"This is for customer service" is so blind to the actual stakes here that its actually sort of amazing. I get not understanding this stuff its wild next frontier world-changing stuff. But at least he can have the humility to not make himself look like a jackass.
The equivalent here is someone in like 1994 being like - "Internet website shopping? I don't get it - I can just go to Caldor or JC Penney and get whatever I need. And why would I pay for shipping? And can't every company just build their own website? What's the difference? I just don't see it."
Microsoft isn't interested in ChatGPT because its going to like automate customer service for Office 360. They are interested in tuning it to take down Google's search dominance. So, that single application of this technology is "POSSIBLY DETHRONING THE CORE PRODUCT FOR THE 4th biggest company by market cap in the entire world."
Now
@nelsonmuntz you might say - Microsoft is crazy!! Surely google, who has a wee bit of AI expertise themselves, agrees with Nelson that this is all commodotized $29.99 chatbots, right?
They are shzztting bricks and reorganizing their entire AI/Product organization to counter this threat mid year -
A New Chat Bot Is a ‘Code Red’ for Google’s Search Business
View attachment 83014
The actual stakes here are are "owning every piece of written or spoken language interaction that you have in your entire life" - across all your commerce and services, your learning and information needs, your personal life and daily schedule, etc. Every piece of technology you use, every product or device that you have need to interact with. Oh, and the moment you deploy it, it takes its billions of interactions of datasets, and now starts creating a highly optimized version of that model and learning you specifically and how you personally talk, the inflections and vocabulary you use, etc etc.
People gravitate toward the best UX for whatever they are looking to do. Its why Apps murdered mobile browsing, and why Google is now modern day all encompassing information center.
The companies that adopt and deploy this technology - which is for 99% of all businesses wayyyyyy outside the scope of their resources to hire teams to build and integrate this stuff, so they need to tap partners/startups/businesses that specialize in it to do it for them - get shockingly good ROI compared to almost any other intervention, and also get their teams experience in conceptually understanding AI/ML at a basic level.