HNHacker News
TopNewBestAskShowJobs

bwat49

347 karma · joined March 23, 2017

submissionscomments
bwat49··on US Military had close call after using AI for hallucinated intelligence report
Why the shipping manifest details were load bearing:
bwat49··on GPT-6 Astra
Its time to move on to the flamingo on a unicycle bench
bwat49··on I Think You Might Be Fooling Yourself with AI
I've been seeing a lot of this in online discourse where people are like "the bubble's gonna pop and then this will all be over!"

it's a combination of wishlful thinking and straightup delusion to think that AI will just poof and disappear (maybe the specific LLM approach eventually, but not the general idea of AI/the industry/product as a whole)

bwat49··on Egg consumption inversely correlated with Alzheimer's
So one of those egg council creeps got to you too huh?
bwat49··on Since Linux 6.9, LUKS suspend stopped wiping disk-encryption keys from memory
I shutdown mine too but only because suspend is still a crapshoot on linux
bwat49··on Claude Sonnet 5
this seems rather counter-productive, wouldn't a model with less cybersecurity capabilities be more likely to produce insecure code? Not to mention, Chinese models don't have these restrictions and can be used to exploit said unsecure code.

I supposed I shouldn't be surprised at how the trump admin is approaching AI regulation, counter-productive is really all they do

bwat49··on There's still no point in gigabit broadband
downloading a game on steam can fully utilize gigabit speeds, and modern games can be 100+gb
bwat49··on Show HN: Files.md – Open-source alternative to Obsidian
I've vibe coded a codemirror table editor (as a plugin for joplin) using a nested codemirror instance for the cell editor, in my case using joplin's built in command to render markup for the inactive cells: https://github.com/bwat47/joplin-rich-tables

it works well but I'm not much of a programmer and probably ended up with a crazy over engineered architecture lol

I was surprised that there weren't any existing libraries for this (there is one now but it was created after I started work on my plugin: https://github.com/ckant/codemirror-markdown-tables)

bwat49··on Show HN: Files.md – Open-source alternative to Obsidian
S3 as well, I sync with cloudflare r2's free tier and it works great
bwat49··on DeepSeek V4 – almost on the frontier
> "actually no", "hang on", "wait that makes no sense"

Claude does the same thing, claude code just hides the thinking now

bwat49··on Claude Code refuses requests or charges extra if your commits mention "OpenClaw"
well love or hate them, their service is at least reliable
bwat49··on Where the goblins came from
gpt also loves talking about handwaving, "I'm going to do X, not just a hand-wavy victory lap"
bwat49··on Changes to GitHub Copilot individual plans
its showing claude code as included in pro now on that page
bwat49··on Measuring Claude 4.7's tokenizer costs
in copilot I find it hard to justify using opus at even 3x vs just using GPT 5.4 high at 1x
bwat49··on ChatGPT for Excel
its baffling how badly microsoft has handled copilot, this is exactly what copilot in office should have been
bwat49··on Introduction to Obsidian
no merging, they just show as sync conflicts
bwat49··on F-15E jet shot down over Iran
It does matter because oil is a global commodity, the fact that the US is a net exporter doesn't stop the prices from going up and other follow-on impacts to the global economy.
bwat49··on F-15E jet shot down over Iran
the big difference with Iran is the strait of hormuz. It doesn't matter how "well" it goes if it stays closed and torpedos the global economy

> inconsistent communication

I feel like "inconsistent communication" is putting it lightly, with trump going back and forth between "we won", "we'll take the oil", and "whatever we'll leave" often within the same day.

bwat49··on Windows 11 after two decades of macOS: okay, but also awful
I haven't had issues with malware since windows vista
bwat49··on GPT-5.4
yeah claude is great... but only if you pay $100-$200 a month
bwat49··on Don't make me talk to your chatbot
Yeah a trend that I've noticed in online comments is people taking LLM generated text and just removing punctuation and making it all lowercase. It's like dude, it's still so obvious xD
bwat49··on GLM-5: Targeting complex systems engineering and long-horizon agentic tasks
> Your objective has explicit instruction that car has to be present for a wash.

Which is exactly how you're supposed to prompt an LLM, is the fact that giving a vague prompt gives poor results really suprising?

bwat49··on Simplifying Vulkan one subsystem at a time
not to mention the OP mentioned 22 LTS which isn't even the most current LTS
bwat49··on UEFI Bindings for JavaScript
Your scientists were so preoccupied with whether or not they could, they didn't stop to think if they should!
bwat49··on LLMs as the new high level language
I feel so attacked right now
bwat49··on Microsoft open-sources LiteBox, a security-focused library OS
the $10 plan makes a great backup to claude or codex and the inline completions are nice
bwat49··on Two kinds of AI users are emerging
I think there's some middle ground possible between those two black and white groupings
bwat49··on Two kinds of AI users are emerging
yeah I actually use AI a lot, but copilot is... useless. When microsoft adds copilot to their various apps they don't seem to put any thought/effort behind it beyond sticking a copilot button somewhere.

And if the copilot button does nothing but open a chat window without any real integration with the app, what the hell is the point of that when there's already a copilot button in the windows taskbar?

bwat49··on After two years of vibecoding, I'm back to writing by hand [video]
> there is pressure to ship crappy code

this is true with or without AI

bwat49··on AI Usage Policy
I would hope that actually testing the changes is done regardless of whether or not AI is used
Page 1 of 7Next →