Watch on YouTube
You Can't Win at AI Without Your Own Data
This video does not allow embedding, so you get a static preview and a link to the original.
Key takeaways
- Why is it so important to build a data-centric business in the age of AI?
- And so, if you want to have a unique mode in some way to a business and not be obviously like the frontier models are all trained on all of the free available information.
- And so, in order to have any kind of edge on those models, you need to have proprietary information.
- You have to have data sets that they don't have access to or that they don't have like call it clones or similar data sets.
Chapters
Full transcript
Locally archived YouTube transcript with timecodes. Speakers are assigned heuristically and may be corrected editorially.
0:00Why is it so important to build a data-centric business in the age of AI? >> Well, you can't have AI without a data layer first, because like AI is trained off of data. And so, if you want to have a unique mode in some way to a business and not be obviously like the frontier models are all trained on all of the free available information. And so, in order to have any kind of edge on those models, you need to have proprietary information. You have to have data sets that they don't have access to or that they don't have like call it clones or similar data sets. Cuz if you have a unique data set, but it creates the same outcomes as a different data set that is not the same as yours, it actually does not have the same edge.
0:32And so, you need to have unique data. Um that's if you wanted to have some sort of AI, you know, call it product as as a thing that you would sell. If you want just to use AI within the business, which I think is 95 or 99% of businesses out there, um it's more in order to have Let me zoom out. If I wanted to have an employee come in, right, and do a job, I would have to train the employee on that job. In order to train the employee, I would have to document steps to doing the right behaviors. The documentation of those steps and the the when done right, it looks like this. That is all data. And so, the more documentation you have around the processes you have in the business, which is what are the inputs and what are the outputs and what is the basically the manufacturing process step by step that takes inputs and turns them into outputs, and how do we know output is good versus bad, which fundamentally is just reinforcement training, which works for humans and works for machines.
1:24Um we need to have that so that we can get the most out of the tools. That is all. Like it's just so that you can get more out of the AI versus having a generic out-of-the-box AI, we have a this is how Jack likes to run the AI. If you like this video and you're business owner who wants to break through your current revenue ceiling, I distilled every lesson from scaling 10 businesses past 10 million and three businesses past 100 million into a completely free scaling roadmap that I've used to go from zero to one, zero to 10, and zero to 100 plus. And so, you can click here and you can check it out. Again, absolutely free. And since you're business owner, I appreciate you and enjoy.