Jake Casey

Claude seems to be fixing a super annoying developer problem. Anthropic announced a research preview feature called Auto Mode for Claude Code, expected to roll out by March 12, 2026. The idea is simple: let Claude automatically handle permission prompts during coding so developers don’t have to constantly approve every action. Sstops those annoying permission prompts during long coding sessions. Before this, you had to use `--dangerously-skip-permissions` to work without interruptions. That method worked fine but took away all your safety nets. This new auto mode gives us a smarter option. Claude will take care of the specific permission choices on its own while still blocking threats like prompt injections. You can finally let long tasks run without watching your screen the whole time. Since it is still a research preview, you should run it inside isolated setups like sandboxes or containers for safety. Expect a small jump in token usage and delay, because the model needs extra time to process the security checks. Once available, you just type `claude --enable-auto-mode` to start. If you manage a team and need people to manually approve actions, you can restrict this feature using Mobile Device Management tools like Jamf and Intune or through configuration files.

Modern Brazil is full of liminal spaces like these and I love them "Liminal spaces often lack people, feature unnatural lighting, lack clear exits, and evoke a sense of uncanny familiarity or nostalgia. Common examples include hallways, abandoned malls, hotel corridors, and gas stations at night. Liminal spaces often trigger feelings of anxiety, discomfort, and strangeness because they violate expectations of how a place should be used or experienced. They can also offer profound moments of deep growth, change, and, for some, a sense of quiet calm. In anthropology, liminality is the middle stage of a ritual or transition, where a person is between roles (e.g., in the, limbo between childhood and adulthood). This creates a temporary, ambiguous state of existence."

In 2016, Yousuke Yukimatsu was diagnosed with brain cancer, that same day he quit his job as a construction worker to become a DJ

Yesterday was GREAT But today I will be honest it's getting harder The game is now 3000 lines long in a file called fly.html and Cursor starts getting into issues To make one change, it has to go over the entire file and applying a change can take 15 seconds Then over half the time it breaks the game, with no errors (so it does code properly), or it removes entire parts (like the plane itself, or the controls etc) I think what I'm reaching is the limit of the LLM context window, so it does the change I want but then it forgets other parts of the code due to context limit Cursor I think works by "diffing" the source and the answer of the AI/LLM, so it then starts removing code that you actually need that it'd forgot One solution is to split it up into multiple files, which I tried, so I asked it to do that but then it creates lots of files like [game.js] function updatePlayer(){ // insert your updatePlayer code here } But of course Cursor SHOULD add the code there, so then I had to go over every file to ask it to put back the code there, but in the end it just gave me a broken game that didn't work anymore I love to give everyone an HONEST idea of what's possible with AI coding now, and obviously it'll keep getting better, so I think you can definitely generate entire apps and games easily, but you start hitting roadblocks after awhile I do know Cursor team and other people are working daily to fix those things though A context limit, IF that is the real problem, is a relatively easy thing to fix though Any way my tip is: make sure Cursor immediately starts coding the app in multiple files, so it just works from the start! I'll keep hacking and trying to make this work though, I got pretty far yesterday which was very fun and cool

So it's a technical test to prove Photo AI can achieve the same abilities as a real photographer with free movement within an AI photo shoot For that you need to be able to zoom out and zoom in essentially forever Next is moving left, right, up and down through a space Total free movement to simulate a photographer is the goal!

Me in da club with AI audio Sounds kinda like some bar in Colombia with live music?