I have a behavior pattern from my father: I'm initially 'happy' if someone is getting hurt (not Bad).
Like someone is dropping something or whatever.
I realized that's not the reaction I want to have and have a core memory of a woman asking a man who drove into the wall (with the back of a truck) if everything was okay with him before even checking the damage to the truck and her house wall.
I want my first reaction to be like that.
My older cousin told me on e that our grandpa was also like my father.
It's an unlogical trained behavior. Nothing more.
And unfortunately that also means there is no reason for people being like that than 'they learned it.
Germans have the word schadenfreude. Is it even a cultural pattern and my family just pushed the goal post further than others?
I'm happy to explain again what makes me so tremendously excited:
It's the first time ever I talk to a system and it does something useful.
If it's able to understand in German and English that I want x and y it's the first time we will have something normal people can use.
It's the perfect working interface.
You know I loved the idea of a chatbot but it was way to stupid. But I only wanted it to do basic things.chatgpt already can do more than the basic system I need.
I'm totally fine of writing good and extensible documentation with good examples and stuff (what I already do anyway) and if chatgpt now can do 90% of basic support queries it will already change a lot.
I'm NOT necessary exited because it can answer random questions in a way that it feels like an agi.
I don't like the unscientific ad for his gf company.
'which helped launch the movement of those opposed to endocrine disruptors, was retracted and its author found to have committed scientific misconduct'
I asked chatgpt just an hour ago to tell me how to put a doc at the first position of the navigation with docusaurus (Facebook static page generator).
It told me the sidebar.js way.
I asked it to tell me another option which I can configure in the doc file directly.
Chatgpt got all of that right. If this is just a text prediction engine, it's crazy impressive how well it just works.
I don't even start designing the prompt to get what I need.
I definitely will follow all of this very closely and already have plenty of use cases for this 'text prediction engine '.
I'm hyped because I'm waiting for ml to break through for a while. Google for example has a huge us hospital under it's cloud platform with tons and tons of data.
We will see more and more expert systems.
Chatgpt will push for more investments.
This is not just a hype it's part of the journey we are for a few years now.
And in comparison to shit like crypto or nft there are already clear benefits.
Stabel Diffusion is fun to use and to explore.
Chatgpt is fun to use and to explore.
Not that this is any indication but the interest in SD was huge but the interest in chatgpt is much bigger. We had multiple internal talks about it and due to the interest in it, more got scheduled.
You will no longer search through documentation. You will just ask the product documentation what you need.
All of those real life use cases pushes more and more money into this topic