Improve AI error messages

[eamodio]

Description

via @justinrobots

Results of investigation into AI failures

• GK AI most common errors:
  • Error: you have exceeded your token limit
• Copilot most common errors:
  • Error: Unable to Generate Commit Message: (Copilot) Request Failed: 400 {"error":{"message":"Model is not supported for this request.","code":"model_not_supported","param":"model","type":"invalid_request_error"}}
• OpenAI most common errors:
  • Error: Unable to Generate Commit Message: (OpenAI) (OpenAI) 429: Too many requests (rate limit exceeded) or your account is out of funds

[eamodio]

For GK AI:

• We already show a Your request could not be completed because you've reached the weekly Al usage limit for your current plan. Upgrade to unlock more Al-powered actions. notification with an Increase Limit button

For Copilot:

• We now show a The selected model is not supported for this request. Please select a different model and try again. notification with a Switch Model button

For OpenAI (compatible):

• We now show a Rate limit exceeded, or your account is out of funds. Please wait a few moments, check your account balance, or switch to a different model, and then try again. with a Switch Model button

I've also improved some of the other messages and added better handling for other cases as well, see:
https://github.com/gitkraken/vscode-gitlens/blob/5cd7b9ce358dc1f192df491f8db4c5e1d57d6fec/src/plus/ai/aiProviderService.ts#L931-L1029