kentm

Not Analemma_ but I'll give my two cents.

In my mind the issue isn't that it was crafted by AI, but that it wasn't crafted with the intention of getting the point across in a way that respects the reader's time or energy. The result is often needlessly wordy with information buried in a sea of flourish. This might be forgivable if the author had a certain entertaining style, but LLMs are aggressively mediocre by their nature, making wading through it painful.

Its certainly possible to get an LLM to write concise prose but thats not how most people are using them. We can talk all day on HN about the ideal LLM user that is conscientious, reviews the output, puts time and effort into stylistic concerns, etc. But none of that actually matters because thats not what people are doing; the entire reason they're using an LLM in the first place is so that they don't have to do that.

> If you want to know why big, progressive cities have the most unaffordable housing, look no further than this law.

This is an unsubstantiated assertion. I'd want to see evidence that things like this actually move the needle on rent vs things like available housing stock, zoning, etc which seem much more likely to actually impact affordability.

No, because the issue is navigating legal agreements with your vendors, not the ease of deploying. With AWS Marketplace you are making a new legal agreement with the vendor of the product, not with Amazon.

> But I won't follow you as far as "expecting developers to have expertise in how and where their software runs is unreasonable".

In my professional experience, management tends to treat this as a dichotomy rather than a range. And its lead to situations where we're choosing to route every decision through a small number of 'experts' and expecting that to scale, or have people make very bad decisions about things like indices, data layout, etc.

I align with you on this. If you are using some sort of system, there is a sort-of minimum bar of expertise one should acquire. When using analytical platforms, having a basic understanding of how data is stored and the levers it gives you to optimize data processing should be standard.

Some examples:

MySQL - I would expect developers to be able to come up with a list of columns to index and why, understand why UUIDv4 is a bad choice for primary keys, know how to write performant SQL by hand and understand when their ORM is going to do the wrong thing, etc.

Redshift - Understand dist and sort keys and have some mental models on how data shuffles and is pruned based on each choice.

Iceberg + Spark/Trino - Understand that data tends to be stored/loaded as batch files. Understand how deletes work. Understand partitions and sort as options. Understand generally how bloom filters work and when you should decide to place a bloom filter. Understand why UUIDv7 is better than UUIDv4 when its part of a partition key.

> Just pretending that AI doesn't exist or doesn't matter is one way to go.

I don't think thats a common position on this site. It'd be more accurate to say that people want balanced content.

I use AI quite a lot in my work, but a ton of AI content is just thinly-veiled marketing and not particularly interesting, IMO.

This is where I’m at. I find AI to be useful but there’s a tendency for it to suck the oxygen out of the room. Other things still matter.

Edit: I'll also note that while its nice to see real novel AI news I find a lot of it is extremely marginal value. I mentally associate most discussion with debating what color of marker to use when marking the edge of your CDs to improve the sound.

> React can render at more than 60fps if necessary, and if the code isn't shit.

Sorry, that doesn't really come across as a ringing endorsement. React apps are typically not doing anything complex so rendering at less than 60fps should only happen if you're doing very computationally intensive things or writing bottom-quartile quality code.

> A badly made native app is worse than a well built react-native app. There are plenty of anti-patterns that will ruin your native app.

Sure, but thats not an interesting observation unless there is something about react-native that makes those apps consistently higher quality than native apps.

On the other hand, like-for-like C/C++/Rust performs better than javascript in terms of CPU and memory use, so the null hypothesis is that re-implementing would , in fact, improve things. There's little reason to think a priori that the end result would somehow be worse.

In many cases, no.

> I don't understand why a police cruiser can sit in a public space (or even a private one) and write down licence plates and descriptions of passers-by with pen and paper, or record everything around them with dashcams and bodycams for later use, but when it comes to cameras on a pole this would require a warrant.

Scale actually matters. Things that are generally OK at a small scale become problematic at larger scales. A single police cruiser writing down license plates isn't able to track you in the same way a huge surveillance network is, and the opportunities for abuse are much lower.

They did not present it “as if they don’t have a choice”. I think your personal biases are impacting your reading here.

No, I'm sorry but I think thats a cop out. The fact that LLM are stochastic and can give incorrect answers is not particularly difficult to comprehend, and the risks that fall out of that are reasonably understandable. The issue is entirely down to bad choices by the people driving LLMs, because they are engaging with what they wish LLMs do instead of what they actually do.

Yes but the key here is doing proper risk assessment. "What is the consequence if the LLM gets this wrong?" "How do we verify the output?" "What are the legal ramifications for using the LLM in this way?" "Who is responsible when the LLM fails?" "Whats the expected accuracy here?" etc.

In the current AI mania, there's a lot of due diligence simply being ignored. Plenty of "Well humans make mistakes too!" going on here on HN too.

Audit logs are only useful if they are used to action abuses. The issue is that police departments protect abusers due to the “thin blue line” mentality.

I don't believe that deportation is inherently morally wrong, but there are certainly ways to deport people that are less moral than others.

Thats not an accurate framing. There are some disagreements that are red lines and others that are not. I think Chrono Trigger is one of the greatest games of all time, and someone disagreeing with me on that point is not a big deal. A hyperbolic example of the other side would be a Nazi's opinion on Jewish people. Given the nature of that opinion, not only would I disagree with them it would also be completely appropriate to fully disassociate with them and highly encourage others to do so.

People draw lines in different places, and what you are doing here is implying that it is unreasonable for people to take that level of issue with DHH's "gypsy" comment. I disagree. I think its completely reasonable to view that as unacceptable and to disassociate from that person.

The bill does come due. I've seen it happen over and over. But management fails to connect current problems to past mistakes, especially when they happen over the course of years. Thats setting aside when they insist that the correct decision was made, but have no real evidence that this is the case, other than the fact that "things worked out" even though maybe it would have been better had a different decision was made.

People do get a pass on this -- you can't rigorously examine every decision and counter-factual or you'd never make forward progress. And I don't really have any great ideas on how to bridge this gap. It'd be nice if there were better ways to measure this but there aren't, so it really comes down to who is more convincing, which is often orthogonal to facts and correctness.

I'd also argue whether ORMs actually save that much time in practice. In Java, for example, the main time sink is the JDBC plumbing and its easy to use something like JDBI that handles that plumbing without abstracting away the underlying SQL.

The application->database layer is pretty impactful and it pays to pay attention to it, because poor access patterns will cause a lot of trouble in the future, and its made worse by not having a very accurate understanding of whats going on in that layer.

I think a lot of developers don't have a good sense on where their time sinks actually are. Boilerplate is not pleasant to write but also not the timeline-destroyer people tend to think it is. And its often not enough of a time-sink to warrant introducing "magic" that will have very large negative future impacts.

Agreed. Usually an updated-on timestamp is sufficient to cover your bases without over-complicating the table. And your PII redaction is probably going to be shredding or nulling the fields instead of hard record-level deletes.

For #4 I always tell people to assume 99% append only, but do consider the need for updates/deletes in edge cases.

If any of the data is PII and you will be subject to GDPR (you probably want to be at some point) then you will need a way to hard delete it.

A large portion of append-only datasets I’ve worked with have run into some edge case that required an update.

Also if you’re doing an append only log plus mutable view then please use triggers or materlialized views. I ran into a few cases where the approach was to just update both tables and that lead to divergences between the two.

Yeah I hate infinite scroll. I have no idea what actual common sentiment is because you’d have to do broad polling to find that out. I always wonder if UI designers fall into the “everyone i talked loved it” where everyone consists of “other designers and developers who have a stake in making this a common feature” or “user surveys that are heavily loaded to arrive at the conclusion I want “

Could be I’m just an old man yelling at clouds though.

You have to get licensed to drive a car and there are rules. If you are caught breaking the rules you can be fined (or jailed if your rule breaking harms others enough). In extreme cases of rule breaking you can lose permission to drive the car.

This is all friction that is intended to make car-driving work at scale.

In 1978 loans were made non dischargeable for the first 5 years and extended to 7 years in 1990. In 1998 the waiting period was eliminated making them non dischargeable in perpetuity. Private loans were made non dischargeable in 2005.

So while student loans were technically dischargeable approx 28 years ago there were some big caveats.

They do not. This is nonsense.

Not that I want my kids looking at porn or violent content, but I’m far more concerned about man-o-sphere influencers than that other stuff.

I don't think anyone is expecting to seriously "derail AI" just by preventing data centers from being built in their neighborhood. But they can at least inconvenience the people they see as the perpetrators. If you can't win the fight, and can't walk away, at least you can give your opponent a bloody lip. And who knows, maybe it gets steam and the people in congress who have been actively ignoring them have incentives to at least throw them a bone.

If there were serious macro consequences then maybe it would be an issue but from the point of view of the communities -- there aren't. Data centers don't bring a lot of permanent employment and tend to be given tax breaks. They are skeptical that data centers are the boon that these people claim they are -- and they are right to be so since people pushing these data centers have been wildly untrustworthy at best.

Saying "You can't stop it so you might as well get on board and hope that in the future you can convince shareholders/billionaires/US Congress to give you a pittance of UBI" is going to go over like a lead brick. There's no reason to trust that UBI will come if you give up any leverage you have, even if it is a minuscule amount of leverage.

I'm not sure why this is relevant:

> This doesn't make sense. The reason Congress is difficult is because of those same powerful people.

Its much easier to put someone who is aligned with your values into a local political position than congress. And its much more likely that your neighbors will vote in a way that aligns with your interests. And you won't get overridden by a congressman from several states away that has different incentives.

Yes, people building data centers can just shop for a new location. But resistance to data centers appears to be pretty correlated with living in communities that are good places to data center, at least anecdotally.

The world I'm looking in is one where citizens pushing back locally has seemed to get at least some measure of success, albeit spotty, whereas attempts to lobby Congress about AI has been screaming into the void. Frankly, I think your position here is completely divorced from what is actually happening in reality.

> This is fair but I mean as long as people are agitating politically it seems like going after the data center is going after the wrong end of the equation, with similarly difficult odds of success.

I don't see why you think that. Its something that:

1) These CEOs and people with power want

2) The populace has some degree of control over, since its local politics vs national.

That makes it an attractive way to push back against powerful people that they see as operating in bad faith. It seems like their chances here are much better than trying to go directly to congress and advocating for UBI.

> if there is UBI?

At least in the US, the public simply does not trust that the United States government will consider such a thing. They won't even consider universal healthcare. No-one is going to go "OK we trust you, you can build your data-centers now and we'll talk about UBI once you've 'disrupted' our jobs."

Yes, a bunch of CEOs are making the rounds talking about it, but talk is cheap. Even if that talk is directed at congress. Have any of them even cleared the flow bar of funding research into how it'd work and what the policy would look like?

This has an incorrect baked-in assumption that arguments should be equally weighed based on political leaning. But sometimes arguments are just wrong -- the right wing argument on climate change was that it was a hoax or not human-made for a long time. A properly truth-seeking AI would not present that argument as valid or supported.