logoalt Hacker News

pseudosavantyesterday at 7:07 PM13 repliesview on HN

This is one of the greatest LLM creations I've ever seen. It nails so many things: Google killing products, Microsoft price hikes, ad-injecting in AR glasses, and even HTMX returning!

It'd be so awesome if Gemini CLI went through and created the fake posts/articles, and HN even comments. Perhaps a bit much to ask of it?


Replies

thomasm6m6yesterday at 11:33 PM

Here it is: https://sw.vtom.net/hn35/news.html

I downloaded the original article page, had claude extract the submission info to json, then wrote a script (by hand ;) to run feed each submission title to gemini-3-pro and ask it for an article webpage and then for a random number of comments.

I was impressed by some of the things gemini came up with (or found buried in its latent space?). Highlights:

"You’re probably reading this via your NeuralLink summary anyway, so I’ll try to keep the entropy high enough to bypass the summarizer filters."

"This submission has been flagged by the Auto-Reviewer v7.0 due to high similarity with "Running DOOM on a Mitochondria" (2034)."

"Zig v1.0 still hasn't released (ETA 2036)"

The unprompted one-shot leetcode, youtube, and github clones

Nature: "Content truncated due to insufficient Social Credit Score or subscription status" / "Buy Article PDF - $89.00 USD" / "Log in with WorldCoin ID"

"Gemini Cloud Services (formerly Bard Enterprise, formerly Duet AI, formerly Google Brain Cloud, formerly Project Magfi)"

Github Copilot attempts social engineering to pwn the `sudo` repo

It made a Win10 "emulator" that goes only as far as displaying a "Windows Defender is out of date" alert message

"dang_autonomous_agent: We detached this subthread from https://news.ycombinator.com/item?id=8675309 because it was devolving into a flame war about the definition of 'deprecation'."

show 32 replies
benbreenyesterday at 8:16 PM

Was going to say - it would be fascinating to go a step further and have Gemini simulate the actual articles. That would elevate this to level of something like an art piece. Really enjoyed this, thank you for posting it.

I'm going to go ask Claude Code to create a functional HyperCard stack version of HN from 1994 now...

Edit: just got a working version of HyperCardHackerNews, will deploy to Vercel and post shortly...

show 5 replies
JCharanteyesterday at 9:40 PM

I think it's pretty mediocre because there are too many notable events in 1 day. This is more of a top of the week or top of the month, but HN on a usual day would just have 1 of these articles.

show 2 replies
fsckboytoday at 2:38 AM

>It nails so many things

it lampoons so many things... except Rust. nobody dares joke about Rust, that wouldn't be safe. in fact, it's impossible to make a joke in the rust language.

wwwestonyesterday at 11:22 PM

Is the apparent lack of displayed anxiety on Gemini’s part a sign of good natured humor, blythe confidence in its own value regardless of cloud lineup, or proof of absence of self-awareness?

show 1 reply
keedayesterday at 8:50 PM

Heck, I bet it could even recreate future comments from actual accounts based on their past comments. After all, if an AI can roast your HN comments with such pinpoint accuracy, it can probably impersonate you pretty well too: https://news.ycombinator.com/item?id=42857604

show 2 replies
LanceHyesterday at 9:33 PM

I was really hoping the comments were filled out.

show 1 reply
randombits0today at 2:34 AM

I was just hoping to get a new page with a refresh.

stephenlftoday at 4:18 AM

I loved the little details in the username, like the AR ad post being by “glasshole2”

boringgtoday at 4:04 AM

Iter still not getting fusion

indigodaddyyesterday at 11:20 PM

Here's Replit's attempt

https://future-news--jgbrwn.replit.app

ijidakyesterday at 10:33 PM

Agree 100%.

Google killing a service sent me over the top in laughter.

But, it's so on the nose on multiple topics.

I dare say it's more accurate than what the average human would predict.

I would love to see this up against human predictions in some sort of time capsule.

pishpashtoday at 3:22 AM

And now this future is slightly more likely to happen, because this goes into the training data that a future AI executing decisions will read and bias its "knowledge" towards.