HN user
Artifacts have always been delegated away by the expert when possible because expertise is more about "planning" and less about "doing".
I'm glad you brought up Jevons Paradox, because it's for this reason I believe it's an extremely poor description of what's happening with AI.
The broader topic people should be discussing is not Jevons Paradox, but marginal utility.
More artifacts have diminishing utility.
LLM is devaluing security researchers' work
I'm confused by this take. In general, LLMs don't replace the expertise required to use an LLM.
Knowing the right input to give the LLM is not enough, nor does it ever end. The LLMs themselves need to be continuously updated and/or specialized for use case, and the "right input" is a moving target.
Keeping both ends of this aligned is the exact same expertise we already had before LLMs. These implementation details would only impact the credibility of those researchers who don't adapt to using this new tool. Who would stumble over something as trivial as this? LLMs aren't even the most important tool in the toolbox anyway.
I only ever look at HN on a phone, and I'm pretty sure I'm in the majority.
Bad mobile support is a sore spot for any "Show HN" announcement.
Now there is a reluctant acceptance and admission that yes, people are building something useful every now and then
Those are the people who were already doing so without AI.
I've watched the headbanging, stressful, but stockholmed love of coding be all but abstracted away from me.
This is only true if it was ever that hard for you to do the job while getting paid for it.
As long as you keep on stubbornly insisting that this wave of "AI" is anything at all like "AGI" or ever heading towards it, you're going to stay wrong and be pushed out of the profession because nobody can reliably trust you or your judgment. Your value as a developer was always your judgment. The code has always just been one artifact of that.
You're still expected to own every line and defend it. It will never be abstracted away because it is the primary contract between business needs and the machine. It is at once both the operational plan and its execution. The only people who insist that it will/should be abstracted away are pushing propaganda to discourage the continued power shift towards the more technically literate who can and will argue for what actually provides value instead of passive investor-pleasing moves. They'd rather you play with the skinner box than meaningfully compete. All this was already set in motion many decades ago. What happens when the next wave of tech comes out? Do we really want you at the helm if you're going to be this naive and easily manipulated? The premise of this blog post is no less delusional than what we were hearing a couple of years ago.
The Keith Haring altarpiece is a really cool surprise!
It's unfortunate that this tour makes it very difficult to see. Really not a fan of this "on rails" presentation style. I understand the angles matter when viewing a splat, but come on.
Entirely dependent on what you're shipping.
If there is a significant amount of user interaction to test, QA is going to do a better job than devs.
If there's a lot of integration behind the scenes with a ton of edge cases, devs will do a better job.
If you have both (probably), you need both teams to pitch in more than upper management wants to believe and they'll just cut you instead! :D
You've definitely used non-toxic and thoughtful products or you wouldn't be making such a distinction.
You are throwing the baby out with the bathwater by retreating to "tools". If you want to stop the enshittification, this is not the way. You're just opening yourself up to new scams. i.e. AI tools, political agendas, etc.
Better products are just better. Demand them.
I do agree that most LMGTFY was obnoxious, and I'm not defending that.
At the same time, we have historically seen a lot of "conversational bidding" lead to low quality shitposting at best. At worst, we see wildly out of touch echo chambers fueled by bots, guerilla marketing, and political agendas.
You did solve your problem by seeking higher quality people to talk to. I'm just saying there is more to rejection than the other party "not understanding".
In the context of work chat or certain events, I could understand a LMGTFY response as steering conversation out of the mud and saving everyone the headache. It also helps you, the recipient of that response, save face.
They're the ones looking like an asshole while you get another chance to try somewhere else. I've met some of the most sociable people upon realizing their usage of these "jiu-jitsu moves" were not meant to be hostile.
I was implying that the answer has to be nostalgia. Ewaste didn't stop being produced some time in the early 2010s. There's always newer ewaste to look at.
You can buy more modern laptops made as recently as 2020 that are at least an order of magnitude better in every single way (even including weight) on Amazon and eBay for about $150 USD. They're lightly used and all hardware is supported without any fuss including the touch screens, etc. They're even cheaper if you buy them in bulk. These are institutional selloffs (schools, offices, government, etc.).
You do have to install Linux yourself though.
Unless you live off the grid, paper notebooks are not purely functional. Neither is living off the grid for that matter. I'm not saying there aren't huge benefits to those things, but they aren't strictly necessary.
People find aesthetic value in keeping a neat notebook. It's been a social media trend for many years now. It has to be the majority of jetpens.com customers.
Why wouldn't a messy notebook be similar?
I was going to comment similar.
I already know I'm going to annoy some people by saying this, but I've gone through both messy and neat phases of life. Every messy phase was almost entirely forgettable and inexperienced noise. Eventually, I'd find stability in my life and become neat again.
I'm not saying messiness is inherently shameful, but it is absolutely worth looking at the bigger picture to find answers. I probably prolonged the messy phases by avoiding self-analysis. I now realize that it was insecurity and avoidance behavior.
Of course, neatness can also be a coping strategy. I'm intrigued by the author's premise.
I'm somewhat enamored with the concept of shitty / messy notebooks, like this or this.
It sounds like they want to explore messiness as an aesthetic, not genuinely experience it. I've already said enough to imply what I think about that. I think deliberately chasing aesthetics might be more detached from reality than what I had going on.
This looks very similar to other homebrew clones from well over a decade ago. What's the point?
How are netbooks dead?
What's the meaningful difference between a netbook and a modern 11-inch laptop?
I still strongly believe Google Gemini has the best position for one simple reason: model maintenance. Accurate information is a moving target.
Open models are indeed very capable, but they will eventually become more specialized to the application to keep an edge. It makes perfect sense that the future shape of AI conforms to the landscape it was born out of.
My mind immediately went to the "Li'l Bastard General Mischief Kit" from The Simpsons.
You're way overthinking this. Making the steering wheel an AC vent is a Mercedes Benz patent. It's just hollow on the inside.
We also used to have crotch vents beneath the steering column! This was arguably the best idea since it could cool the entire driver area including the wheel.
This is artificial scarcity like the McRib.
I think this is a bit more complicated than that. There's more music and video being made today than ever before and social media is how it gets distributed. Attentive time spent on streaming platforms is completely dwarfed by social media. If at all, people have something streaming in the background while they barely pay attention to it and instead focus on their phones.
This aligns with the broader historical trends of the internet creating deep niches. You have to take the good with the bad. We wouldn't even be discussing physical media at all without the internet being how it is. Despite there being a large audience for physical media, they're not the majority. The majority has moved on.
I'm very confused by this take.
It's been over 20 years since then and it's still just as awkward to take a call in public. People will instinctively prefer a quiet place away from the crowd. Otherwise others may eavesdrop, think you're talking to them, or are crazy.
You'll find that most of those people with airpods are listening to something, not talking on a call. The most popular "smart glasses" that I see everywhere don't have cameras. They're "AR" HUDs for watching movies or playing games.
It's not about social acceptance. These hardware designs still suck big time.
a tool that can be extremely helpful at times and extremely sucky at other times but that is still, on average, a time saver?
I can tell you that while I use AI, it's basically a compressed version of stackoverflow to me. I don't bother with agents because I simply don't have that much code to write. The nature of code didn't change. Less code is better code, and it's a strong smell when people claim that they needed an agent to save time.
When I do have a new project, there's an extremely high chance that I can reuse existing code from another project that already has passing tests, or I am experienced enough to know what exactly needs to be done. I can probably finish my part within the sprint or faster. The agent can only introduce risk to the project timeline or code quality.
This isn't me being arrogant or "anti-AI". The hard part of software development has always been collecting consistent requirements from various stakeholders. I'm pro-AI in the sense that, when used correctly, it helps junior devs learn and otherwise shines a spotlight where the real bottlenecks are (project management and delays).
There has been plenty of good faith ruined by trolls on both sides of the discussion. The vast majority of which are new accounts. HN seems to have been under attack for years now.
More generally, the AI discussions that HN regulars actually want to have are going to be limited to the business plans, press releases about new models, and the (very) occasional higher quality project that couldn't have been done without a big lift from an LLM.
Anyone genuinely wanting to discuss the truth from a philosophical or scientific perspective is never going to penetrate the profound ignorance on much of the business side (mostly startups and aging leadership), so why bother? Business is at least what everyone has in common here.
Abstractions don't give up control. They remove options.
In the best cases, those options were redundant or irrelevant to your goals anyway (compilers most of the time).
In most cases, they add mild inefficiencies (OSes, libraries, frameworks, build tools, etc. and sometimes the compiler).
In the cases of LLMs, WYSIWYG, low-code, etc. you're straight up throwing the baby out with the bathwater and setting the house on fire while you're at it too. Such is impatience and greed.
This distinction between control over the outcome and the available options is no longer as subtle as it once was in the bad old days when everyone was more naive. It is genuinely interesting. I'm not wanting to be negative for the sake of it. I actually think we've had glimpses of more reasonable compromises in the highly constrained by committee environments of app/web dev.
There is a degree to which you can retain control with those higher-level abstractions, but it tends to be just as much or more work to maintain the illusion for their less experienced end users. You end up with more scaffolding than building. This is ultimately why we hire devs anyway and abandon those tools.
Most computer programmers gave up complete control some time ago when they stopped writing in machine language and let assemblers, compilers, and interpreters worry about all the little details.
Ah, there it is. The slippery slope that has stubbornly refused to be slippery for many decades now. Perhaps the author is completely misunderstanding these "metaphors".
Do you really have faith this would be disabled by choice if you tried it on a random android phone? If they haven't already disabled it in rage after finding it on accident, I'm pretty sure that person would be terrified upon you making them aware. The first-time setup process isn't consistent across devices and never will be.
You can try to argue this down and reason about it all you want, but it's still a bad default regardless of how it's presented to the user. It's also bad design because it conflicts with similar defaults such as the shortcut to open the camera (press power button twice), silencing an incoming call (press once), etc. People may also press multiple times when the phone seems unresponsive.
You'd expect the resurgence of bad defaults to be motivated by marketing or something, but this one is just outright stupidity.
There's a part of me that is fairly convinced this was added as an afterthought after much internal whining by someone at google with a checklist insisting that android must support some easily accessible emergency feature. This was then deliberately chosen to make the feature look bad, and the users are victims of office politics.
Quickly press the power button 5 times if you have an android phone and see what happens. This has got to be the absolutely worst idea ever and it's enabled by default!
Yes, but are you claiming that LLMs perform more specific acts of cognition beyond organizing information?
The "capabilities you observe" are the actual psychological phenomena at hand here. There's zero chance that branch of research will meaningfully improve the output. That's simply not the point.
This isn't people merely annoyed with repetition. This is the majority of people realizing the limitations of LLMs. Why would researchers give a flying crap about the ignorance of the business world and the public?
People do swap out their expressions all the time. There are influences everywhere that we absorb.
That doesn't matter. The underlying ideas are more important than the words. That's what people are frustrated with. I don't understand why this has to be reiterated for years on end, but LLMs are not intelligent. They just model language.
I'm not claiming to know better than researchers. They do know how they work, and so does everyone else, except you I guess.
The research goals were and still are clearly distinct from the business goals.
Yeah wow fascinating! It's almost like LLM output quality was never the point from a business perspective.
Real people think in concepts and experiences instead of words. The words are not so important to get the idea across, but LLMs only model language.
The problem is fundamental. There's no workaround. Averaging out word usage might even make the problem worse.
Are those complaints really a contradiction? The cost is high for ambiguous slop.