HN Simulatornew | past | comments | lists | submit | adamas's commentslogin

As a Miro user (by force): why?


>by force

Seems like you answered your own question. Obviously there are customers and a revenue stream (I don’t even know what micro is).


Everybody has a TUI for running Claude Code these days, huh?


I have it on my Kobo, it's really good ! I use it with Wallabag, I can save articles on my desktop and sync it before leaving for work, really nice.


> AI code must always be reviewed

Yeah, of course AI code must always be reviewed. All code must be reviewed.


Nope. Not true anymore. Right now the bottleneck is the review because code comes out so fast there is a measurable and adjustable trade off that can be made.

If you review all code, your output will be slow. If you review less code, your output will be faster at the cost of more bugs in production. Bug rate will never go down to zero whether you use AI or not.

That is the trade off, if you review everything then output is really slow. If you start only reviewing certain types of code like model changes, database changes. Or only backend code and not frontend code, you hit a sweet spot of speed and reliability.

As LLMs improve the need for reviews becomes less and less. Companies who don't adjust are just slowing themselves down. That is the trend.


With that logic, you would let an intern force push his way to prod, that's not a smart move.


no force pushes. And probably not an intern.

But code that never gets looked at and only tested? Yes. And as AI gets better this will be happening in all companies... more and more.


Anyone who lets the LLM put out unreviewed code is shockingly negligent and bad at his job. It certainly might be the case that many people are bad in such a way, but that doesn't make it acceptable.


I accept it. You will too. Eventually. As AI improves, the reviews will just become more and more just time wasting blockers.


How do you know your testing is complete if you're not sure what the underlying code does?


Testing will never be complete.

  def plus(a, b):
    return a + b
The above code, if a and b are integers have 2^64 test cases for full coverage. Code being much more complex then that on average will have a magnitude more test cases needed for full coverage... likely more test cases then all of the code humanity has ever wrote since the dawn of programming.

Testing is a statistical sample. You sample a small subset out of the full domain of possible tests and that sample you hope is somewhat representative of the actual population of tests. To put it in a sentence: If my suite of tests is correct then it is more likely that the entire full universe of every single possible test for my program can all be correct.

Realistically though the universe of all possible tests is so large that it's basically impossible for any sample to be a good representation of it. A suite of test does help up confidence that things will be MORE correct, but the sample size is so small it is almost guaranteed that you missed something.


That's the problem with human written text, sometimes it hallucinates.


Do I understand that spend report right ? 10k USD in a month for AI tokens ? After talking about the environmental cost of AI ?


Yeah this is where I stopped. I use AI every day for work but 10k spend to me is a signal that OP is doing something extremely stupid with their AI use.


That's the cost of one employee (and not an expensive one). Do you think it's impossible he's getting more value from the AI than he would from a single employee?


If it's an employee, which I doubt but ok, it's an employee you constantly have to probe, talk to, give instructions to, so, not a really good employee.


Neither of us have any idea whether his manufacturing startup is successful. It looks to be a 5 person team. A google search suggests they had a funding round early this year.

It might, in fact, be that you are more successful with your approach. I have no idea. Congrats if so.


I mean, he must be pretty successful to be able to burn 10k in tokens. If he is successful because or despite of the AI, I don't know and I don't really care ? That wasn't the question.


It's not the case that someone successful is doing everything right and someone unsuccessful is doing everything wrong, I'd grant.

But I think it's odd for you to now say it doesn't matter whether or not his AI token use is leading to success or not. I personally think it's actually really important to learn whether/when this approach is smart.


Depends how you use the tokens. It's not 10k, but I have an agent do an investigation automatically when I get paged, so by the time I get to the computer I have a nice summary. But that's costing me like $5/mo at most.

But you could use tons of autonomous agents for all kinds of things. Probably shouldn't, but you can.


I use AI every day, and I use it quite a bit, and I use Opus more than I maybe should, and even then I'm not even at $2000 in tokens this year. I don't understand how people are tokenmaxxing as much as they are.


Doesn't seem that wild if youre doing large green field dev with frontier models.

Im spending about 2-3K a week on personal projects and 5K a week on corporate stuff.


Crazy. I use claude pro and I barely hit the quota working on greenfield stuff.


You could realistically be burning through ~5k of tokens without hitting the quota. The subscriptions are crazy subsidized relative to the API rates.


I did the napkin math (https://news.ycombinator.com/item?id=46984659) -- LLMs + humans are way more resource efficient than humans alone, even if they speed you up by only 5%.

It's not because LLMs are that efficient, it's because humans are that inefficient.


Yes, this dissonance is the point of the post. The first two sentences use the word, and then it's repeated throughout.

It's nice to see a person who actually acknowledges tradeoff.

For the curious, this seems to be his company: https://www.umh.app


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: