@SoftpaleGames Thank you. We're building Dust of Battle. Fast, tense, tactical battles on the 200AD Silk Road.
https://t.co/ovmndUSGQx
If you've ever wanted to see a Roman centurion face off against a Han Imperial Knight we have just the battles for you!
I know it���s not entirely equivalent but a lot of bad novels were produced when the word processor enabled a massive influx or writers and even more so when Kindle gave them easy distribution. Books didn’t fall apart.
@lucas_builds I’m multi-modal I use coding agents on some projects, on others I use LLM only as a research partner while I write the code. Where it matters I write better code.
did you know your samsung tv has a service called rmdemon (that's not a typo, demon not daemon) that starts a PACKET LOGGER on your network.
what possible legitimate use case could there be here, samsung?
I've been reviewing a lot of code written by Opus 5.5, and it still seems to make a lot of bad software engineering decisions. The kind that will accumulate/calcify into unmaintainable software over time. IMO even the agents will struggle to maintain vibecoded software.
Ever since I first read about the Enigma machines and the work to decode them, I've been totally fascinated by them.
Getting to see them up close at Bletchley Park, I've always wanted to take apart and understand exactly how they worked.
So now I can...
https://t.co/1S4F55cHve
If I could train an LLM purely on my own code (I realise that in practice that won’t work but stay with me) would it be okay for me to use it? Would it still require me to disclose LLM use?
The came for the weavers and I didn’t say anything because it was 180 years before I was born. They came for the miners and I didn’t say anything because I was 13 and too busy playing on my Acorn Electron. Then they came for us and I have no idea what to do and nobody cares.
@dunsincodes I think about the example of Steve Jobs obsessing over the wiring inside the Macintosh. “But nobody will see it” they said to him. His response “I will know it’s there. Try again.” I guess it comes down to how much you care. In 10 years I might not, today I still do.
I came back to Twitter to help promote the game @SpiralChris and I are building and to avoid the politics that drove me off-platform last time. It’s different politics now but… here we are.
Marketing Mallet Mayhem '84 has been tough.
I can't tell how much of the low reach is the algorithm and how much is my approach.
I'm trying to find the people who'd enjoy a bit of old-school arcade air hockey. #gamedev
@azamsharp My experience too. It gets the simple calls right but is quickly out of its depth. It has a super narrow lens and misses the big picture. Understandable when you know how they work. Obvious if you have the experience yourself.
@alanb It’s interesting that even at this point the training still seems to suggests to it that the person asking the robot is then going to do the entire work by hand.
@EuphorIQGameDev@Giulio6686 While I agree with your fundamental point the danger for me is that slop clones create noise and an expectation that make it harder for people to discover quality. We can’t put LLMs back in the box so how do we help the quality rise to the top?
@Giulio6686 Your list seems good. Additionally“What is a prompt?” becomes a relevant question. The results from a newb typing“Copy game X” are, at least right now, going to be very different to an experienced team working from a creative design document and iterating hard on play-testing.