Bill Thompson

What if you're paying for Google Drive?

I don't know about their terms of service now.

That is one of the best things that I use with large language models is any product I download, I have the neural network examine the terms of service.

And then you can pretty much understand, like, here's my focus.

Here's the 40-page terms of services document.

When you click that link that you got, what are they able to do with my data?

So that's how I sign up for apps.

And that's one of the great uses of a large language model, in my opinion,

is to quickly understand how these things are being used.

And that's why I say with Apple, with Meta, with all of these large information, you are more the product than the product's the product.

And that is because they're trying to build the most powerful, capable artificial intelligences, which I think is a misnomer.

And again, we can get into it later.

But

They're trying to build these hyper-competent artificial intelligences.

And you need two things for that, really, is training data and you need compute.

And that's why you start seeing them coming out with like Meta's building its own nuclear engineering facility or nuclear facility or something like that.

And they need more training data.

So if I want to build a replica of Joe Rogan that I can make hyper-realistic AI videos for, I need every picture of your face from every angle.

I need every wince, every squint, everything you've ever done so I can introduce more training data to better train that neural network in order to generate more hyper-realistic versions of yourself.

And so when a company is offering you something for free and it's fine, like if people are fine with that idea, then by all means, download all the free apps that you want.