edude03

Most SaaS these days are really about gatekeeping some resource (typically data, or a marketplace) so while you can replicate the gate, it's really unlikely you can replicate the resource on your own. Like random example, I could vibe code a social network, but I can't make people switch to it.

> I worry that we are being over sold on a promise that LLMs will magically make up for a lack of proficiency in an area

I saw a post on twitter about how game devs were using ChatGPT for localization and when you translated the text to English it said something like “as a chat assistant I’m unable to translate this concept” or an explanation instead of the translation.

This is exactly the sort of future I imagine with AI - not that the grunts on the ground will be sold on it but management will be convinced they can fire the people who know what they’re doing and replace them with interns armed with a ChatGPT subscription

More likely because otherwise companies would have no reason to support Safari, they'd tell everyone to download chrome like they do on desktop - then google would have no incentive to optimize chrome for iOS because what would you use otherwise?

Sounds like the roughly do the same thing as ablation - run the network in a way that’ll get the undesired result and multiply it with vectors that prevents it from going that direction

You might have better odds buying scratchers - assuming you even have $400/m to invest in an enterprise without cash flow

First time someone I know in real life has made it to the HN front page (hey sharnoff, congrats) anyway -

I think this post (accidentally?) conflates two different sources of slowness:

1) Building in docker 2) The compiler being "slow"

They mention they could use bind mounts, yet wanting a clean build environment - personally, I think that may be misguided. Rust with incremental builds is actually pretty fast and the time you lose fighting dockers caching would likely be made up in build times - since you'd generally build and deploy way more often than you'd fight the cache (which, you'd delete the cache and build from scratch in that case anyway)

So - for developers who build rust containers, I highly recommend either using cache mounts or building outside the container and adding just the binary to the image.

2) The compiler being slow - having experienced ocaml, go and scala for comparisons the rust compiler is slower than go and ocaml, sure, but for non interactive (ie, REPL like) workflows, this tends not to matter in my experience - realistically, using incremental builds in dev mode takes seconds, then once the code is working, you push to CI at which point you can often accept the (worst case?) scenario that it takes 20 minutes to build your container since you're free to go do other things.

So while I appreciate the deep research and great explanations, I don't think the rust compiler is actually slow, just slower than what people might be use to coming from typescript or go for example.

Many docker users develop on arm64-darwin and deploy to x86_64 (g)libc, so I don't think that'll work generally.

I think the student example is wrong but I'm struggling to articulate why - here's my attempt:

Students often cheat because they claim "they won't need to know this in the future", but school isn't about memorizing facts as much as learning how to research, learning how to communicate, learning to manage their time etc. When I think of someone who's "smart" by my definition, I would expect them to have all these skills that these students are (semi purposefully) avoiding.

Furthermore, - and this might be the sleepwalking bias - the conclusion that "I won't need to know this" is subtly against their best interests, because imagine for example, you spend $200,000 and 10 years to go to medical school, you "cheat" on everything, and pass only to find out, that DRs aren't in demand anymore because AI knows everything (and more that) you would have learned in school, so now the lay person just uses that instead. Wouldn't you have preferred to avoid wasting your time and do something niche that the AI doesn't know instead?

And of course - what are students typically doing instead of learning? TikTok, instagram, fantasy football, youtube shorts - all things that "we as a society" have decided are brain rot.

So it's hard to say that someone who choses things against their best interest for no real upside, who hasn't learned the skills to survive in society is smart by whatever definition

> We tell ourselves it’s substance abuse, so we can pretend it’s their own fault and couldn’t happen to us

I don't disagree that many people disregard the problems of the homeless however I want to be abundantly clear that that I'm arguing against

> by simply losing income and not being able to afford rent anymore.

It's not simply that, things have to be going wrong in your life (not necessarily your fault) that losing your income puts you on the streets. Which goes back to the gradient - for some people losing their income will have no visible impact, for some they'll be under a bridge but which one it ends up being will depend on other factors in your life (having savings, friends, etc)

> but if you give someone keys to a home the problem of them not having a home is solved.

I guess this is the heart of the debate though - is homelessness caused literally by not having a home? And bonus question - would changing zoning laws as OP suggest solve this?

I'm arguing no - more often than not there are contributing factors other than mental health/abuse/drugs that cause homelessness to be your only option that wouldn't be solved by someone giving you keys. Most obviously - food, utilities, is the house close enough to work that you can walk there, who is paying for maintenance of the house, will you be housed with people that are actually homeless because of mental health issues - and will they do things that cause your own mental health to be impacted etc.

> However, what people miss are the tons of people living in their cars, wearing regular clothes, not panhandling and showing up to work best they can.

Sure, this is what I'm talking about when I say there is a gradient. But for argument sake/illustrative purposes though - what would housing have to cost to end homeless? And bonus question - what would zoning laws have to look like to get to this price?

I imagine the answer would have to be "free" but we know for a fact that wouldn't end homelessness because homeless shelters exist that cost $0 and homelessness still exists (yes, even where there isn't overcrowding).

And sure yes, I'm sure there are people that are hanging on by their finger nails - if rent goes up $100/m they'd have to live in their cars but I'm skeptical that zoning law reform is the thing that's going to save them (or end homelessness as the OP suggests)

> If those jobs don't pay enough to afford to live in LA

There's a huge gap between "I can't afford to buy a $700,000 house" and "I live on the streets" though. Renting is an option, renting (or buying) and having room mates is an option. Not living in Beverly Hills is an option etc.

> I will move away from LA") will never be enough to fix it.

I disagree - people generally go towards where the money is. People want to be a hair stylist in LA instead of North Dakota because of the high population density and general wealthiness of the population. During covid, we saw an exodus of people out of the HCoL areas *because* the "non essential" jobs they relied on also moved out (not clear which came first, but the point remains). So generally yeah, if you can't get your hair cut in LA because all the stylists moved out of the city, you'll either commute, or move where everyone else did (which I guess is gentrification in a nutshell)

Either way. I don't think the author is talking about the average starbucks employee here

knative might be what you're looking for

bko in the sibling comment said nearly exactly what I was going to say. While I do agree that you can "fall into" this start and it's very difficult to get out of - I'm very skeptical that with no other contributing preexisting factors you wouldn't be able to figure something out. You'd have to be extremely unlikely to

- not have any family

- not have any friends

- not have any savings

- not have anything they could sell

- not have any ability to do even temp work (IE, had some traumatic debilitating accident)

- Not being able to move somewhere that's cheaper

- Not being able to take advantage of any social programs (I'm canadian, it's pretty easy for example to get a large portion of your school paid by the government)

And if _all_ of this was true, other than potentially avoiding the impacts of being homeless - how would being given a place to live (especially in a High cost of living area) fix any of your other pretty serious issues?

Now, this is a whole other philosophical debate - but why does anyone need to live in LA? I really doubt that people of sound mind are living their homes willingly to be homeless in LA. Sure - there is the crowd that think LA is where they need to be to make it - and I'm sure more than 0 of them believe in the grind so much that they do decide to be homeless on purpose HOWEVER (with no data to back it up) I'm confident that's not the majority of homeless people in LA.

I'm saying of sound mind not as pejorative to be clear but I think for most people of "sound mind" they could figure something out that's not living on the streets (getting roommates, moving somewhere cheaper etc)

I think anyone who has actually been hands on trying to improve the lives of the unhomed know that 1) it’s a gradient between “I get kicked out of x everyday and go back at night” and “I’m sleeping under a bridge” and 2) rarely is it the actual cost of the home that’s the problem. Often it’s mental health / substance abuse / lack of a support system. Heck I think there is enough movies/shows about how people fall through the cracks that if you really cared you could learn about the problems in a fun night at home watching Netflix.

So no hate to the author but this feels like pointless political posturing

Could be, I've seen a weird trend of using AI to write content when it doesn't make sense. Sure, using it to write a blog post about a topic is "slop" but I can see arguments for it. Using it to improve thoughts you have in your head, by making up details and add emojis however, I can't understand.

For example as a heavy FB Market place user I see a lot of stuff like:

[picture of an iPhone 12]

- iphone 14 - new battery - delivers to [enter your state here] - comes with [enter accessories it comes with]

Like they were too lazy to even fill in the brackets or ensure some level of accuracy. What's the point?

You know counting calories by taking a picture literally doesn't work right? In fact, I'd argue it's one of the best examples of when an LLM confidently feeds you misinformation. An obvious example, a chicken breast fried in oil and a chicken breast air fried look identical, yet the oil fried one would have 3-4x the calories at least. Don't take my work for it though - this was just on HN the other day: https://news.ycombinator.com/item?id=44220135

Sorry, should have been more specific - I don't care if people know what I downloaded - I just wonder how stuff I definitely didn't download is attributed to my IP address. Can't be that my dynamic IP address changed, because of the small time between stuff I did download and stuff I didn't. So then, is the scanner wrong, or do I have rouge device(s) on my network?

huh, weirdly it has stuff I did download and stuff I didn't download within minutes of each other ... should I be worried?

I asked a friend who works on datacenter this so it's secondhand information but - apparently it's cheaper to use mist cooling where you essentially spray cold water on hot parts and it evaporates off instead of the normal closed loop system that you'd see in a home water cooling for example. And while it's not destroyed, it means you need a constant supply of clean water, and the evaporated water doesn't necessarily go back to the body of water you got it from.

Hypervisor yeah but it’s still possible for code in debug mode - there is even an app that automates the process for you https://github.com/osy/Jitterbug (requires a paid dev account)

Can’t you already do that? If you pay 99/year you can sign your own apps with whatever entitlements no? You just can’t submit them to the Apple Store for obvious reasons

Where did you get that idea?

Basically this - network effects are huge. People will definitely by hardware if it solves a problem for them - so many people bought blackberries just for BBM.

I feel vindicated! I'm building a tool with VLMs and I've noticed the answer is always what I expect to see, but wrong if the input is slightly different than expected.

Just like the article - if I have picture of a cup, it says cup, if I have a picture of a dog, it says dog, if it's a dog with a cup, it says a dog with a ball (noticed this with Qwen and InternVL).

> Maintaining it myself?

You say that like it's an absurd idea, but in fact this is what most companies would do.

I built something like this a few years ago. Essentially how it worked is you'd specify the path to a nix flake in the kernel parameters, and once the installer was booted it would pull that flake and run it (which ran the installer, formatted the disks and restarted the system). It was part of a larger project but I wonder how the nix team would feel about a small binary that would be included in the default installer image that does this

I know this isn't what you're asking for - I wasn't either - but I found a used surface pro (arm or x86) is better for this use case than I imagined. They're so cheap used on eBay or FB marketplace that I think it's worth trying if you're already willing to buy a new iPad anyway.

I have two now - the SPX - they're ~$200 used, with LTE and 16GB of ram, and a SP8 - i5/16gb of ram ~$350 used from FB marketplace. The SP8 runs Fedora 40 and it's light enough that I just keep it in my backpack whether I'll need it that day or not.

I don’t know anything about IMS but I assume they have to stay on the call long enough for the debug headers to be sent (like the tracing the call thing in every spy movie but real) and if that’s the case can this be mitigated by “just”* not answering calls from unknown numbers?

*yes I’m aware that means people you know who have your number could also exploit this