> It could be IQ, cultural-specific, polarized against authority, much of which deserve monitoring.
It doesn't seem so confounding to me. If you were unemployed and needing assistance, do you think you would be happier or less happy having someone require you to report if you've got a job yet and they can take the income away?
> I'm not against AI summaries being on HN, however, users should verify and cite sources so others can verify.
I don't see how they contribute anything to a discussion. Even a speculative comment organically produced is more worthwhile than feeding a slop machine back into itself. I don't go out for coffee to discuss LLM summaries with friends, and I can't imagine why anyone would want to do that here.
Earlier today I asked Gemini Pro to find information on a person's death that was turning up nothing for me otherwise, and it just imagined finding verbatim Obituary quotes in every source, cobbled together vaguely related names, plausible bits and pieces from wherever, almost like it was 2023 again.
It ain't search, and it ain't worthwhile; I'd much rather someone ask an llm the question and then post a question out of curiosity based on it, but without the summary itself
> It is search if you ask it to produce a list of links.
Not in the example I mentioned. It can imagine the links, and the content of the links, and be very confident about it. It literally invented an obituary that didn't exist, gave me a link to a funeral home that 404'd, came up with "in-memoriam" references from regional newsletters that never contained her name. It's actually really scary how specifically fake it was.
I asked it to produce verbatim references from any sources and the links to them, and none of the text it produced could be searched with quotes on any search engine.
I think that's the tricky thing. I'm not saying it's not useful when it is, but you really do need a keen and skeptical eye to be able to know. The problem kind of reminds of bloom filters, such that they're useful for situations when you want to know something might exist or definitely does not exist in a set. Exact truth has some level of permissible error rate, as it does in any situation, but definitely wrong is pretty important to know about.
The issue as I see it is just straight copy/pasting its output. You want to use it as a search tool to give you pointers on things to look up and links to read? Great. Then use that as a basis to read the sources and write your own response. If you aren't familiar enough with the subject area to do that, then you also shouldn't be pasting LLM output on it.
It's not even copy/pasting in some cases. In my example, it confidentially produced "verbatim" references that don't exist anywhere, to specific pages that never mentioned this person's name or contained any of the text. Sometimes completely different people, 404 pages, huge waste of time
Yeah I agree, I've seen the hallucinated references also. Sometimes used by people in internet arguments to make their bullshit seem more legitimate. What I meant though by copy/pasting is people getting the LLM output and then just directly feeding that to the people they're conversing with, instead of looking into what it's saying or really even engaging with it in any way.
Ask it to solve a tough Euler Math puzzle with the search button on and it just copies the answer from the web. Turn search off and it actually computes the answer.
Funny how the search button is taken away though.
Which is "fine" so to speak. We do this with using AIs for coding all the time, don't we? As in, we ask it to do things or tell us things about our code base (which we might be new to as well) but essentially use it as a "search engine+" so to speak. Hopefully it's faster and can provide some sort of understanding faster than we could with searching ourselves and building a mental model while doing it.
But we still need to ask it for and then follow file and line number references (aka "links") and verify it's true and it got the references right and build enough of a mental model ourselves. With code (at least for our code base) it usually does get that right (the references) and I can verify. I might be biased because I both know our code base very well already (but not everything in detail) and I'm a very suspicious person, questioning everything. With humans it sometimes "drives them crazy" but the LLM doesn't mind when I call its BS over and over. I'm always "right" :P
The problem is when you just trust anything it says. I think we need to treat it like a super junior that's trained to very convincingly BS you if it's out of its depth. But it's still great to have said junior do your bidding while you do other things and faster than an actual junior and this junior is available 24/7 (barring any outages ;)).
Bit of a tangent: I don't really subscribe to the introvert/extrovert divide personally, but do eventually hit a wall with socializing, and am happy to explicitly isolate myself in my own world or with a smaller group for extended periods of recharge. Unfortunately, I've committed to attending my good friend's costume NYE party, and have betrayed myself somewhat because... I'm just tired of costumes, he's a very theatrical film person and I'm... a web dev, who's just never really had an affinity for dressing up in that way—even less so since it's been a socially packed autumn. I'm considering bailing, but I feel like that would be a bit of a fail.
I think as a nerd, I'd need to make it a challenge and a small hobby like you have, but I also am trying to quit YouTube. Can you picture yourself in my situation? Any tips on finding a seed of interest?
This is a bit longwinded, so apologies: I tried sculpting because I saw a video on YouTube where this guy, I think he goes by Craftyart, or Craftyarts - he had a speed video where he sculpted, cast, and painted a version of the Joker, but it was Willem DaFoe. It was incredible, and it just gave me an itch. I watched it and wanted to do that, to make that.
For me, I'd often have these ideas of things I wanted to try, or do, or challenge myself with, and then for some inexplicable reason I'd never do them. In this instance, I told myself to get off my ass and just give it a try. It may have helped that I was in therapy at the time and making efforts to address a lifetime of issues. It has lent a certain proactiveness to my being. For me, addressing my mental health is a driving factor in having made any of this possible.
Finding a seed of interest: if you mean directly with making a costume, I don't know. If you're not interested in costumes, I don't think it is something you can force. Overall though I think anything that causes that itch, that pull, maybe even a sense of yearning "to do" is enough to get you going on a path. I had a feeling when watching the video that reminded me of what I felt when I was a kid and I would see something and I'd get excited to do the same.
I don't know that any of this would have come together for me had I not been on a journey to improve my mental health, and making efforts to find something that connected with me. Something outside of a screen. But in the end, what I connected with was surprising. It looks like it makes sense in hindsight, but at that time, it felt like it came out of left field.
Hopefully there are some tips somewhere in this mess of words. If not, my apologies for wasting your time.
> For me, I'd often have these ideas of things I wanted to try, or do, or challenge myself with, and then for some inexplicable reason I'd never do them.
I think without a mentor or point of reference for why or how you'd go about doing something like that, it's just a completely abstract domain, much like software is to anyone who hasn't spent a lifetime coding or figuring out how computers work. The mental health work and the video by Craftyarts seem like the perfect timely combination to allow for peeling back those layers, literally and figuratively, further allowing curiosity to be actionable.
I've been doing that a bit with electronics, and a recent example that seems similarly daunting for me would be watching the end to end process of building a custom keyboard pcb. At first it seems like an immense rabbit hole, but dedicating a bit of money and time incrementally is insanely rewarding in aggregate, moreso the further away from your mainline discipline it is. I tend to avoid these until I have a specific challenge in mind.
The seed of interest question was framed poorly, but it was related specifically to the latex mask subject, and I guess I was just curious if there were any adjacent ideas that might be worth exploring, since you do seem to have an interest in vaguely related areas
Adjacent ideas: 3d printing - I think it is cool but I'm not into it myself because it would require more time for me in front of a screen and working with digital tools. But, there are a lot of things to be explored here. One idea is digitally scanning a analog sculpture and then 3D printing the mold for it. This would be huge as you could make very complex molds and they'd be essentially perfect. And then when your mold broke down you could just print another one. No need to work from a master sculpt.
I have a friend who made me 3D printed keychains for swag at events. He embedded NFC chips in the keychain and this links to a linktree on my website. Coming up with cool swag like this could be something to explore. People found them really cool and it was a relatively simple thing to do. I'm sure there are some wild things 3D printing could be applied to for things like this.
1. If you have that urge to go, it is probably for a good reason, agreed :) Wouldn't call it a fail though if you didn't end up going. We all require balance.
2. Parties are for getting together, costumes are just a dress code. They'd love your company even if you didn't dress up -- that's why they invited you after all. So don't stress over it. You can come in something silly or minimal fuss.
> They'd love your company even if you didn't dress up
This is generally true and reassuring, and but in this case I have to put at least something reasonable together since I half-assed it last time lol. I'll probably just try and attend every third costume party in the future
> If it turns out you can't copyright code that was generated with the help of LLMs a whole bunch of $billion+ companies are going to have to throw away 18+ months of their work.
Hmm, it is interesting to think about that situation. Intuitively it would seem to me like there's some nuance between whether work would need to be "thrown out" or whether it just can't be sold as their own creation, marking some kind of divide between code produced and used privately for commercial purposes vs code that is produced and sold/provided publicly as a commercial product. The risk in doing the latter, or entirely throwing out the code, seems like it would be a relatively cheap risk that those companies do anyway all the time.
However, if I as a small business owner made a tool to help other businesses based on LLM code that used some of my own prior work for context, then sold the code itself as a product or sold a product with it as a dependency, it would be a much greater liability for me if it turned out to include copyrighted && unlicensed work that was produced by an LLM that further can't be claimed as my own.
Privately, on servers or in internal tooling not sold commercially, it would perhaps be next to impossible to either identify or enforce those limits. Without explicit attribution to an agent, I have no idea (with certainty anyway) which code anyone on my team has produced with an LLM, and it's not available publicly—aside from pure frontend web stuff—so I wonder in what capacity it would even be possible to throw specific chunks out if it was hypothetically enforceable.
By your response, it really seems like you read their first sentence as advocating for vibe coding, but I think they were saying something more to the effect of "While it's exciting to reach those milestones more quickly and frequently, as it becomes easier to reach a point where everything seems to be working on the surface, the easier it then is to bypass elegant, properly designed, intimately internalized detail—unavoidable if written manually—and thus when it comes time to troubleshoot, the same people may have to confront those rapidly constructed systems with less confidence, and hence the maintenance burden later may be much greater than it otherwise would be"
Which to me, as a professional SWE, seems like a very engineer thing to think about, if I've read both of your comments correctly.
Justin Trudeau actually appealed to many younger Canadians as well, until he severely didn't, but that's how he got elected afaik. I don't think the numbers are out yet, but it was largely boomers who voted Carney in, everyone else that I've spoken to under the age of 40 is like "meh", and some regret voting him in after the Liberals' predictable reaction to the Air Canada strike.
Much like the rest of the g7, we have an aging population and a mega generational class divide. Our youth unemployment rate is high, jobs have dried up, it's a shitshow, and Carney hasn't tried to address this much.
So whether he's popular or not needs more context. He'd certainly be most popular with the richest and most populated generation ever, and potentially business owners, but we'll see.
Considering how much government spending goes to the elderly, either directly via programs like OAS and tax benefits or indirectly via healthcare, and it was only a matter of time before young people question their position in it all (higher schooling tuition/debt, bad job market, expensive housing, etc). OAS doesn't even start getting clawed back until personal income is over $90K and is only fully clawed back at $150K! And it's double that for a couple!
The timing of the last election was perfect for Carney when there was a window where the whole country was going WTF with Trump and PP was still railing against various "woke" grievances and mentioning Trudeau by name. The fact that he wasn't turfed after not only losing the election that was his to win, but also losing his seat, is everything wrong with the myopic federal Conservative Party (whose core members refuse to "compromise" with the rest of the country).
There is a real generational tilt happening and young Canadians no longer defaulting to left leaning ways of thinking (not that they ever were as much as people thought).
Yep agreed on all points. I like that there are a few local orgs (GenerationSqueeze, Missing Middle) bringing light to things like the portion of the federal budget allocated to OAS and how it's structured, and generally being real about present day inequities.
Carney will (hopefully) have to reckon with those in the coming year, while Pierre seriously missed a (the?) boat. It does feel like something big is shifting slowly.
> There is a real generational tilt happening and young Canadians no longer defaulting to left leaning ways of thinking (not that they ever were as much as people thought).
It's my impression that the balance between economic prosperity and social good needs to be constantly curated and revered as an inherent virtue of a democracy with strong social safety nets. It's much easier to get working age people to compensate for the ails of generations past if there's no doubt in their mind they'll have a roof over their head next year.
Progressive, often barely tangible issues, necessarily become internalized as luxurious if the people who could support them can't even pay for food.
Using nearly the exact same rhetoric and party platform, with the exception of austerity measures. Only difference really seems to be that he knows how to pass a budget, but hasn't really offered anything substantial in terms of hope to the working age population
This is what my nostalgia for native macos editors rests on. I've wanted to buy Coda despite VSCode and other derivatives being more productive, and where would editors now be without BBEdit, Textmate, Espresso/CSS Edit, which all did particular things very well, given the constraints at the time.
> Does this new ban move kids to using email to keep in touch with friends and family?
You had social media but no ability to send DMs?
In an attempt to not deliberately misinterpret you, next to zero of my current ability to keep in touch with anyone in my life via the internet, distant or otherwise, depends on social media, so forgive me if this seems like a strange take. Kids need access to YouTube in order to talk to their family?
> Are they now completely isolated from the rest of the world?
It's only in extremely recent history that anyone, especially kids, had access to the rest of the world in any meaningful way, or at the resolution available now. I don't think it's remotely healthy for adults to concern themselves with the hourly regional issues wherever they're occuring in the world; it costs society a great deal more than it earns imo (but it's very profitable for the companies on this list)
> In an attempt to not deliberately misinterpret you, […] Kids need access to YouTube in order to talk to their family?
Your attempt has failed; obviously I’m not taking about YouTube, but about things like WhatsApp, Telegram, WeChat, Instagram, and other social media which families actually use to talk to each other on a dialy basis.
Perhaps you don’t use these, but most of the world population uses some of these (or something similar) to keep in touch with family and friends.
Heck, even when I was a teen (before smartphones) I kept in touch with friends over social media. We’d even organise meeting up through it.
Out of all of those, only Instagram is on the list (or I have heard has complied despite not being on it).
You seem to be confusing messaging software and social media ?
Do (the others than Instagram) have an algorithmic feed, or (effectively) do not work without making an account ?
I guess that there's also Discord (that isn't on the list but has still complied) that is in an awkward in-between ?
(IMHO both Instagram and Discord ought to be banned anyway, for everyone, because they're deep web platforms that are owned by Meta/Tencent, and are therefore a threat to the open web and liberal democracy.)
I think you established too broad of a scope for discourse to be within the parameters you were hoping.
Immediately upon reading your comment, I thought about the general overprotection and over-supervision of kids which leads parents to drive their kids everywhere, prevent them from learning to use the subway on their own, or even live in cities. But what I think you were getting at is more about smaller hypothetical physically analogous places, but it's hard to think about what those places are in real life without relying on assumptions that may be more likely to occur online than in any significant concentration in the real world.
Imo, the most threatening place for kids to be in real life in terms of external factors, day to day, is around cars, bullies, bad actors within the family, and then maybe church/sports teams, but all of those are usually safe unless they're not, you can't realistically do anything productive about that without sacrificing their development as a human, except prepare them and guide them.
Online, it's just a whole different beast, and I'd think it would be games and social media, anywhere a gaurd would be let down, but imo the greater threat isn't criminality as much as it is nearly every other aspect except basic chats.
It doesn't seem so confounding to me. If you were unemployed and needing assistance, do you think you would be happier or less happy having someone require you to report if you've got a job yet and they can take the income away?