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.
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.
> 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.
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.
Thats a double-edged sword in some scenarios. If you're trying to keep info private then feeding it into a shared agent basically means that you can't guarantee that privacy. I'd imagine the approach here would be to have separate agents for private data and then restrict Slack access, but I could imagine tons of accidents from managers that habitually @Claude without understanding the implications.
> I'm probably in the minority, but I do not want a "connection" with a business. I want transactional interactions that actually work.
Every time a business tries to make a "connection" its really just an avenue to exploit or manipulate me. I've never had them making a connection for my benefit (i.e. take a hit to their bottom line).
I'm not asking for altruism but I am asking for them to drop the pretense and quit bullshitting me.
I find it can often write correct code but not maintainable, performant, or reviewable code without additional human guidance. The "solution" frequently given is that humans don't need to maintain it anymore so its not actually a problem. But the agent can't be accountable for mistakes, so unless that changes or the risk of a defect is close to zero, one still has to put forth effort to keep the code maintainable.
To be fair, there are plenty of situations where throwaway code is perfectly fine and/or defect risks are low enough to make the trade-off worth it. I don't think a lot of developers are thinking about it in that context, though.
It should be but often isn't. There's been a lot of threads on HN where the response to huge PRs wasn't "Don't do that, use AI when authoring better" but "The reviewers are actually the problem, they're missing the AI train". And I see this in industry too.
Its exactly this. I have had a few LLM coding sessions where I reviewed the resulting work and thought "I don't think my team can safely PR this." I then went back and broke it down into smaller PRs, still using LLMs but at a size that is easy to review. And I reviewed the output myself before I asked a reviewer to commit their time.
The problem is that this is increasingly seen as a non-productive workflow slowing everyone else down, so the pressure is growing for writers to just shove massive PRs out the door and reviewers to use LLMs to make that tractable. I suppose those advocates have more faith in LLM output compared to humans than I do.
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.