The trend seems to be going towards "AI Copilot for everyone".
In the years to come, no matter what your field is, if you are a knowledge worker, you will probably (have to) work with an AI copilot (assistant).
Large language models used for generative AI, like GPT-3, are impressive already.
Rumors already suggest that GPT-4, planned for release in the next couple of months, will make another leap in how AI generates text & content.
A similar leap than GPT-3 (mind blowing) was to GPT-2 (just OK).
(think about what GPT-10 will do! 🤯)
And domain-specific applications show amazing value already.
For example I'm using Github Copilot when coding.
It's a big time saver and I'm amazed at how it can guess the next line(s) of code - mixing both best practices, and my own code uniqueness.
With domain-specific language models, you can imagine AI copilots for everything - leveraging the logic & lingo of your specific field.
For example, if you need to write a legal document, an AI copilot will help write it in "legalese" (using the legal lingo) and surfacing relevant legal references.
What will happen to society & the economy, when every knowledge worker's productivity & output quality increases massively thanks to AI?