Yes and no. I think it's obvious there's a ton of value in the product and it's a massive force multiplyer for certain types of tasks. But it inherently cannot

Yes and no. I think it's obvious there's a ton of value in the product and it's a massive force multiplyer for certain types of tasks. But it inherently cannot be trusted and still requires someone with expertise to verify and implement. I don't think they're going to achieve real AGI. I don't think we ever will. I think they'll get something "close enough" and claim they have it, but I don't think the path to AGI is through LLMs.