Almost all of the Crypto community has moved to AI including a16z. Crypto was a solution looking for a problem. The parts of AI will be beneficial and will improve productivity. However, we do not know which companies other than the few we already know will ever make a profit and open-source models are a big threat to them too. However, the training cost may limit the open source movement unless we can train with far less data and resources. If someone is just creating a wrapper around ChatGpt API, they do not have much moat. What would be the moat in the future data, model, or ability to automate the whole interaction for example I want to travel to England and I provide the parameters the whole thing is fully automated and I validate the whole thing once it is ready and approve, the APP will do everything from creating the complete itinerary to booking to keeping track of the trip in case something changes. Even in the last case, it is not something you can monopolize easily. I think the data is the key but most of it is behind the firewalls and copyright laws so we will see new kinds of jobs where people will be creating training data by coding or writing books for AI models (some companies are already hiring poets) or providing their expertise in a format that it can be used as training data.
If there is money to be made, the hype will be created like we have right now and was happening around Crypto and most people will lose money except a lucky few if it does not work out, people will move on to the next new thing for which they can create hype. Rinse and Repeat!
Almost all of the Crypto community has moved to AI including a16z. Crypto was a solution looking for a problem. The parts of AI will be beneficial and will improve productivity. However, we do not know which companies other than the few we already know will ever make a profit and open-source models are a big threat to them too. However, the training cost may limit the open source movement unless we can train with far less data and resources. If someone is just creating a wrapper around ChatGpt API, they do not have much moat. What would be the moat in the future data, model, or ability to automate the whole interaction for example I want to travel to England and I provide the parameters the whole thing is fully automated and I validate the whole thing once it is ready and approve, the APP will do everything from creating the complete itinerary to booking to keeping track of the trip in case something changes. Even in the last case, it is not something you can monopolize easily. I think the data is the key but most of it is behind the firewalls and copyright laws so we will see new kinds of jobs where people will be creating training data by coding or writing books for AI models (some companies are already hiring poets) or providing their expertise in a format that it can be used as training data.
If there is money to be made, the hype will be created like we have right now and was happening around Crypto and most people will lose money except a lucky few if it does not work out, people will move on to the next new thing for which they can create hype. Rinse and Repeat!