Hacker News: kevinwang

New comment by kevinwang in "Claude Code Down"

kevinwang — Mon, 06 Apr 2026 15:39:51 +0000

Still down for me. (And still nothing on the status page!)

New comment by kevinwang in "Pentagon formally labels Anthropic supply-chain risk"

kevinwang — Thu, 05 Mar 2026 21:35:13 +0000

Even in your analogy, it's appropriate to reject the terms of marriage and not wed this person. But it's unprecedented to also vindictively ruin their life (e.g. by unilaterally putting them in jail)

New comment by kevinwang in "California is free of drought for the first time in 25 years"

kevinwang — Wed, 21 Jan 2026 02:45:47 +0000

From the link alone, it looks like the state actively funds cloud seeding research, not active practical cloud seeding?

New comment by kevinwang in "Salesforce regrets firing 4000 experienced staff and replacing them with AI"

kevinwang — Thu, 25 Dec 2025 16:59:37 +0000

Thanks. Link should be changed to this.

Edit: oh wait, this article isn't the source either. It references an article by "The Information", which I assume is https://www.theinformation.com/articles/salesforce-executive... There's also this follow-up: https://www.theinformation.com/articles/story-salesforces-de...

It's paywalled, so I can't verify.

New comment by kevinwang in "XY Problem"

kevinwang — Thu, 04 Dec 2025 15:10:45 +0000

The YXY problem?

New comment by kevinwang in "Cuddle Fish – A Soft Floating Robot for Safe Physical Interaction"

kevinwang — Mon, 01 Dec 2025 15:09:12 +0000

It says:

We see potential uses in companionship and affective interaction. The robot’s ability to elicit spontaneous touching and positive emotional responses suggests it could serve in therapeutic settings or as a social presence for people who spend time alone. The quiet operation and gentle movement make it suitable for environments where noise and sudden motions would be disruptive.

New comment by kevinwang in "'Calvin and Hobbes' at 40"

kevinwang — Thu, 20 Nov 2025 16:46:50 +0000

Well said. I feel similarly despite being born about 15 years later.

New comment by kevinwang in "Lee Felsenstein"

kevinwang — Mon, 10 Nov 2025 13:49:13 +0000

Good idea! Anyone should be able to add it if it's in the public domain.

Can Anyone Rescue the Trafficked Girls of L.A.'S Figueroa Street?

kevinwang — Sun, 26 Oct 2025 13:26:34 +0000

Article URL: https://www.nytimes.com/2025/10/26/magazine/sex-trafficking-girls-la-figueroa.html

Comments URL: https://news.ycombinator.com/item?id=45711682

Points: 4

# Comments: 1

New comment by kevinwang in "Advent of Code 2025: Number of puzzles reduce from 25 to 12 for the first time"

kevinwang — Sun, 26 Oct 2025 12:42:25 +0000

A little sad for me because I've enjoyed the global leaderboard aspect for years but of course my second reaction has to be to take a step back and appreciate all the joy that this one man has given us for all these years.

And he's made it clear from that start that he never intended the global leaderboard to be the point, plus AI the last few years messes it all up. All good things come to an end, and I gotta appreciate the good run that we had, and the voluntary work of one person that gifted it to us.

Luigi Mangione's Missing Months

kevinwang — Tue, 21 Oct 2025 16:52:21 +0000

Article URL: https://www.nytimes.com/2025/10/21/us/luigi-mangione-asia-trip.html

Comments URL: https://news.ycombinator.com/item?id=45658089

Points: 3

# Comments: 2

New comment by kevinwang in "Diffusion Beats Autoregressive in Data-Constrained Settings"

kevinwang — Mon, 22 Sep 2025 22:16:50 +0000

Not sure if you meant this because it doesn't cite the paper you mention, but it's a similar work: "An Investigation of Model-Free Planning", Guez et Al. (Deepmind) 2019 https://arxiv.org/abs/1901.03559

New comment by kevinwang in "DeepMind and OpenAI win gold at ICPC"

kevinwang — Wed, 17 Sep 2025 23:56:54 +0000

For problems that human teams eventually get correct, they seem to have submitted mostly 1 time -- occasionally 2 or 3. For problems that they did not get correct, there are some problems with up to 16 submissions.

New comment by kevinwang in "Conway's Game of Life, but musical"

kevinwang — Thu, 11 Sep 2025 15:37:23 +0000

Darn, I can't hear the music on iphone safari :(

New comment by kevinwang in "The elegance of movement in Silksong"

kevinwang — Tue, 09 Sep 2025 13:20:31 +0000

Thanks, that was a fun read!

New comment by kevinwang in "The elegance of movement in Silksong"

kevinwang — Mon, 08 Sep 2025 22:34:46 +0000

I remember reading an essay (probably from here?) about how a great way to build a game is to build it around a "toy" -- something that is pleasurable to simply interact with, even without objectives. I can't find it anymore -- the closest I can find is https://medium.com/@keerthiko/toys-to-games-25d35b40425d but I don't think it was that, although it's based on the book "The Art Of Game Design" which may have been a common inspiration.

Anyways, I've often thought about Super Smash Bros. (particularly, Melee) as a prime example of that idea.

New comment by kevinwang in "Pebble Time 2 Design Reveal [video]"

kevinwang — Wed, 13 Aug 2025 18:14:31 +0000

Oooo, that looks a lot better than the initial designs!

New comment by kevinwang in "How to teach your kids to play poker: Start with one card"

kevinwang — Tue, 12 Aug 2025 01:52:10 +0000

A quick take:

> reasonable possibilities are too many to analyze, one is forced to pretty much guess.

You could say the same thing about chess, but an experienced player wouldn't, because they know which candidate moves are reasonable and which lines to delve into through intuition shaped by experience.

Similarly, you might say the same about poker. The possible hands your opponent has are actually quite large, but an experienced player can have a reasonable idea of the possible hands and their probabilities, which may involve eg ignoring most hands as unrealistic and bucketing hands into classes.

New comment by kevinwang in "Prediction markets create harmful outcomes: a case study"

kevinwang — Wed, 06 Aug 2025 14:34:22 +0000

This seems to have just happened: https://polymarket.com/event/another-dildo-thrown-at-wnba-ga...

(and it could happen again: https://polymarket.com/event/another-dildo-thrown-at-wnba-ga...

https://polymarket.com/event/dildo-thrown-at-wnba-game-on-au... )

New comment by kevinwang in "Hierarchical Reasoning Model"

kevinwang — Tue, 05 Aug 2025 22:15:55 +0000

Am I the only one that doesn't get this at all? Why use a recurrent architecture at all if you're only going to backprop through 1 step?