Hey, I think I am probably overreaching it but here is what I think the entire crypto-web3-AI dream will take us tomorrow. I wrote the article - Vision Electronic Cloud in 2012, it has been 12 years since I wrote that and we are nowhere near the Vision Electronic Cloud today. It seems that solving a problem is more important as compared to giving people vision. There are homes today which are like Electronic Cloud but there have not been massive transformations in energy, and the repair and maintenance cost of IOT devices ultimately acted as a negative proponent of Electronic Cloud.
I don't think anyone is sitting on the realization here of how intelligent LLMs and GPT actually are. They can act as wonderful brains. However, there is still the problem of how to use them by providing the right context? How would the privacy of a user work if they are not given the right context? What if a third-party organization requests access to some of the results computed on such data? Who will own their personal data?
Here is what the vision encapsulates - everyone will have their personal LLM which will be trained on their own data. This data will be stored in vectors and can be anything digital under the sun. - Images, Videos, Audios, Texts. There will be a schema along with each item which will tell how this data was captured and other statistics about the information. Each person will have cameras embedded into their glasses which will upload everything that they see to the cloud, and it will upload everything they hear to the cloud, this data composed with artificial intelligence will be revolutionary. Everything they read, hear or write will be recorded.
The personal LLM and
How would privacy work?
This could potentially lead to privacy violations on an unimaginable scale. No entity should have entire control over this user's data, this could lead to changing privacy policy or being under government pressure to vilify someone's data. The request to access this data would be made using a smart contract transaction,