Doruk
banner
dorukyilmaz.net
Doruk
@dorukyilmaz.net
From being woefully optimistic to being hopefully optimistic.
March 1, 2025 at 10:17 PM
Reposted by Doruk
This is the Pixel World, a small 64x96 2D simulation populated with AI creatures making their own decisions to survive.

They explore, gather resources, build, talk with each other.

There's no scripted behavior model. Just a prompt with a list of actions, short memory, current state, world info.
November 28, 2024 at 5:51 PM
At somepoint, idea of a curated feed of sound people with one's interests felt like the ideal internet usage to me.
Overtime many things changed, how people interact w. social medias (became m. personal), the perceived anonymity of the internet and so on.
I am not interested in a such place anymore.
November 26, 2024 at 1:26 PM
hate-based development
October 21, 2024 at 11:36 PM
Reposted by Doruk
Took this photo of a cheeky goat today
April 26, 2024 at 8:50 PM
I wonder if there is any generic/abstract advice that applies to most cases of any kind.
Anyone got a clue?
March 17, 2024 at 8:20 PM
Someone actually made an Operating System game
store.steampowered.com/app/2075070/...
Looks interesting, horror themed.
I will deffinetly checkout when I get a chance.
February 13, 2024 at 7:31 PM
Alphaville
February 9, 2024 at 8:09 PM
Reposted by Doruk
US Survey Foot xkcd.com/2888
January 31, 2024 at 4:08 PM
Reposted by Doruk
Spelling xkcd.com/2885
January 25, 2024 at 1:21 AM
As with everything good in life,
the gmail basic html version will be killed very soon.
For those who don't know, it was a very fast to load version of gmail, with very little sensory overload.
Here is an image from interwebs as I don't want to leak my gmail.
January 22, 2024 at 6:20 PM
Immutable Wind
January 17, 2024 at 6:03 PM
January 17, 2024 at 6:01 PM
Reposted by Doruk
I like how, at the time, the Netflix Marvel shows weren't really seen as very good so weren't really considered canon, but now the rest of the MCU isn't really seen as very good either they slot RIGHT IN
January 10, 2024 at 9:28 AM
I wonder if the main issue ML projects have stems from their approach, that trying to quantify things as right and wrong.
Maybe there is no pattern to match.

Just like how training/memorizing is not the way to learn something.
January 14, 2024 at 8:22 PM