It´s really so much fun to play around with the whole AI stuff 😀

Now I thought why not just run my own ChatGPT.

Doing it would even have some advantages.
You probably wouldn´t have to deal with the “ChatGPT is at capacity now” error anymore and you can customize the settings.

In the OpenAI Playground you have settings for the language model, temperature, maximum length and more

The model for example can impact the quality of the response and with the temperature you can make the output more random.
Then you have also setting for maximum length/tokens to keep the costs under control.

To run an own ChatGPT like script you need to register on OpenAI to get an API token.

Then I just went to ChatGPT and gave this instruction:

twinaxe:

Create following website:

– Multiline text field for input
– Enter button to submit input
– Multiline text field for output
– API call to OpenAI AI with model=”text-davinci-003″
– Use input as question and write response to output field

This is the response:

ChatGPT:

As an AI language model, I can provide you with the steps to create a website with the requirements you’ve mentioned. However, I can ...

This thread has 4 more replies.
WANT TO SEE THE REST?

Join STM Forum