A lot of people are using Claude Code which many consider to be a noticeably better for coding than the other models.
I think also they tend to be generating non-C++ code where there are more guardrails and less footguns for LLMs to run into. Eg they're generating Javascript or Python or Rust where type systems and garbage collection eliminates entire classes of mistakes that LLMs can run into. I know you said you don't use it for Python because you know the language but even experienced Python devs still see value in LLM-generating Python code.
That’s funny bc I linked my post to a server I’m on and I also was told to use an agent.
My worry about an agent is I’m trying to translate the math with full fidelity and an agent might take liberties with the math rather than full accuracy. I’m already having issues with 0 to 1 indexing screwing up some of the algorithm.
Humans are also non-deterministic code generators though. It can be possible that an LLM is more deterministic or consistent at building reliable code than a human.
My first impression of the idea is that what you're pitching is a lot like Kickstarter or Indiegogo (I'm assuming you're talking about physical sales because you mentioned Shopify). Kickstarter even holds money in escrow as a stronger signal of demand.
Interesting! Kickstarter is more goal oriented, our initial users just want to know if someone is willing to buy (or places a pre-order for) their product and then start growing their business if the market pull & size makes sense. Would you happen to know anyone around you who is starting a business or wants to start one?
No, DNC picking a bad candidate should be responsible for a D electoral loss, it is not responsible for Trump's electoral win. A lot of people seem to think only the DNC has agency in elections, only they should be responsible for outcomes.
The essay reads to me as 5000 words on why epistemic standards should be lowered for effective fraud investigation. Requiring retrospective evidence - the highest epistemic standard - is too slow and expensive, so the government needs to copy the private sector's epistemic standards where a collection of solid heuristics and smart observations is sufficient to preemptively block payment.
It's an insightful point, but it seems consistent with those epistemic standards that it's unimportant whether the 50% figure is valid in retrospect. The observation of "industrial-scale fraud" is sufficient to act, it doesn't need a retrospectively validated 50% figure and he would like you to get all the way off his back about it.
Sorry I meant that humorously in the vein of Ryan George's comedy YouTube series "Pitch Meeting" where one of his catch phrases is "I'm gonna need you to get all the way off my back on that"
No government body is run completely on the ball, which is why it's such an effective bad-faith demand.
I don't even know where this belief comes from. I'm certainly not aware of any historical scenario where authoritarian regimes end when their opponents finally embody perfect behavior above reproach.
It would be a bad-faith demand if I was asking people to assume every blue-state government was bad. But I'm not: I'm simply asking to recognize one that clearly is.
You don't need to ask people to assume every blue-state government was bad, Nick "name a Democrat city that's prospering right now" Shirley will do that after you, you just have to tee him up by saying the first part: "we need every government body we run to be completely on the ball".
I doubt you yourself are engaging in bad faith (of course I recognize your username) but it's still a bad-faith demand to expect "completely on the ball" behavior above reproach and your intentions don't matter when you echo the demand.
For the author, it wasn't enough to simply recognize a failure to prosecute fraud fast enough, such a failure must be characterized as the cause of the irresponsible demagoguery that followed. Then turns around and wonders why his article isn't treated as the apolitical dissection of fraud that he claims it to be.
I don't understand this at all. The DFL-controlled government of Minnesota royally fucked up and allowed fraud against a social services program on an industrial scale. That fraud isn't a small crime; it's a grave crime, victimizing the most vulnerable people in our society. It's a very big deal. This is a technical post discussing a variety of different ways in which program administrators could hope to prevent something like it from happening in the future.
How are people finding ways to downplay or dunk on this? I just don't understand. What do I care how "apolitical" it is? I don't care. I do not care. The fraud is what we should care about. That's what the post is about.
You don't have to care how apolitical it is but the partisan political nature of the post, which it starts and ends with, is why the HN thread is reacting to and discussing partisan politics. What makes it partisan is the shift from admonishing the government to justifying the partisan "irresponsible demagogues" that are currently brutalizing Minnesota by pointing to the blue-state government's slow prosecution of Somali immigrants.
When Charlie Hebdo was bombed and shot, I suppose what people should have been writing is a technical post about the poor quality of their work with tips on how to convey the same artistic point in a way that doesn't invite fanatics to bomb and shoot them, concluding that by not reining in their bad work they have ceded the field to people who will not be gentle in their proposals. Then you can comment things like "What do I care how apolitical it is? The art is what we should care about. That's what the post is about".
Edit: maybe a better analogy would be 9/11 with the US and Al Qaeda, where the US would be less innocent in your political sensibility than Charlie Hebdo and the dynamic I hypothesized was more real.
The partisan politics are in the story and part of its substance (and really how could it not be? what he's suggesting has substantial political consequences even setting aside the naked partisan jabs). The presence of technical details doesn't negate that, many polemics have technical details.
That is not a plausible objection to this article and as someone actively involved in Democratic politics and a compulsive HN participant I find this whole thread really embarrassing.
AI is not the only way to make synthetic music. If you have an exact idea, you can use virtual instrument plugins for software like Ableton Live to produce music.
The difference I see between a company dealing with this as opposed to an open source community dealing with this is that the company can fire employees as a reactive punishment. Drive-by open source contributions cost very little to lob over and can come from a wide variety of people you don't have much leverage over, so maintainers end up making these specific policies to prevent them from having to react to the thousandth person who used "The AI did it" as an excuse.
When you shout "use AI or else!" from a megaphone, don't expect everyone to interpret it perfectly. Especially when you didn't actually understand what you were saying in the first place.
All major platforms have mechanisms to identify ban evasion. It's not so easy to create another account when, for example, they ask for a phone number.
In the U.S. at least, it is trivial to buy a new SIM anonymously. But really, you should refuse to use any platform that requires a phone number in the first place. These companies make it implicitly very clear that they want to control you and extract every bit of information that they can from you.
reply