Maybe the future of the programming is the human-language-level programming.
Maybe it will take a lot of iterations. For example (limiting myself to standard CSS button options only):
USER: Make a button
AGI: Background color of the button?
Attachment:
button colors.jpg [ 19.42 KiB | Viewed 8441 times ]
(USER clicks on his choice)
AGI: Button size?
Attachment:
button sizes.jpg [ 20.51 KiB | Viewed 8441 times ]
(USER clicks on his choice)
AGI: Button padding?
Attachment:
button padding.jpg [ 24.81 KiB | Viewed 8441 times ]
(USER clicks on his choice)
AGI: Radius of button edge?
(USER clicks on his choice)
AGI: Button border color?
(USER clicks on his choice)
AGI: Does the button color change when you move the mouse?
USER: No.
AGI: Does the button have a permanent shadow, or does the shadow appear only when the mouse is over the button?
(USER clicks on his choice)
AGI: Can the button be disabled? Under what conditions?
(USER clicks on his choice and somehow describes the conditions affecting this property)
AGI: Do you want to specify a non-default width for this button?
(USER clicks on his choice, or enters a specific value)
AGI: Is this button part of a group of horizontal or vertical buttons, with or without borders?
(USER clicks on his choice, or enters a specific value)
What are the opinions of the programmers about the future of programming?