Funny. I had claude recreate Thrust based on wiki article alone - and it nailed it. I then spent a week trying to implement a perfect autopilot for it. And then a solveable level generator. I have this running as a screen saver in one of my apps.
Your reconstruction is so close, I'm a little surprised you didn't stick with the original controls! Even though the BBC had a rather weird layout for all the non-alphanumeric keys.<p>It's been a very long time since I played Thrust, and I've played plenty of WASD games in the meantime, but it still felt really strange to control it that way rather than the classic BBC Z/X/*/? for left/right/up/down.<p>Pretty sure it was return to fire, space for shield/tractor; and maybe * for thrust? <i>Edit to add:</i> ah, no, shift to thrust. Works very well on modern keyboards too.
Maybe I should pop in an original keyset as an option - I think the original uses the caps key maybe. I'm pretty sure it's weird on a modern keyboard layout. I think the version on the bbcmicro site keymaps it by default, in any case if you want to revisit the original its here:<p><a href="https://bbcmicro.co.uk/game.php?id=432" rel="nofollow">https://bbcmicro.co.uk/game.php?id=432</a><p>Edit: I'm going to have to go and revisit it again now... I was sure it was infuriating on a modern keyboard.
Someone should update the Lunar Lander page on Wikipedia -- it doesn't mention Thrust.<p><a href="https://en.wikipedia.org/wiki/Lunar_Lander_(video_game_genre)" rel="nofollow">https://en.wikipedia.org/wiki/Lunar_Lander_(video_game_genre...</a>
I hate when people say "AI" without specifying which model. It's like saying "plane" without specifying if you're talking about a Cessna or an F-22 Raptor.<p>> I asked Claude<p>There is no such thing as 'Claude'. Claude is a brand of model, some incredibly dumb and some incredibly capable. Was this 4.6 Sonnet? 5 Fable? Without specifying, the post is essentially meaningfless.
> This is where things got interesting. Not because AI wrote the code — the code itself isn’t complicated, it’s a 1986 game that ran in 32K of RAM — but because Claude turned out to be an extraordinary tool for interrogating 6502 assembly.<p>Complaining about slop with slop.
> The tick loop waits at least 3 <i>centiseconds</i> per frame, giving an effective rate of about 33.33 Hz<p>3 centiseconds instead of 30 miliseconds, totally not a robot
I was about to complain at you for jumping to conclusions, but your cited example contains two emdashes and a nested "it's not X it's Y". It certainly looks like slop. In my own writing I'm increasingly conscious of trying to avoid the appearance of slop, I would like to think I would have caught this.
A lot of people don't hate slop, they hate <i>other people's</i> slop. Of course their slop doesn't stink, they prompted it better or something.
I tried to use his writeup to oneshot the game and you can see my results here along with the prompts I used:<p><a href="https://kemendo.com/thrust-one-shot.html" rel="nofollow">https://kemendo.com/thrust-one-shot.html</a><p>Notably while the game "works" it's not even close to an a "reproduction" as far as I can tell - moreso an interpretation.<p>This has one level that doesn't level increment, none of the adversarial sprites are correct and the color and iconography are incorrect.<p>Granted I didn't give it much to work with but I figured I'd see what happens. As far as one shots go, I've seen worse.<p>I used commodity GPT 5.6 HIGH on firefox via chat interface
Hmm - the article isn't dated and it doesn't mention which models were used for the initial slop version. The initial commit in the git repo is from Feb 15, 2026.<p>I wonder how the initial pass would fare now with Fable 5 or 5.6 Sol?
I think it was Opus but can't remember the version. Would have been whatever was available in Claude Code in February.<p>Managed to find some screenshots. It did work but gravity wasn't really gravity - things fell at a constant rate. Turrets were floating in mid-air. It was missing the shields and you couldn't pick up the ball from the pedestal.<p><a href="https://www.jamesdrandall.com/old-thrust/1.png" rel="nofollow">https://www.jamesdrandall.com/old-thrust/1.png</a><p><a href="https://www.jamesdrandall.com/old-thrust/2.png" rel="nofollow">https://www.jamesdrandall.com/old-thrust/2.png</a><p><a href="https://www.jamesdrandall.com/old-thrust/3.png" rel="nofollow">https://www.jamesdrandall.com/old-thrust/3.png</a>
I've been saying for months that any time you say something about the capabilities of the AI tooling you should say what exact model, agent, and effort level you're using. Because I've taken a few of those "AI can't do X" and had AI do a fine job at them.<p>I have Sol 5.6 ultra working on Thrust right now.
I'd be interested in seeing the "slop" version for comparison.