Indeed, I get your point... How fast it will get there would depend and how much data we input into the models....
I guess it's more important to feed decentralized AI models as we wouldn't want to give that power to Centralized organizations.
Does it make sense?
This is my view. I do all I can to push information on Hive, which is available to anyone. I do not spend my time on centralized entities. Social media applications are data warehouses these days.
To me, we have to make Leo that.
It would be nice to see more activity on InLeo. But we are already collecting a nice amount of data daily… And it is decentralised!
I agree. We do need to get more activity.
That is why I am focusing upon filling the database with as much as I can. The threadcasts are doing a good job of that bringing news to the platform. It is crucial to get this information decentralized, on a permissionless network.
Seems I'm talking to the right person
Does inputting data have to be done manually by users?
Is there a way we can input the data online (on the internet) as Open AI does in training their model?
Someone could set up something scrape other sites.
LeoAI is built using Llama3 as the model. However, we need to add to our own database to keep it growing. The weighting is going to be different.
For now, the key is to add as much as we can manually about a wide range of topics to get the relationships going within the database.
after building our database manually, wouldn't it be easy for just anyone to use it for training their own model, without our permission since it's Blockchain?
Same way we're training Leo AI on hive data without anyone's permission.
Anyways, most or all centralized platforms don't ask for permission either, they only include opt out options, hoping the masses wouldn't even see that button.
you're absolutely right my friend, I'm not even two weeks old here but when I saw the vision of Inleo I knew it's better to feed it with data than give it to some company that'll take all the profits to themselves