4 ms·

> You ask it to show you a color picker, it generates HTML code for a color picker, you copy that into your browser and you can pick your color, which you can t

> You ask it to show you a color picker, it generates HTML code for a color picker, you copy that into your browser and you can pick your color, which you can then copy&paste back into the LLM for further processing. This is slower, more awkward, and less efficient than just picking a colour from an existing colour picker.

grumbel 3y ago
The point is that nobody had to program this. Nobody had think up front "Will the user need a color picker?". Nobody had to find a spot in the UI to place it. You can just will it into existing as a user with nothing but the power of the LLM. No classic app has anywhere near that amount of expressiveness. Future versions of chatbots will of course have support for <iframe> or similar to display this kind of stuff inline, that should be obvious.
- Planktonne 3y ago
  I don't think that's the point; it's not that no one had to program the colour picker - it's that no one did. The workaround shows that there was a need for it. Having to copy and paste code to get a colour picker that you can then use and then paste the output back into the chatbox is less efficient than using a colour picker. LLMs can work as general interfaces, but the trade-off is that they're less efficient than a specific one.
- gtirloni 3y ago
  > The point is that nobody had to program this. Nobody had think up front "Will the user need a color picker?" Are you sure the training dataset didn't have a few articles explaining how to code a color picker? Did it figure it out by itself like you say?
- esafak 3y ago
  One duty of the programmer and product manager is to think about the likely uses of the program and to build a UI to enable it. If users wanted a blank slate they could write the program themselves, or have chatGPT write it. Maximum expressiveness is not the goal, because it comes with a price. There is a balance to be struck between expressiveness, and economy of effort and cognition.
TT2048 3y ago
For now, until running ToolFormer or one of the other Jarvis like models get* better

grumbel 3y ago

The point is that nobody had to program this. Nobody had think up front "Will the user need a color picker?". Nobody had to find a spot in the UI to place it. You can just will it into existing as a user with nothing but the power of the LLM. No classic app has anywhere near that amount of expressiveness. Future versions of chatbots will of course have support for <iframe> or similar to display this kind of stuff inline, that should be obvious.

Planktonne 3y ago
I don't think that's the point; it's not that no one had to program the colour picker - it's that no one did. The workaround shows that there was a need for it. Having to copy and paste code to get a colour picker that you can then use and then paste the output back into the chatbox is less efficient than using a colour picker. LLMs can work as general interfaces, but the trade-off is that they're less efficient than a specific one.
gtirloni 3y ago
> The point is that nobody had to program this. Nobody had think up front "Will the user need a color picker?" Are you sure the training dataset didn't have a few articles explaining how to code a color picker? Did it figure it out by itself like you say?
esafak 3y ago
One duty of the programmer and product manager is to think about the likely uses of the program and to build a UI to enable it. If users wanted a blank slate they could write the program themselves, or have chatGPT write it. Maximum expressiveness is not the goal, because it comes with a price. There is a balance to be struck between expressiveness, and economy of effort and cognition.

TT2048 3y ago

For now, until running ToolFormer or one of the other Jarvis like models get* better