HNHacker News
TopNewBestAskShowJobs

yunwal

2,080 karma · joined June 8, 2021

submissionscomments
yunwal··on Growing tribe of jobless techies is stuck in Silicon Valley's new reality
This is really SDE pay coming back down to earth. You're still making way more than even most white collar jobs at a startup.
yunwal··on Chess puzzle I found in my dad's old book
If there is, it's not listed as the solution on the puzzle website
yunwal··on LLMs Are Not a Higher Level of Abstraction
You literally can contain a logic failure. If I execute logic on my computer that’s not connected to the internet it can’t get out of the box. Done. Contained
yunwal··on Mercedes-Benz commits to bringing back physical buttons
It's impossible to completely evaluate a product before buying. That's why it's important to shame once you have enough experience with it
yunwal··on LLMs Are Not a Higher Level of Abstraction
> but how many unpredictable edge cases are you going to cover?

This is the wrong question. The correct question is what specific subsets of cases do you allow, similar to any security question

yunwal··on LLMs Are Not a Higher Level of Abstraction
> A reasoning error has an infinite, unpredictable blast radius.

Says who? It’s quite easy to limit the blast radius of a reasoning error.

yunwal··on Spotify adds 'Verified' badges to distinguish human artists from AI
> The "song of the summer," the latest TikTok trend

These are 2 cases where you absolutely need a personality to go along with the song. Department store ads are probably already AI.

yunwal··on Approaching Zero Bugs?
> If we assume that we fix bugs faster than we introduce new ones and we assume that the AI tools can improve further, the question is then more how much more they can improve and for how long that improvement can go on.

Why would we assume this? Historically it's obviously inaccurate since the world began with 0 software bugs.

yunwal··on Uber torches 2026 AI budget on Claude Code in four months
do you... honestly not believe that system design is real???
yunwal··on Your phone is about to stop being yours
No, hypothetically apple lets you run arbitrary code. In practice, they lock people out of their developer accounts all the time.
yunwal··on LinkedIn is scanning browser extensions
Reversible encryption wouldn’t be required for fingerprinting. They’re doing something even more sinister here.
yunwal··on HERMES.md in commit messages causes requests to route to extra usage billing
How is it overreaction?
yunwal··on DeepSeek v4
> Uyger-genocide

I'm gonna go out on a crazy limb here and say that this is on par with the genocide in Gaza? Mass sterilization, forced labor, sex, and torture on a larger scale than Gaza. Certainly we can argue about which is worse, but they're both incredible atrocities. The only thing that makes China less scary IMO is that they currently aren't the empire ruling the world and at the center of the global economy. If that changes, as seems likely, I don't see any reason to believe China would be a better or more compassionate world ruler than the US.

yunwal··on Palantir employees are starting to wonder if they're the bad guys
> like the aforementioned free-range eggs

I noted this in another comment, but the "free-range" designation means almost nothing. Hens have access to the outdoors, but that can mean a packed coop with no grass where part is missing a roof.

Look for "Certified Humane" or research the farm directly.

yunwal··on Palantir employees are starting to wonder if they're the bad guys
Just FYI, the designation "free range" on eggs means essentially nothing. It means the hens have access to the outdoors, but that could still mean a tiny, packed space, just missing a roof.

"Cage free" and "no antibiotics" are probably the only USDA-regulated terms worth caring about, but they're fairly low bars. "Certified Humane" designation is a higher, well-audited bar, but many farms that might qualify forgo it due to the costs associated.

yunwal··on OpenAI ad partner now selling ChatGPT ad placements based on “prompt relevance”
I don’t think they’ve been successful enough at monopolizing to get away with this to an egregious extent like Google has. Anthropic and Google both have debatably better models with ad-free platforms (so far). And open models are not so far behind.
yunwal··on OpenAI ad partner now selling ChatGPT ad placements based on “prompt relevance”
how is this different than what OpenAI is trying to do?
yunwal··on The buns in McDonald's Japan's burger photos are all slightly askew
Seems like you just completely ignored the point.

It’s not about sticking it to the rich guys, it’s about charging them an extra 30 cents.

For McDonalds, that 30 cents might be tripling the margin on that item.

yunwal··on I still prefer MCP over skills
How do you shut off particular api calls with an agents.md?
yunwal··on Claude mixes up who said what
The correct word for this IMO is "chaotic" in the mathematical sense. Determinism is a totally different thing that ought to retain it's original meaning.
yunwal··on The threat is comfortable drift toward not understanding what you're doing
Determinism just means you don't have to use statistics to approach the right answer. It's not some silver bullet that magically makes things understandable and it's not true that if it's missing from a system you can't possibly understand it.
yunwal··on Claude 4.6 Jailbroken
Is anyone pretending like models are not vulnerable to prompt injection? My understanding was that Anthropic has been pretty open about admitting this and saying "give access to important stuff at your own risk".

https://www.anthropic.com/research/prompt-injection-defenses

Now, do I think that they sometimes encourage people to use Claude in dangerous ways despite this? Yeah, but it's not like this is news to anyone. I wouldn't consider this jailbreaking, this is just how LLMs work.

yunwal··on "Vibecession" reflects an increasingly difficult economy and society
Vibecession is such a stupid, patronizing term. It’s been so obvious for so long that basic needs like housing are becoming more expensive, and yet economists can’t stop talking about their “basket of goods” as if the way that’s decided upon isn’t just vibes
yunwal··on Claude Code's source code has been leaked via a map file in their NPM registry
> The source code of the slot machine is not relevant to the casino manager.

Famously code leaks/reverse engineering attempts of slot machines matter enormously to casino managers

[0] -https://en.wikipedia.org/wiki/Ronald_Dale_Harris#:~:text=Ron...

[1] - https://cybernews.com/news/software-glitch-loses-casino-mill...

[2] - https://sccgmanagement.com/sccg-news/2025/9/24/superbet-pays...

yunwal··on Swift 6.3
Things that are not simple about go tooling

- Comments as directives[0]. Nobody finds this intuitive. I've never met anyone that saw what directives were doing and thought to change a comment. - Comments as commands for the compiler[1] - The state of go linting[2]. There are 30 different formatting tools that all frequently conflict with each other. It got so complicated that one project exists to manage installations of all the other ones (golangci-lint). Only difficulty is that most IDEs use gofmt and if your golangci-lint uses a different gofmt version than your IDE, you end up getting different formatting when you save vs. when you commit. This is only a problem because golangci-lint can't be installed using the go toolchain[3], so you can't just have dependency resolution solve this for you.

I can go on

[0] - https://groups.google.com/g/golang-dev/c/r4rdPdsH1Fg/m/yjOOz... [1] - https://essentials-of-go-programming.readthedocs.io/tooling.... [2] - https://go.dev/wiki/CodeTools [3] - https://github.com/golangci/golangci-lint/issues/1657

yunwal··on Claude Code runs Git reset –hard origin/main against project repo every 10 mins
Where did they say the title is ai coded?
yunwal··on Swift 6.3
Go and “simple tooling” don’t really belong in the same sentence. Powerful tooling, sure, but simple?
yunwal··on Thoughts on slowing the fuck down
Doesn’t really seem like that’s the way the market is heading though
yunwal··on FCC updates covered list to include foreign-made consumer routers
Incredibly obvious domestic surveillance scheme. Quite creepy
yunwal··on You are not your job
> I think people in general, themselves, are just less interested in the counter cultural ethos.

Gen Z is less patriotic than any other generation, so the counter-culture of their generation is really just... culture. Everyone kind of just agrees that patriotism is a joke, politics are a joke, taking any of this seriously or contributing to it more than necessary is a joke. Gen Z protests as much as any previous generation. The fact that it's not taken seriously in traditional politics is just another signal for them to opt out.

The issue is not that there's no counter-culture, it's that we've designed a system of governance that can withstand it without budging (or at least pretends like it can). Long term, the system will starve. Good luck finding anyone competent to join the military, engage passionately with their jobs, or contribute to civil service. It's not gonna happen. The american dream is now to be on the winning side of a scam.

It might take some time for something else to arise from the ashes of all this, but the first step is already occurring.

https://www.theupandup.us/p/genz-sinking-american-pride-patr...

https://www.yahoo.com/news/marine-officer-explains-why-gen-1...

← PreviousPage 5 of 34Next →