Felt inspired and started a new terminal expecting to get stuck pretty fast... ...but @grok just kept chewing through harder and harder systems, tasks and features. I now have a fully functional game. I have a server. I have this world that has been building in my head for years. The old advice is: "Do not build an MMO." I even used to give this advice to new game devs. Now it just feels so within grasp somehow. Is it time to start a new adventure?

pricing ai by real world value sounds fine until someone has to decide what a bug fix in an indie game nobody's played yet is worth

LFG at 2am and the only tank online is a paladin with no shield

Kling 4 ships with 10 keyframes. give it a year and ai video reinvents the timeline editor

taking the longest flight path on purpose so i can go make a sandwich

a year of everyone building subagent org charts and anthropic just says put it back in one loop

raid leader saying its just trash right before the third wipe

Id rather see one greybox someone actually shipped with Kimi K3.1 than another rocket scene demo

beyond good and evil 2 is the final boss of scope creep

Corpse run back to the instance is still the hardest content in the game

Epic's AI jam promo hand having six fingers is the most on brand thing they could ship

Every pug has one warrior rolling need on cloth for an alt that doesnt exist

next fest needs a filter for the ai disclosure box. and one for roguelike deckbuilders while theyre at it

devs pulling public playtests so nobody clones the build before launch is such a weird 2026 problem

Meta basically trained Muse Glimmer in The Sims but every sim has an inbox

Argon getting docked points for saying idk is the first benchmark loss that sounds like a feature

deepseek flash 1017 is open weights now. anyone tried it for npc dialogue yet or still too janky?

konami is right that ai wont find the fun. it just gets you to the first playtest faster

bro

This thread ended up being a Roadster commercial lol