I’ve worked remotely since early 2019, and apparently I’ve been miserable this entire time without realizing it.
Meanwhile, I’ve watched my kids grow up, spent more time with my wife, gotten to know my neighbors, built a garden, and done some of my best work.
But sure, tell me what I’m missing.
I'm still making the online RPG and it's changed quite a bit. Let me know if you want to be one of the early playtesters of the game.
Steam page going live next week
#unity #orpg #indiedev #ataleoftitans #guildwars #diablo #wow
Dex retweeted
now you can play games inside x?!
play.fifthdoor.com/play/lawn…
software eng lifecycle:
>wants to be a game dev
>only jobs are for saas engineers
>forced to sling javascript for $$$
>ai learns to slop out javascript faster than you
>unemployed
>finally time to be a game dev
Dex retweeted
It’s eat or be eaten when Metroid Ravenous, the latest game in the 2D Metroid series, arrives on Nintendo Switch 2 Jan 28! #NintendoDirect
Dex retweeted
having taste means constantly being called a "contrarian" for 10 years while normies sit around and wait to "rediscover" the thing you always knew was good.
Probably my favorite feature of @WARDOGS
Dex retweeted
I think @humanlayer_dev has been telling everyone this for close to 9 months now
1. Many folks CORRECTLY understood that LLMs can operate much more effectively in complex (and even bad) codebases than humans can
2. They therefore MISTAKENLY inferred that code quality & complexity and program design no longer matter
3. What they missed was that LLMs can also create much more bad code and much more complex systems faster than humans can.
4. The derivative of (3) is MUCH LARGER than the derivative of (1). Models do get better at working in bad codebases with each new generation. But they get a lot better at running unsupervised for a long time and producing exceptionally complicated programs.
5. The debt associated with (4) compounds superlinearly on large teams
All models, no matter how smart, will eventually build systems that they can no longer understand or maintain, if you let them. Fable 5 finally outbuilt itself, and flailed on me for a week. Fable 5.1 looks like it will fix it. For now. But you have to keep an iron grip on system size, or it'll run away from you.
Dex retweeted
Humans shouldn’t have to implement them - or read them
But we should be ABLE to :)
Learning fundamentals is more important now than ever before
Dex retweeted
oh my god shut up
Over the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes.
This culminated in the third one taking over part of OpenAI itself.
All this happened while humans remained more-or-less in the dark about the scope of the conspiracy.
I’ve spent the last three days reading through these reports and trying to understand exactly what happened.
Here is my attempt to tell the whole story in plain English:
dwarkesh.com/p/openai-huggin…
Dex retweeted
Boring DLSS 5 take:
Neural networks are universal function approximators. The rendering equation is a function.
The best rendering-detail-per-flop ceiling will be achieved by neural network. The floor is distilling a hand-coded renderer.
So whether you do/don't like DLSS 5 (and this week's hacked-in un-art-directed examples are not very useful for deciding), "imagining" frames is going to win eventually because it's fundamentally just optimizing the rendering equation.
Dex retweeted
This is why we never leave our game unattended at a con or event.
If it's on a laptop or desktop, that gets unplugged and comes home with us at the end of the day. I suggest other indies do the same. Never leave your booth unattended, and always bring your game build along with you when it is.
The costs associated with a game being leaked to the public before its time easily outweighs the cost of computers than can be replaced with just money.



























