#459 – DeepSeek, China, OpenAI, NVIDIA, xAI, TSMC, Stargate, and AI Megaclusters

3325.788 - 3341.193 Dylan Patel

So you have the stressful period where it's like flat or might start going up. And you're like, what do I do? Whereas there are also lost spikes that are, it looks good. And then there's one spiky data point. And what you can do is you just skip those. You see that there's a spike. You're like, okay, I can ignore this data. Don't update the model and do the next one and it'll recover quickly.