The Importance of a Soft Template in Data Science Workflows

I actually agree with you as someone who knows quite a bit about how these are built. I am extremely skeptical of that ever being solved. The models are all the same as what we've been doing forever. It's taking this problem, it's mapping it in some latent space and it's doing some linear stuff in latent space. But these problems aren't solved by making that space better in my belief. So there's an opportunity there for, hey, I'm starting a new project. Can you just give me the template? Like, can you give me this stuff I'm going to need for this problem? And I don't need it to be perfect. Right?