southwindcg - Hackdot

Re:In 1982, a physics joke gone wrong sparked the invention of the emo... on Thursday November 27, 2025 @09:47AM (#46069734)

Emoticons have been around since the March 30, 1881 issue of Puck magazine, if not earlier.

Re:Show HN: What If Your Phone Was Your Bank? (Offline, Zero Fees, Ful... on Sunday November 16, 2025 @03:21PM (#45948088)

Show HN is not for reading material. Also, your GitHub link is a 404 error.

Re:Russia's first AI-powered robot walked on stage on Friday November 14, 2025 @02:27PM (#45930990)

And that this happened with the theme from Rocky playing (a film franchise in which an American underdog beats up a Russian super-athlete), and there were broken plastic robot bits left on the floor... it really, really does seem like a comedy skit.

Re:Higher Thermal Conductivty Than Diamond Found in Boron Arsenide on Friday November 14, 2025 @02:20PM (#45930867)

Specifically when produced with "exceptional purity"...

Wikipedia [0] lists room temperature thermal conductivity at 1300 W/(m·K) and says "Chemical synthesis of cubic BAs is very challenging and its single crystal forms usually have defects." I guess it remains to be seen how cheaply and easily this "exceptional purity" can be achieved.

[0] https://en.wikipedia.org/wiki/Boron_arsenide

Re:Day 1 of Light Speed Up: When transparency matters more than perfec... on Tuesday November 04, 2025 @10:38PM (#45818720)

Your links to the original Show HN just point to a comment.

Re:UBCO study disproves the simulation hypothesis on Monday November 03, 2025 @12:46AM (#45796192)

Yes, this is exactly my problem with claims about the 'real' universe if we are, in fact, in a simulation. It might be literally, programmatically, impossible for us to infer anything about it. The analogy I like to use is Pac-Man believing that the entire universe exists within the confines of a blue-walled maze.

Re:Paris Had a Moving Sidewalk in 1900, and a Thomas Edison Film Captu... on Monday November 03, 2025 @12:39AM (#45796150)

I believe it was H.G. Wells, in his A Story of the Days to Come (1897) and When the Sleeper Wakes (1899).

Re:Creator's sue Facebook over broken moderation algorithms on Monday November 03, 2025 @12:05AM (#45796023)

Actual title: "How an Oregon court became the stage for a $115,000 showdown between Meta and Facebook creators"

Re:Steve Phelps lies about NASCAR for 6 continuous minutes on Saturday November 01, 2025 @02:41PM (#45784178)

Per the guidelines, please don't editorialize titles. Add a comment if you want to do that.

Actual title: "Phelps says NASCAR 'trying our hardest' to settle antitrust suit"

Re:Why do some people play their phones out loud on buses and trains? on Thursday October 30, 2025 @04:50PM (#45765159)

Maybe they're p-zombies.

Re:Samsung makes ads on $3,499 smart fridges official with upcoming so... on Tuesday October 28, 2025 @06:09PM (#45739892)

Only eight years for a dryer is definitely not pretty good in my mind. It's barely acceptable unless you have a huge family and are doing laundry daily. I had a low-end Capri (Sears house brand) that was 21 years old and still going strong when I moved away. It was serviced once, by me, to replace a fuse. If I'd paid twice as much and gotten only eight years out of one, I'd be furious.

Re: on Tuesday October 28, 2025 @05:47PM (#45739660)

Most of them are flagged/killed. You must have Show Dead turned on. While I agree that this user is basically spamming, the community seems to be removing most of the posts. Maybe shoot an email to Dang if you feel the domain should be flagged automatically.

Re:I co-wrote a 1000-page prophetic trilogy with GPT on Friday October 24, 2025 @04:27PM (#45698731)

This is at least the third different account you've used to post this AI slop "prophetic" trilogy here. Please stop.

Re:Show HN: hist: An overengineered solution to `sort|uniq -c` with 25... on Thursday October 23, 2025 @02:55PM (#45685410)

Oh, I meant me, personally, I don't have a use case for it. Without a doubt a lot of people are going to find this speed improvement valuable.

Re:Show HN: hist: An overengineered solution to `sort|uniq -c` with 25... on Thursday October 23, 2025 @01:24PM (#45684433)

I don't [currently?] have a use case for this tool, but I love seeing existing tools made faster or more efficient.

Re:Scripts I wrote that I use all the time on Wednesday October 22, 2025 @02:36PM (#45673321)

True, everything depends on what one is trying to do at the time.

Re:Scripts I wrote that I use all the time on Wednesday October 22, 2025 @01:54PM (#45672747)

Regarding the `line` script, just a note that sed can print an arbitrary line from a file, no need to invoke a pipeline of cat, head, and tail:

    sed -n 2p file

prints the second line of file. The advantage sed has over this line script is it can also print more than one line, should you need to:

    sed -n 2,4p file

prints lines 2 through 4, inclusive.

Re:A Closer Look at Piezoelectric Cryst on Wednesday October 22, 2025 @01:28PM (#45672365)

*Crystal

Re:Denosing Images of Cats and Dogs with Autoencoders on Wednesday October 22, 2025 @04:41AM (#45666353)

I pointed out this spelling mistake when OP posted it yesterday, and they've done it again. Strange.

Re:AI-generated version of Without Me, Eminem recommended by Spotify on Wednesday October 22, 2025 @04:34AM (#45666307)

It looks like everything this IAX Music channel has posted there is AI slop.

Re:New Scams on Tuesday October 21, 2025 @04:50PM (#45661521)

Please use actual titles when you can, unless they're obviously clickbait.

"Meta ramps up scam detection features in WhatsApp and Messenger"

Re: on Tuesday October 21, 2025 @01:15PM (#45658443)

This is the second obviously-LLM-written physics post this morning and I expect as the chat bots get better, we're going to see so many more of them. Maybe people are seeing the likes of Terrence Howard getting attention for nonsensical "science" "theories" and wanting to get in on the grift... I don't know.

Re:Denosing Images of Cats and Dogs with Autoencoders on Tuesday October 21, 2025 @11:03AM (#45656718)

*Denoising

Re:I Have a Question on Tuesday October 21, 2025 @10:51AM (#45656558)

Prefix your post title with 'Ask HN:' and add a short query. Flesh out the question with more detail if necessary, in the post text.

Re:SpaceX launches 10,000th Starlink satellite, with no sign of slowin... on Tuesday October 21, 2025 @08:07AM (#45654822)

"Starlink satellites in low-Earth orbit are de-orbiting and falling to Earth at the rate of one to two per day."

Re:Unified Gravito-Electromagnetic Resonance (UGER) Model on Tuesday October 21, 2025 @07:58AM (#45654762)

Your previous post claimed to explain "The Nature and Purpose of Dark Matter" while this post is "... eliminating the need for separate exotic forces like Dark Matter and Dark Energy."

Are you a physicist? It seems like every few weeks, another "independent researcher" posts another LLM-generated Grand Unified Theory, or Millennium Prize solution, or paradigm-shifting scientific discovery, without having the relevant math or physics background to spot when their LLM of choice is hallucinating, and why it doesn't work.

Re:Show HN: Unstable Singularity Detector update 1.4.1 on Monday October 20, 2025 @12:25PM (#45645725)

Your link is 404.

Re:Show HN: Copilot AI Agent on Monday October 20, 2025 @12:23PM (#45645707)

Wow, what an original name! Microsoft's lawyers definitely won't be contacting you...

Re:Marc Benioff: I no longer believe National Guard is needed for SF on Saturday October 18, 2025 @12:38AM (#45624941)

A Fox News talking head literally said this recently, but had to apologize for saying the quiet part out loud...

https://www.cbsnews.com/news/fox-news-brian-kilmeade-apologi...

Re:Ask HN: Why are group chats with AI not a thing yet? on Saturday October 18, 2025 @12:29AM (#45624897)

My friends and I had a very simple version of this up and running perhaps fifteen years ago, where we fed-and-scraped Cleverbot[0] and had a second bot posting its replies in our chat. Obviously it was drastically inferior to state-of-the-art LLMs, but it was still amusing. It would only 'speak' when addressed by name, so we weren't flooding Cleverbot with input. Of course, it only ever had one line of context at a time.

[0] https://www.cleverbot.com/