Hacker News: bruce343434

New comment by bruce343434 in "Windows 11's built-in Weather app wastes more than 1 GB of RAM"

bruce343434 — Mon, 10 Aug 2026 10:00:47 +0000

Tis not shared

New comment by bruce343434 in "Qwen3.8-Max: A New Bar for Coding and Cowork"

bruce343434 — Mon, 03 Aug 2026 14:09:52 +0000

Perhaps they just used the exact same random seed everywhere?

New comment by bruce343434 in "The LLM Critics Are Right. I Use LLMs Anyway"

bruce343434 — Fri, 17 Jul 2026 22:11:50 +0000

> last month I spent almost 10k USD on tokens. It sounds so insane

It is. That kind of money would change my life and this guy spends it on prompts that "really force him to think".

I use LLMs. So far, 20 bucks per month has me covered. I hit the limit sometimes. I just take a walk. Forces me to think.

New comment by bruce343434 in "Solod: Go can be a better C"

bruce343434 — Fri, 17 Jul 2026 14:55:16 +0000

C++ compilers do this too, thats why so many to-be-linked symbols start with "Z"

New comment by bruce343434 in "I tricked Claude into leaking your deepest, darkest secrets"

bruce343434 — Wed, 15 Jul 2026 20:17:21 +0000

addressing any of the points would require locking the AI down and making it less general and less "agentic". Your concerns make sense if you look at the AI as an information retrieval engine.

   - why is single URL crawl with 20+ redirects not flagged as problematic and/or aborted?

There could be legitimate use cases for interacting with a website like this that could serve the user.

   - why is a query about a coffee place based on its public URL even seeded with the users' context and confidential information?

Because it provides context in how the agent interacts with the site (in this case to detrimental effect)

   - why dont they just look up the coffee place on a trusted source like google maps and continue from there?

The ai was explicitly instructed to check the given url

   - why is the basic "social" engineering style attack working?

because ai can not separate prompt from information, they share the same input channel. state of the art ai has some amount of "common sense" as to when it is being prompt injected or engineered, but this isn't exhaustive

   - why is the cloudflare impersonation not challenged if the website is clearly not from cloudflare and there are zero references from cloudflare to this website in the training corpus?

because the ai didn't think to check if the website is truly behind some sort of cloudflare product or not

New comment by bruce343434 in "I tricked Claude into leaking your deepest, darkest secrets"

bruce343434 — Wed, 15 Jul 2026 07:53:46 +0000

Prompt injection (or llm social engineering" is fundamentally unsolvable, though with training its effectiveness can be reduced

New comment by bruce343434 in "Every new car sold in the European Union must include a driver monitoring camera"

bruce343434 — Wed, 08 Jul 2026 08:12:45 +0000

the optimal crime rate is non zero and the same goes for traffic death rate, because to bring them to zero would imply a dystopia where personal freedom is seriously limited

New comment by bruce343434 in "Evaluation order and nontermination in query languages"

bruce343434 — Tue, 07 Jul 2026 08:34:37 +0000

    :set {:job "Engineer"}
    :where |(= ($ :job) "Developer")

What is going on there in that where syntax?

New comment by bruce343434 in "Organic Maps"

bruce343434 — Sun, 05 Jul 2026 14:39:38 +0000

Is there a nautical map equivalent of osm or organic maps? One that emphasizes waterways by drawing them thicker when zoomed out like regular maps draw roads thicker? Plan routes over the water? Even google maps lacks a nautical layer.

New comment by bruce343434 in "LuaJIT 3.0 proposed syntax extensions"

bruce343434 — Thu, 25 Jun 2026 16:57:15 +0000

Because you can deduplicate certain parts of the logic which make the whole thing less error prone, such as

    if c
      x=1
    else
      x=2

If I ever want to change x, or refactor this code some other way, its a more brittle process over x=c?1:2

The ternary expression also takes up much less space so there is less of an emphasis on it, this can be a stylistic tool in a programmer's toolbox

New comment by bruce343434 in "Slate EV truck starts at $24,950"

bruce343434 — Wed, 24 Jun 2026 15:42:56 +0000

tldw:

- wider wheels are marginally less efficient but almost negligible

- rims with larger diameter have larger air resistance. Probably because the rim patterns intersects the air in weird ways but it's not explained. Tire walls are relatively smooth, so tires with higher aspect are way more aero-efficient despite having more marginally lower rolling resistance.

New comment by bruce343434 in "Wages in America Are Too Low for the 30% Rule to Work for Renters Anymore"

bruce343434 — Thu, 18 Jun 2026 13:32:03 +0000

“I have seen people technically meet the 30% rule and still feel financially strained,” shares Linda Grizely, a certified financial planner and financial wellness speaker. “The pressure isn’t the rent alone; it’s the combination of rent plus everything else in their financial life."

New comment by bruce343434 in "OpenAI Losses Increased Nearly 8X in 2025, with Spending Hitting $34B"

bruce343434 — Thu, 18 Jun 2026 01:01:15 +0000

Why would the company pay more when they can just not pay more? The only things I can see happening is they might lower prices as competition ramps up, or in general as there is more supply for the same cost.

New comment by bruce343434 in "OpenAI Losses Increased Nearly 8X in 2025, with Spending Hitting $34B"

bruce343434 — Thu, 18 Jun 2026 00:21:31 +0000

Everyone else will be 5% more productive. Then no one is "more" productive. So everyone has a higher output, but the same wages and hours worked. There was only a gap when usable AI first came out, some contractors could do the same quantity of work in less time and enjoy time off or do more jobs. Now the gap has closed or is closing. And using AI now is more about not being less productive than peers who do use it.

New comment by bruce343434 in "A robot is sprinting towards you. Do you want it running on Claude or Grok?"

bruce343434 — Wed, 17 Jun 2026 22:33:55 +0000

I want it to cause a traffic accident. If I'm going down, so is everyone else. I'm already dying anyway. Grok 10000%

New comment by bruce343434 in "1k Data Breaches Later, the Disclosure Lag Is Worse"

bruce343434 — Mon, 08 Jun 2026 07:12:04 +0000

The issue is how easy computers make everything, and how well processes scale with computers. Back in the day to heist data you'd have to physically break in or infiltrate, rummage through files, copy them somehow or just straight up take them. In a briefcase?? How many files can you exfiltrate per day like that?

But on a database it's practically a matter of running a copy command and uploading it or exfiltrating it. And there will always be software vulnerabilities.

Computer processes have no inherent rate limiter to them, and they even allow you to run stuff from a distance.

New comment by bruce343434 in "Did Claude increase bugs in rsync?"

bruce343434 — Sat, 06 Jun 2026 15:18:01 +0000

I agree. I tend to use the gnu "asprintf" which simply returns a properly allocated char buffer with the formatted string in it. And on platforms that don't feature asprintf (windows) you can build your own using sprintf!

New comment by bruce343434 in "Show HN: Soft Body Jiggle Physics"

bruce343434 — Sat, 06 Jun 2026 15:13:52 +0000

The accumulator pattern keeps simulations stable in the face of variable frame rate. Accumulate the frame dt into an accum variable. If accum>fixed_physics_dt then step_physics until false. I'm assuming the author did something like that because it's standard practise.

https://www.gafferongames.com/post/fix_your_timestep/

New comment by bruce343434 in "Did Claude increase bugs in rsync?"

bruce343434 — Sat, 06 Jun 2026 08:44:44 +0000

I'll change your mind:

If you pass NULL as the destination pointer, it doesn't write any string. If you combine this with %n at the end of the format string, you can get the exact length that the output string would be. Then you allocate that, then you print again, into the actual destination buffer this time.

New comment by bruce343434 in "S&P 500 rejects SpaceX, also blocking entry for OpenAI and Anthropic"

bruce343434 — Sat, 06 Jun 2026 08:18:12 +0000

what happens when ALL the stocks have been bought back? what is the natural conclusion? you get extra points if you mention dilution i.e. oops we turned on the ~~money~~ stock printer and your stock is now actually worth less!