_heimdall

Well then you're taking the opposite view. Your argument wouldn't be to stop data centers due to water use, it would be to secure water use for clean drinking water first before other uses are considered.

That angle opens the argument to concerns from various fronts, including data center use but also concerning things like agricultural use if that's a concern for you.

The drinking water concern also would hinge on having to define "clean drinking water." Humans have made do with drinking from streams, rivers, and ground water for most of human history. Today into he west at least we largely drink recycled waste water treated with bleach and various other chemicals. Drinkable yes, but I'd personally rather drink some rain water than bleach water.

If you aren't familiar with the distinction between correlation and causation in scientific research and statistics I really would recommend you look into it.

Modelling studies can never prove likely causation, at best they prove likely correlation. Meaning, two things seem highly likely to happen together but we can't say one caused the other.

Carbon emissions and carbon in the atmosphere has been shown highly correlated with the increase in average global temperatures, weather pattern changes, etc. That doesn't directly say carbon emissions caused those though, they could both be side effects of a more fundamental root cause.

Said differently, we know carbon emissions and average temperatures after the year 1850 both went up at similar rates. We also know, though, that many other things changed over that time. We can't run a controlled study of the entire environment is impossible, we can't isolate variables and compare outcomes when only one facto changes (in this case carbon).

Its always a very interesting architectural question when deciding at what level(s) the business logic should live. There absolutely are valid reasons for some to go in SQL, though I tend to avoid putting the most complex logic there when its really tricky.

When it really gets hairy, or when the business logic keeps changing under my feet, I'll try to find constraints I can put in the db as a final backstop while leaving most of the logic somewhere in the application stack.

I can count on one hand the number of people I've worked with that really know SQL well enough to pick up complex business logic at that layer and work with it easily. I've been mainly in small companies for the last decade, I'm sure at larger orgs there are more data engineers running around that could own it.

The big challenge with using the platform in my experience is that we have spent north of a decade now trying to make the web into an operating system rather than a document system.

Web components really do well at a small scale, being able to craft a custom input in roughly the same way a browser engineer does is a nice super power.

If I want my entire site to be one big JS application, though, web components don't easily scale there.

I don't like frameworks like react and am happy to have a site or app that is mostly server rendered HTML with a bit of custom JS and web components to fill in the interactive parts. That seems to not be many devs' cup of tea though so they tend to reach for client-side frameworks that turn everything on the page into JSX rendered on the client dynamically.

Oh there's nothing wrong with them offering maintenance, they just shouldn't put up arbitrary walls to stop me from doing it myself.

Yes, those are all well researched and well understood concepts based on a large body of modelling studies. Both can be true at the same time, and just because the data can only support correlation doesn't mean its irrelevant or not useful, it simply means we can't extrapolate to an argument of causation.

I think I may agree with you here, but in this context can't you add more details? Your comment here is pretty high level and not specific to this topic or how we might go about trying to handle it.

I'm not actually sure where you land on climate change being causally linked.

Don't get me wrong, I think carbon emissions matter but primarily because we should simply do better if we can. But I've never found a study that proves causation, at best its correlation shown primarily through modeling.

That doesn't mean we should stop trying to reduce carbon emissions, but it does mean that reducing carbon specifically isn't guaranteed to reduce or reverse climate change.

Surely you must agree thst such a complex system as the global climate can't be driven by a single factor, or a single element on the periodic table.

People shouldn't be latching onto side effect costs of AI data centers like water or energy use, it seems like an abstraction layer above the core complaint. If we don't want AI, or LLMs, or data centers that should be the complaint. Once it is abstracted to water use or similar it opens the door for data to show the actual water use is comparatively small.

Its like arguing about climate change by zeroing in on carbon emissions. The issue people should be raising is wasted energy, over consumption, not knowing "enough", etc. Once the issue is pinned to a single externalized factor the argument lives or dies by that factor.

> may not perfectly map to the parties labeling themselves as such.

Well that's for sure. In the US I don't see the Democratic as liberal, nor the Republican party as conservative based on policies they promote.

"Pro police" and "tough on crime" don't accurately describe a view that is actually pro surveillance.

The two former terms are political in nature because they don't accurately describe what they are supporting and instead attempt to put lipstick on a pig, as it were. Sure you could argue that "pro surveillance" is political, and technically any term can be made political, but it does describe the actions being promoted.

Why would it be okay for traffic enforcement?

The issue with privacy violations isn't how it is being used today, its how it may be used later and the precedent it sets for future invasions of privacy.

Didn't congress leave it as expired around 2020?

Its surprisingly difficult or impossible depending on how they set it up. I had an Chevy with firdt-gen OnStar, disabling it was as simple as unplugging the OnStar box under a seat.

In a first gen Chevy Volt that's not possible, the computer and screens are so integrated with OnStar that physically disconnecting it leaves the car effectively useless.

Auto manufacturers lost me when they began being so hostile to people wanting to maintain or repair their own vehicle.

Much of my 80s model pickup can be disassembled with a basic set of screwdrivers and wrenches. If my Chevy Volt bricks because the battery has an aging cell, it has to be towed to Chevy to pay $200 for them to turn of a software alert and allow it to charge.

I do like electric vehicles but its insane how absolutely terrible they are when it comes time to work on them.

I really dislike such politically motivated names. People do generally use those terms to promote surveillance, but I can 100% support strong policing (meaning well funded police investigations and community outreach) without abandoning my own rights.

I do want police to be able to investigate potential crimes that our society collectively agreed we don't want to happen. I don't want to give up my rights to privacy, protections against search and seizure, etc.

I've often been annoyed by the left/right divide being spoken about as linear. Its always felt more circular to me, you go far enough to the led or the right and its hard to tell the two sides apart.

Authoritarianism as a second axis is interesting, though I don't know what the opposite would be. Communism, the actual idea with effectively no one in charge? If so, in practice those extremes seem to end up in the same place, one or a few people in charge of everyone else with a very tight grip on power.

I don't necessarily live in an area where I expect people around me to behave. Most people do and I have some good neighbors, but its also a rural area with a drug problem.

I also don't want surveillance to protect me though. Its my job to protect myself, and ideally to be part of a community that looks out for each other. It isn't the governments responsibility to protect me if said protection comes at the expense of my own freedoms.

I don't want to be a rat in a cage just so I can feel safe.

That would depend on how you define "pro surveillance." I've talked with plenty of people who support surveillance cameras in public space, are fine with government agencies purchasing private data that would require a warrant for them to collect on their own, and regularly prefer convenience features even when I have tried to share the privacy concerns with how they work.

People can choose whatever they want, my goal isn't to make people want more privacy. I rarely talk about privacy issues with someone and hear them agree or be concerned enough to think twice, they tend to just laugh it off when they talk about something and then immediately see ads for it.

It also doesn't answer the question of how they might even recognize LLM generated code in contributions to PopOS directly.

I have yet to understand how maintainers can't distinguish beyond (1) PRs that literally include Claude co-author notes or (2) low quality code contribution regardless of the creator.

Are you arguing that LLMs can't replace humans at cognitive jobs or that AI wouldn't be able to? If the former I'd agree, if the latter it must just be a definitional issue - I don't consider LLMs to be artificial intelligent and would expect anything earning that monicker would be able to replace humans at most or all cognitive tasks.

I'm not sure how my ability type out a message here on a computer is hypocritical of the argument I made above. My point was that we can't yet know if these turned out to be net positive inventions for humanity, or as you said to everyone. I didn't argue they are bad or that they must be avoided, only that it Amy turn out to be a net negative in the long term.

I must assume that you don't find arguments related to climate change compelling given that they require considering scales beyond 80 years. If you are concerned with climate change I'd be very curious how you square that circle.

Maybe we just have different ideas of what AI would be?

I don't consider an impressive text predictor to be artificial intelligence. Maybe that means LLMs aren't AI, I honestly don't know because no one seems to care about understanding how they work or solving interpetability first.

I do expect anything that earns the banner of AI could ask good questions, weed through a bunch of word vomit to find the key nuggets, and act on them similar to or better than a human could. Anything less than that doesn't seem particularly intelligent.

Correct me if I'm wrong, but I think this argument hinges on AI not really being intelligent.

If we're talking automation in the way that humans define a process, boundary cases, etc up front then yes you can't really automate away one-off monotony.

If AI is actually artificial intelligence, it will be able to figure that out similar to how a human would.

And how do we know that those inventions are actually better for us?

We created nuclear weapons and yes it does seem to have given us 80+ years without another direct global conflict. That's a good thing in the short term, but what about longer? Are we to expect that nukes will never be used? Are we also to assume that the invention of nukes doesn't incentivize people to invent an even worse weapon?

The same goes for all the others. Computers have allowed us to do amazing things, there are also many in the industry warning that they may one day replace us - some even want that to happen. I don't know what benefit you can point to for pesticides and herbicides, maybe if you look at how much tonnage of food is produced and ignore all the terrible side effects of polluting the planet with synthetic chemicals that specifically end life? As far as Pax Americana, maybe its been a good thing? I don't know how to distinguish between the relative peace from the threat of nukes vs the threat of economic warfare, and our K-shaped economy does seem pretty directly related to the incentives created by Pax Americana.

My point isn't that all of these are bad, or that there aren't some good from any one of them (again, except for maybe all the poisons). My point is that we don't know what the results will be on any meaningful scale - 80 years is a drop in the bucket when discussing life on this planet.

I tend to start a project by making a classless stylesheet of sorts to set a baseline for the theme and branding.

Its almost never all you need for an actual project, but it is a nice starting point to get the basics in place.

I find that I look at the endless list of classless stylesheets shared online as a starting point, most have at least one or two interesting tricks worth borrowing. I don't find using the off the shelf ones particularly useful though, it ends up with a very generic look to it like the old default bootstrap site or newer default tailwind site.

Which ones? The often made claim is that LLMs, or maybe some other architecture for AI, will be better than humans at most/all cognitive jobs. Why would one hire a human instead?

What specific wins do you have in mind? Are you leaning on globalization in general? Or products that spun out of war technology like herbicides and pesticides?

For any seamstress that didn't have other economically valuable skills they were able to pivot to, their world did kind of end.

If LLMs can replace most humans at most cognitive tasks, what do those working in various cognitive jobs pivot to?

Nobody has a moat, but everyone is f*&$ed. Competitors do seem to leapfrog each other, but each step is getting closer to beating humans at most tasks. Once that happens, that do we do?