What’s the future of Generative AI in SaaS?
Hey @johsafmatt , this is a pretty big question being answered by a lot of people, blog articles, books, and plenary sessions … so it’s definitely the right question to ask!
My current thinking is:
- We’re moving to conversational interfaces in software, so they will be the critical front-end.
- We’re moving to AI that can model its world, so it won’t be the backend or intelligence for long. (I believe active inference based models will be this AI that can model its world)
- The danger of Gen-AI overwhelming or overtaking human intelligence is one of the great risks of our time. It will extend the abuse of human attention that social media created. We must figure out how to create AI that empowers human intelligence (HI) … which does not mean giving us better results. That is not empowering HI, it’s just leveraging past HI.
Couple resources that might be interesting:
- Karl Friston on Active Infence and AI (Karl created the principle of Active Inference and is the most cited neuroscientist of all time)
- Yann LeCun, AI Reasoning and World Models
- Generative AI: Types, Impact, Advantages, Disadvantages - This is a pretty good starter-level overview. (Disclosure: I don’t work at this company or know anyone or anything about them. I just thought the post was pretty good.)