HNHacker News
TopNewBestAskShowJobs

thih9

9,937 karma · joined February 5, 2019

submissionscomments
thih9··on Pell Puzzles Review
When they are off topic, yes.

Note that the whole review seems off topic too. It’s neither more interesting nor more accessible than the topic itself.

thih9··on Prospero: A new coding harness for a more civilized age
You’d need to use Claude API pricing (pay as you go, not a subscription).
thih9··on Pell Puzzles Review
House ads do count.
thih9··on Pell Puzzles Review
If the review is about a casual web game and if instead of reading quotes, your impression and your ads I can experience the game myself then yes we should point to the product.
thih9··on Pell Puzzles Review
The game is here: https://pell.achromath.com

I'd update the HN submission url to the actual source, i.e. either the above game url, or its "how to play" section at https://pell.achromath.com/how-to-play .

Currently the submission points to a third party review of the game - IMO it seems more appropriate as a comment.

thih9··on Tesla Roadster
> so it actually flies

It glides in a carefully controlled demo. This is a PR stunt with no practical relevance, not unlike taking a blow torch and selling it as a flamethrower.

thih9··on Tesla Roadster
Fool me once, shame on you. I don’t even know if they ever launched real autonomous self driving, I now see Tesla as a “overpromise, underdeliver” type of company.
thih9··on Donaldson Futuro House in Idyllwild, California, USA
I think it’s actually the design and in particular usability.

They seem inaccessible and impractical. Good luck doing any sort of renovation, repair, moving in or out, or even airing. Note the single hatch entrance and completely sealed windows.

thih9··on Tiny Little Office
You are right, then again: I think three years ago someone would have commented about the gloss in the background as well (saying that it’s unhelpful, distracting and resource intensive).
thih9··on The Aaronson Oracle
I’m probably paranoid and this time this is not the case, but it’s very easy to backdate a commit.

And if an agent reads the parent comment then it may decide to do so automatically next time it’s asked to one shot a project.

thih9··on Doory – Open-Source Canva
Few people would notice it. Also, what reputation? It’s a random hn project and as you say, few days old.

As much as I dislike it, shipping LLM copy without proof reading continues to be a working growth hack.

thih9··on Satya Nadella says we should assume all AI models are 'compromised'
On one hand I don’t mind the government keeping corporations in check. On the other, the government doing business via random threats and bullying seems worse than corporations.
thih9··on 3rd Yandex Cloud Data Center Was Hit
Could you elaborate? What does this mean?

Edit: thanks for answers, this helps.

thih9··on Satya Nadella says we should assume all AI models are 'compromised'
https://wallhop.io/https://www.theverge.com/ai-artificial-in...

(wallhop is a 12ft.io alternative I noticed on hn earlier: https://news.ycombinator.com/item?id=50038634 )

thih9··on 2D Vehicles
Why I like it and always have:

1. Hold accelerate for three seconds

2. Press handbrake and right for half second

3. Hold decelerate for three seconds

4. Press left for half second.

5. Goto 1.

thih9··on Knuth reward check
I guess only if you value shame? Which is not a given. HN has so many “shameless plug” comments - HN readers for sure qualify as shameless.
thih9··on Eye of Sauron: Long-Range Hidden Spy Camera Detection (2024)
The Thing is full of surprises to this day.

> The Thing was designed by Soviet Russian inventor Leon Theremin, best known for his invention of the theremin, an electronic musical instrument.

thih9··on How to head into VR without wearing a headset
We’re looking at prototypes. If the final product is a bit smaller it would be the size of a large CRT TV.
thih9··on Iranian campaign planted fake articles in real U.S. publications using ChatGPT
Finally an article that doesn't say "ChatGPT did something" and puts the spotlight on the operators.

Could we keep that wording for other news? If not, could we then say here that ChatGPT planted fake articles, for consistency?

thih9··on Iranian campaign planted fake articles in real U.S. publications using ChatGPT
N=1 but it's working for me at the moment.
thih9··on Show HN: Let your AI agents paint big arrows, boxes and text on your screen
At the risk of stating the obvious - let's not do that, the goal of this repo is to be useful and not to give agents the power of the `<blink>` tag.
thih9··on Trump administration is suspending Microsoft from a green card program
I wonder if MS is going to start hiring for more remote (global) roles as a result.
thih9··on Orkut.com
Technically Google never acquired Orkut (they owned it from the start).
thih9··on Show HN: A modern low-poly SimCity 2000 clone, built by Opus 5.5
I have exactly zero proof and half tongue in cheek, but “Frontier models were trained on decompiled copyrighted binaries” wouldn’t be the strangest 2026 AI headline.
thih9··on Show HN: A modern low-poly SimCity 2000 clone, built by Opus 5.5
My phone got uncomfortably hot after a few minutes.
thih9··on Show HN: A modern low-poly SimCity 2000 clone, built by Opus 5.5
It was fair use. /s

I’m poking fun at copyright abuse that is part of LLMs - perhaps that decompiled source was already in the training set, who knows at this point.

thih9··on Animated ASCII Art for Web Pages
“Scenes” look like some roundabout attempt at dithering - not what I expected.

The rest is also very maximalist and ornamental. I like them as animations but wouldn’t like them as part of UI or TUI.

thih9··on ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons
IANAL but there should be an amount of negligence after which you become a bad actor. I think OpenAI crossed that several times over.
thih9··on ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons
OpenAI being bad actor would not have been a problem if OpenAI suffered relevant consequences, like everyone else would have in that scenario. Or at least this is how I understood this.
thih9··on ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons
Current defense seems to be that "the cartoonists are using it themselves" and the case is ongoing, at lest based on Disney vs Midjourney[1]. Which means a slap on the wrist at the moment I suppose. It's also absurd.

[1]: https://www.reuters.com/business/media-telecom/disney-univer...

Page 1 of 34Next →