codingdave

Your categories are broad enough that no matter what future someone predicts, you'll be able to say, "Oh that fits category X". But anyhoo, to answer the broader question:

The truth is that LLMs caught people's attention so strongly that standard tech strategy was thrown out the window. Before this, "tech" companies would test out new tech, tech-adjacent companies would be "leading edge, but not bleeding-edge", and more standard corporations would sit back and wait for the kinks to get worked out before adopting anything new.

Well, now everyone is going for the bleeding edge. AI Everything. Use new models immediately, adopt new ways to use them on practically day 1, churn your IT solutions alongside with that bleeding edge. So it is by definition a bubble. There is a reason people used to hold back, wait, and see how new tech evolves and where it stabilizes. This stuff isn't stable yet, and the pain of keeping up is real.

The future is that people will get burned and will step back from that edge. They'll be OK not being at the forefront, and people will establish best practices for where LLMs truly help vs where they are just a dangerous experiment. That will drive a business model where the AI vendors do not need to solely focus on the race, and can set up longer-living platforms using older models that allow non-tech companies to have stability in their IT shops while still using LLMs. It will become viable to succeed as an AI vendor without having to lead. If history repeats itself, the early vendors will crash out, a new crop will come in to run lower risk playbooks, and hopefully sanity will resume in the industry.

How long will that take is the question - We're seeing the cracks in their business model today, so I think we're closer to 2-3 years out than 8-10, but time will tell.

You just explained why it is not taken seriously, though - your life does not depend on it. Most software is not critical to life. Once you cut out the software that exists purely to streamline capitalism, then cut out entertainment... you've eliminated most software from the picture. The stuff that remains, that drives our infrastructure, cars, anything where a failure could harm someone... that is the stuff where quality matters.

When I worked in IT at a hospital, we even had that as an explicit line drawn between 2 departments - IT supported everything that had zero impact on patient health, there was an entirely different department for anything that could impact a patient. And yes, their work was often simple mechanical fixes to health care equipment, while our projects were multi-year reinventions of patient record systems and upgrades to the physical infrastructure of the network, etc. But we never got confused as to which department was actually more important to the patients.

I'm not saying the attitude is OK. It would be lovely if everyone cared about quality. I'm just saying there is a somewhat legit reason for it.

Who did you talk to to validate the product design when developing your MVP? They are your first users.

Actual Title: Introducing Paper Office

From the guidelines: "please use the original title, unless it is misleading or linkbait; don't editorialize.

https://news.ycombinator.com/newsguidelines.html

Agreed, but that is exactly the point. The mods believe that AI slop is the desired top content in their sub. It exemplifies what is wrong with AI-generated content, as it places people's comments secondary to the AI's summary.

That is an impressive level of tone-deafness to see an automatic AI-generated summary of comments as the top comment on the thread.

I have talked to teachers recently about their options when choosing classroom curriculum and tools. None of them have said a word about AI. Most of them have complained of lack of autonomy because the tools and curriculum are decided at the district level, and the teachers have to use whatever the district sets up.

The reason I started those conversations was specifically around when children are learning critical thinking skills. When we broke down exactly what types of exercises teach those skills, it surprised me to find that it starts way earlier than it used to. It is now a late elementary school topic, when it used to be high school.

That doesn't mean that slop isn't creeping into the classroom - it is. But what to do about it is a completely different conversation when it is pushed down from on top vs. assuming teachers are choosing it.

Nifty idea, but the premise seems flawed. This won't help people read, it will help people avoid reading by focusing on which words to skim. It also seems like it would have a problem detecting sarcasm.

The thing about the icons next to each phrase is that they are useless. Forget how they look, instead ask why they are there? Icons in UX design give a visual representation of what a button does, to help people understand the interaction with the UI. Well, these are posters. There are no buttons. It shows a fundamental misunderstanding of why an icon exists in the first place. (And yes, this is true for many of the places where AI throws in icons.)

That is why, even if the look of a poster is OK, they are failures. Because they forget that the reason these advertising posters exist is to communicate something that is going on. A nice look to them is a bonus, not the core purpose. So claiming that AI isn't doing a horrible job is because the author is misunderstanding what the job is in the first place.

Doesn't seem to work. I tried a few terms that I know have results when using the existing search at: https://hn.algolia.com/, and it returned nothing.

> If coding is a nearly solved problem, if AI is making it so easy to make anything,

If you are open to the idea that those premises are incorrect, you have your answer. Coding is helped, not solved. AI makes it easy to prototype, but not easy to launch a product with production-grade code, support, sales, marketing, ops, scalability, etc.

With that change in perspective, you'd see exactly what we are seeing - scads of prototype-level product attempts that don't succeed when launched into a market.

Actual title: Eliminating the Discretionary 60-Day Grace Period (Thanks - the title has now been corrected.)

This is a "Proposed Rule", and the comments window is open. Use that "Comment" button on the page to share your opinion with the decision makers.

In short, no.

Most of the folks I talk to in my age group are somewhere between anti-AI and "AI works for specific use cases". I find it is the younger folk who think it is miraculous and want to use it everywhere.

Hey, this is my niche - I wrote what was once the leading policy management system for K-12 districts in the USA. It no longer holds that position, but it originated the niche, and it really taught me how policy is created.

Talking to the school board is absolutely a key piece of that, so that was a good move, but talking to the superintendent is an even better move. Most districts delegate the policy work to the superintendent. Most superintendents consult with legal firms who specialize in educational policy. At the same time, most school boards cannot sit down and have a discussion with you about a topic - everything they do has to be in the public eye, so these little blurbs at meetings are the limit of feedback, unless you can get them to put an agenda item at a meeting to have a longer discussion (Which you may be able to do when specific policies are up for reading or votes).

The workflow of how policy really works is that the school board (while being the actual decision-maker) is the last step, not the starting point. The actual starting point is in your state legislature as they set the state-level requirements. Attorneys will keep up on that and let the superintendents know what is needed to be in compliance with changing laws. Superintendents do make their own decisions, but will consult with policy consultants and recommend policy changes to the board. The board will have the policies presented at a meeting, provide time to read them and give feedback, and then finally have a vote on them.

In terms of influence, if you really want to invoke change at the policy level, start at the bottom of that workflow. Talk to your state legislature. Talk to your superintendent. Respond at board meetings to specific policies. And sure, give general feedback at a board meeting. They all help.

(Of course, not all districts follow that model... but most do.)

> Please don't use HN primarily for promotion. It's ok to post your own stuff part of the time, but the primary use of the site should be for curiosity.

https://news.ycombinator.com/newsguidelines.html

> Here's the API key...

> ...that can't really be stopped easily.

Sure sounds like you built in an easy lever to pull to stop it. After all, even if you accept your premise that it will clone itself and run independently, that doesn't give it a clone of the models it is using. Pull the key, stop the process.

Also, remember that there was a day in the early internet when a worm took over, and to fix it the entire internet had to shut down and all wipe it from the systems at the same time. There is always a way to stop things.

Yet another take on this could be: Just don't ever do such an experiment. You wouldn't set up autonomous weapons and set them loose in a city just to see whether they hurt people. So don't do the AI equivalent of such a negligent act.

> all they'll find out

Who is "they"?

Flock collects data which is given/sold to police departments or other entities. Regardless of how you feel about law enforcement having this data, how do you feel about Flock being a middle-man who controls the flow of the data? How do you feel about knowing Flock is VC-driven and will be making decisions based on ROI, not societal benefit?

AI has no autonomous goals or personal ambition. The only way they would "take over" in any sense is if people give them a goal that drives them to that direction and give them access to freely access any system.

What needs to stop is people doing exactly that. The setup of, "Look, I gave my AI a goal and system access!" is exactly the wrong thing to do. As we've seen already with the various anecdotes of people's LLMs destroying their data.

There are a plethora of tools and systems in this world that can do great harm if used improperly. AI is now one of them. We need people to start thinking about them in that context. "This is a tool. There are helpful ways to use it and harmful ways to use it. I should take care to stay on the helpful side."

From the guidelines: "HN is a community—users should have an identity that others can relate to."

https://news.ycombinator.com/newsguidelines.html

Actual Title: "Snapchat's New Business"

https://news.ycombinator.com/newsguidelines.html

"...please use the original title, unless it is misleading or linkbait; don't editorialize."

This question is asked often, and the answer is always that it depends on who your audience is. Channels that work for one person might fail for you, and vice versa. Every product solves a different problem, which means different people who seek out information from different places.

Step 1 is to understand who those people are for your product, and figure out which information sources they trust. As implied above, almost nobody trusts online ads.

> and loads of other huge tech things that are now wildly profitable.

They also said it about loads of other huge tech things that died. Pointing at the ones that made it is literally survivor bias.

This was clearly a popular post, but please do be aware that blog posts are explicitly not "Show HNs".

Actual Title: "Why Smart People Have Bad Ideas"

If you want to quote an essay, then write a blog post and quote the essay. Linking to the reference material, putting in the wrong title, and then commenting about the one sentence you wanted to discuss is not only unclear, it is against the submission guidelines.

https://news.ycombinator.com/newsguidelines.html

> (k) to interfere with or circumvent the security features of the Service

IANAL. But I think reverse engineering their data structure, identifying a security measure - even one as weak as obfuscation, and publishing code to circumvent it is clearly against your license.

> you and I know it misses the point.

wat. I didn't lie about my perspective. If you disagree, by all means, share your thoughts. But that statement puts a really odd spin on your response.

"Asking for a friend" is a great example.

I dabble in art - maybe more than dabble. I have a degree in Fine Arts / Art History, and have been creating art for a few decades. But I don't do it professionally.

With that context in mind, my take on this is that it might help people who don't appreciate art take one step in that direction because you've added more to the experience. Not bad, but not really what drives appreciation either.

I've found that what really makes people appreciate it is connecting it to the people behind it. When they see a work in the context of an artist's life, or the history of the time it was created, and understand it beyond just "look at this picture" and more of "Look at this person, their experience, their feelings, their problems and their joys, and see how they expressed all of that through art." Even if the art was a hired commission, which was historically common, just as it is today, the story of "I practiced a craft and got hired to do this" is still a story.

That really is what it boils down to - story telling. Tell a story, get appreciation. Fail to tell a story, get people who glace at it and respond, "OK. cool. Anyhoo."

That existed for decades in the USA just by dialing '411'. It only went away a few years ago in many locations.

I worked with some of the folks who introduced coding into school districts way back when that movement started, so I have a bit of inside knowledge.

The biggest anecdote I'd share is that learning to code as early as 1st grade unexpectedly led to children taking the ability to troubleshoot problems and using it across all their subjects in school. They started to treat struggles in other classes as debugging assignments: "What went wrong? What do I need to fix?" Their entire attitude changed from pass/fail into optimization. It has led to academic improvements across entire school districts, which was not at all expected but was a delightful surprise.

So the question posed seems fairly irrelevant to me. Maybe some of those kids grew up to be coders, maybe they did not. Either way, it helped their education.