World Building Aspirations and Keyboard Shortcuts Introduction

There were three big model releases today. The first two are interesting, but the third one is crazy. The first, Anthropics Cloud Opus 4.1 was released. This is basically just an across-the-board update on Opus 4 model on agentic tasks, real-world coding, and reasoning. It's good, and it shows Anthropics' commitment to just continually improving its product. Additionally, OpenAI announced that they're releasing an open-source model. And this is a pretty big play, since Meta has signaled that they're out of the game, and Chinese open-source models are getting released and really shaking up the market. Interesting that OpenAI is throwing their hat in the ring. It's a reasoning model, and it's called GPT-OSS, and it's available in two different sizes. So one is big, it'd be for data centers, and the others would be small and run locally on someone's desktop. And running this comes with the big benefit of it being more secure, since all the data would be stored locally, and then also not having to pay for the API usage if you're a developer. The smaller model is said to be on par with OpenAI's O3 mini, and then the larger one is said to be similar to capabilities of O4 mini. So pretty good. So if you're a business that's looking for a really secure solution, you don't have the funds to create your own proprietary model, and you're kind of freaked out by these other models that are open-source that have been popping out, and you like that OpenAI brand, well, that could be a solution for you. Finally, the last announcement, and the craziest, is that Google DeepMind announced Genie 3, which is a world-building model that creates these interactive, playable environments from a single prompt. So basically, you can create this simulated world through a prompt. And it's not just like video game-type graphics. It's like VO3-level, hyper-realistic looking. Like you could write that you're walking in a field, and there's a waterfall behind you, and you just navigate and see the waterfall that you created. There isn't public access yet, so all we have to go off is what Google has shared. With a grain of salt. But it does seem like the physics are pretty good, and that there's object permanence, meaning that they showed a shot of this person on a jet ski going through a waterway in a town. And you can see the water from the jet stream coming up, and it kind of flows out and falls like water would. It all seems very realistic. And then it splatters onto the concrete, and you can navigate, look away, and then looks back down, and the water is still on the concrete. That's hard to do, and it's figured it out. It's happening. I loved playing Roller Coaster Tycoon and Sims growing up, and now we're on this trajectory to create and experience those worlds. And this currently isn't something you can do within virtual reality. It's just 2D images that you can navigate within this 3D space. But, I mean, that's where it's going, right? And since this all lives within Google, it's been speculated that Google will probably, at some point, integrate its Google Maps data. And so then we're looking at, you can really create a digital world based off of our physical world. So for businesses and marketers, definitely keep your pulse on this one, because this is the beginning of new immersive media.