njyx

Assuming you know when they hallucinate?

Also, Github CLI budgets exploding :-)

In theory though, there is now a new way for community to support open source, but running vulnerability scans in white-hat mode, reporting and patching. That way they burn tokens for a project they love. Even if they couldn't actually contribute code before.

There should be a way to donate your unused tokens on every cycle to open source like rounding up at the chekout!

A little crazy that Google seems to be the lead investor though...

Super convenient for switching out models, and they also play a vetting function. Cool to see a useful service raising!

(I'm the author.)

Ha ha - just found this. Thanks for posting. Happy to answer questions if people have them!

Agreed that you need a rewarding core loop and also one that becomes richer over time (more characters? more actions, more power, less power...). But, I think you also need a longer term engagement element beyond the core loop: being engaged with the goals of game. If there's only one (e.g. only one ending) then if you love the game you'll try to reach it at least one. If there are more you might be tempted to try them all.

Another thing that helps and isn't discussed much in the article is the "meta" game around a game. Especially if the game has liveops and new characters come out, there are events etc. Most by themselves probably won't make someone replay. But if there's enough new stuff and a community which is constantly discovering new stuff to do then that can spark the desire to play again.

Great stuff: how often is the underlying data refreshed?

So sad to see. May the cellular automata replicate in his honor for ever...