Did you try running it on some standard heuristic programming competitions? (i.e. atcoder's AHC or topcoder's marathons)
Would be nice to see it benchmarked vs something where the human effort can be easily quantified
Oh hey, a new steam fest. Let's see if I have missed any interesting puzzle games lately...
Oh man, I almost missed it. His Jibaro & Witness are my favorite animations from LD&R.
Microsoft: Xbox, Xbox 360, Xbox One, Xbox Series X, Xbox Series S
OpenAI: Hold my beer!
Oops, the correct link is: sudokupad.app/qq3kyjvu5q?s... (hides the grid)
Also, check out the official post on Logic Masters: logic-masters.de/Raetselporta... that includes the FAQ
Someone combined sudoku with metroidvania and the result is the best puzzle game of the 2025.
sudokupad.app/qq3kyjvu5q
I don't even like sudoku, but this is absolutely amazing.
Note: it's really difficult and quite long, but well worth it if you like logic puzzles.
Link to (ugly) code: github.com/FakePsyho/cp...
This result is surprising to me, as it took ~4h + I didn't think it would be possible to get close to 1st without doing any of the fancy rail placement (either optimizing at the end, splitting existing rails for a new station etc.) and/or complex BS eval.
* BS is very slow, I'm grouping states by turn and eval is expected money at turn 800 + small bonus for partially fulfilled trips.
I have also added DP at the end (same as nika), during the contest I had greedy instead of DP and incorrectly assumed that it doesn't add too much
AHC043 happened while I was on vacations and I finally had time to revisit my solution.
During the contest I had greedy+(basic, but heavily tuned) SA, where state=order of stations, with very simple rail placement.
Turns out changing greedy -> BS (+DP) would be enough for 1st*
All this proves is what we all knew anyway: that competitive programming is just glorified copy-paste /s
For y'all that don't do programming contests
2700 is top200 in the world
1900 is top5000
Majority of people can get to 1900 within one year when dedicated. Many can get there in a few months.
Most people won't get to 2700 after a few years of training.
This is 3 months apart.
It looks that I have qualified for the AtCoder World Tour Final!
I haven't been to any onsite in a really long time (since they have disappeared post-covid), so I'm looking forward to it.
Not that I disagree with anything that you've mentioned, but all I had in mind is that AI training thing is a silly take.
I'm happy that people are moving here from Twitter, but doing so over "I don't want AI training on my data" just shows your ignorance.
No matter if you're a republican, democrat or moderate, it is in your best interest to get rid off EC. The only people that are happy with the system are full-time politicians as it gives them a monopoly.
Both parties are screwing you over with their joke version of democracy.
Something I will never understand about US politics:
Why is there no nationwide movement to abandon Electoral College and shift to a ranked voting system?
EC forces a 2-player game that only adds polarization, since shitting on your rival has the same value as promoting yourself.
Thanks for the hints, I think I clicked "show me less..." thing few hundred times already, but so far no results :( Maybe it updates user preferences in batches and takes a few days to kick in.
UFO50
Progress: 33/50
I've never expected that this will become my pick for GOTY, but it's not even close.
I mostly post about interesting/unusual games, immersive entertainment and competitive programming. From time to time, I also vent my frustration about the terrible state of AI safety.
Please like, comment or subscribe 😅
PS. What do I do to remove all of that anime from my discovery tab
Hi Bluesky, I'm Psyho!
Few things about me:
- I've spent 20 years of my life in various fields of AI (including OpenAI before it was loved/hated)
- I believe, I'm the only person that made a living out of competitive programming
- My heart is in game design, though I'm yet to publish anything