For my collaboration with OpenAI, I created some new Rocketganj images using ChatGPT Images 2.0

what level are you at?

smokes and ladders // artefacts from rocketganj

made this in claude btw, some fun new workflows emerging

some days you gotta let the loop reset

never leave earth without the essentials

Still thinking about Digger. Such a great film. Extremely original too. Doesn't neatly slot into any genre, which almost never happens anymore and I think is part of the reason why some of the reviews are what they are. I think maybe we've gotten used to films telling us exactly how to perceive them, in the marketing and in the films themselves. Digger has refused to do both. It's an audacious choice and probably very financially risky, but I'm glad that they made it. Will see it again soon. You should see it too. IMO Iñárritu has made a future classic.

doom machaale

made this image earlier this year. possibly my most seen work by now. keeps showing up on meme pages

hell yeah

once upon a time

Rainquotes

My very late contribution to the Diet Coke discourse: ThumsUp XForce is just so much better

when you fall in love with your ai agent

I think this is unlikely to happen in Hollywood. Billion dollar grossers need a certain critical cultural mass pre-release. Usually that means a well-known IP/franchise/sequel. The IPs with that kind of critical mass are currently all with major studios who for various reasons will not move to 100% AI productions for those franchises by 2030, and there no AI-era IPs yet with the kind of cultural cachet needed to spawn a billion dollar movie. Could happen in China though, but I think a film with that ambition will likely not be “100% AI”, and that label may be less useful over time as more of this tech gets absorbed into current workflows. Which is probably healthy since I don’t think the AI/non-AI binary is conducive to groundbreaking collaborative creative work. Even if one works with the assumption that all shots are generated - that’s still only one part of making a film. Someone will still need to write it and direct it, and it will still need editing and post-production and all other relevant parts of the filmmaking process. Plus a film with that ambition would want to also bring on top tier actors - so performance capture and voice dubbing, etc. - in principle this is not very different from how films are made today, i.e, a synthesis of various tools and methods, and imo is a natural evolution of tech-forward filmmaking. What I think is a higher-impact unlock with AI video is what’s already starting to happen - solo creators and small teams using these tools to create outside the current system and in that process develop new aesthetics, distribution strategies and business models. Exploring this has been one of the focal points of my life and work for the last few years and I’m going to be spending a lot more time building towards this thesis this year. In time, I think some of these AI-native projects popping up all over the world may find the scale and the cultural impact to support a billion dollar grosser movie of their own.