00:00 Speaker A
Open AI kills Sora. This is my diary. They exchange ideas and have this story. The sudden collapse of Open AI’s most hyped product since Chat GBT is said to be the video generation tool Sora. What do you think of this story? How did this go from massive hype to closure?
00:27 Dan
I think there are two things. Oh, they’re refocusing. They obviously knew they had to do that. In other words, we’re going to cut down on the redundant stuff. Well, I also think you can’t ignore what Google has done with its own video generation capabilities, right? It’s, it’s, and its photo manipulation tools. I think they came out with Nano Banana and people were just like, “What is Sora?” You know, that’s it, the main idea here is definitely Open AI reduction. And we’re trying to make sure they compete and get that money back, right? And that’s the big question everyone still has, how are we going to pay for all this? What does that mean for Oracle or what does it mean for Microsoft? You guys have to pay for what you spend. Well, I guess that’s where it all goes. Me too, I mean, the idea of AI being used in art makes some sense to me. You’d use something like Photoshop, but then you’d say in a big way, “Okay, make a new Disney movie with this.” It’s just like, “Okay, well, you don’t want to see that, do you?” It’s like, there’s no artistry involved, just telling me that. Hmm, III seems to be on the side of not being a fan of full AI development videos. But it’s just me and I’m just one guy.
01:45 Speaker A
Boban seems to agree with Dan. This was just Sam Altman potentially preparing his company for a public debut. Improves concentration.
01:54 Boban
Yeah, it increases concentration, but I think Dan said something great earlier. It’s also a monetization issue, but Sora didn’t have a clear path to monetization. And I think the other thing was GPU usage, memory usage. You might want to assign it to your enterprise, right? I think that’s what made the decision, considering the bigger piece of the pie.
02:14 Speaker A
Is that like a broader story for both of you, Dan, that the real monetization is ultimately the agents and the companies and maybe not the fancy consumer tools?
02:24 Dan
that’s right. That’s what most people believe is the case at this point. Sora and the whole idea along those lines seems to be, let’s see what sticks. Throw everything against the wall and see what sticks. And it seems like the one everyone is most excited about right now is the agents. It’s true that 2025 is supposed to be the year of the agent, and while it started out as a way to get up to speed, I think it’s really starting to take off now. Open-claw folks, apparently it’s impossible to get a Mac Mini in San Francisco. Well, people went out and bought it so they could run their open claws on their Mac Minis. So I think that shows that the interest is definitely there. And while the GPU is still very important, I think we also need to consider the fact that the agent side of things is also beneficial to the CPU when it comes to AI. You know, the GPU may be running the AI, but the process the AI is running is based on the CPU. Therefore, those people will continue to enjoy its benefits. You see, ARM announced its own CPU. Oh, and Nvidia has a new CPU rack. Obviously, AMD and Intel are already there. So it’s one of the things that’s really interesting to see this kind of evolution of AI. I think the idea with Sora was, “Look, how cool is this?” You can make a banana flag or something. I don’t know.
