were already considered during implementation), and while all 5 were theoretically accurate, 0 of them were actually reachable.
I wrote them back to Claude point by point and got this back: "I checked all five claims against the code. Your list is correct on all five. I downgrade my verdict: none of the five findings blocks approval. Details and one subtlety below."
This is a usual experience with SOTA models for me and others in my team.
Don't get me wrong, I use it daily; sometimes as a crutch, other times as an accelerator, mostly as a validator/peace of mind. I however don't think that neither velocity or quality is meaningfully changed. But I do spend much more time arguing with a machine.
I use Opus 5 every day, and it just works. It is very rare that I need to do any manual editing.
For example, a few weeks ago I asked it to analyse my code and find duplicated code that could be extracted into reusable libraries. It found several good examples and did the refactoring. I reviewed the result, and it was 100% how I would have coded it myself.
I do almost no manual coding these days. Opus 5 just works for me.
So it is strange for me to read comments like yours saying that it doesn't work. It would be interesting to find out what makes the difference?
For what it's worth I/we do much less manual coding now too, but it almost always takes a handful of re-prompts to reach expectations.
We can do hooks to reduce superfluous comments or do a second specs-adherence pass, but other discrepancies aren't uniform enough to enforce in prompts or context.
I'd love to know the difference too. Or whether long term velocity would be higher by vibing through PRs and deal with issues when things explode or refactor down 2M lines of LLM creep after the fact.
Other than that, interacting with my boss would become less stupid (or at least be his own variety of it instead of ChatGPT's variety of stupid) and nothing meaningful would change.
Everyone who's saying that they wouldn't get anything done at all, you're already doomed and I can't wait to watch the collapse after this crap has ruined so much.
I could theoretically maintain my output, yes.
Practically nope, not because I can't do the work without AI, as much as because of burnout - both from technical (Ansible, Python - neither are a thing I like using) and social (dealing with corpospeak) reasons.
However if everyone else stopped using AI we would be broke, so please don't do that.
The main issue I see is that currently managing 70 repos would just be physically impossible without re-hiring people versus like now with 8 monitors in grid and averaging 32 threads (4 per monitor) and being a bot for 10 hours a day, saturating about 20 subscriptions (mix Anthropic, OpenAI, Google, Xiaomi, Minimax...) in total currently using our own tools.
While I loved the era of before, I would genuinely be instantly depressed to go back to it knowing that the problematics have completely shifted, this isn't a 20% improvement or something, it's straight 10x increase (at least lately).
I use AI daily as well to build tools for my daughter, toddler video-games and so-on, it would be so frustrating to lose this.