Tim Davies explains why companies need clean, structured data to get real value from AI. Large models like Gemini 1.5 train on 35 million books, but your competitive edge comes from maintaining your own data well. Most organisations skip this and lose out.
But if you don't get your data cleaned and put out there, then nobody's going to get anything from you. I've just lived in this world at the time where I was training, and then you contributed nothing to it.
And it's. It's the two sides, isn't it? By writing that book.
Yeah.
That's been trained. It's been part of a training data set for models. I mean, Gemini 1.5. Read 35 million books.
Yes.
And knows them and can quote a line from any of them.
Yes.
And how many people do you know that can accurately quote lines from multiple books?
Oh, my goodness.
It's insane.
It's a win.
Yeah. But the flip side to that is as an organization, the CEO, he's got within his organization a vast amount of data, which actually is probably the only moat, the only thing that is truly defensible within the business is the data and insights that they've got. But they don't have those insights because they haven't done any work on their data to make sure it's appropriately structured and mined.
EP 60: Why Founders Become the Bottleneck — And How AI Changes Everything | Thea Brook
