Upgrade DALL-E image generation using the OpenAI API

[Kardbord]

The DALL-E API is now available in public beta. Its image generation capabilities are significantly improved over the bot's existing dalle-flow implementation. Due to these increased capabilities, it is a paid service, but I can set usage caps so my end-users don't break my bank. :) Understandably, it also comes with a host of usage policies to follow, so there will probably be some research involved to make sure I'm compliant. My game plan is roughly as follows:

• [x] Implement Go bindings for the OpenAI image generation API. This will be a separate project in a separate repo.
  • [x] Ensure the library complies with the Usage Policies.
    • [x] Build in calls to the moderation endpoint?
  • [x] Link to Usage Policies in the library's README for library users.
• [x] Use the library from the previous step to implement a new image generation command for the bot
  • [x] Decide whether or not to keep the existing dalle-flow implementation around as an unpaid option
  • [x] Disable the new OpenAI image generation command if no API key is present. This key should be optional.
• [ ] Add the improvements suggested in #73 to this implementation as well.

[Kardbord]

OpenAI bindings for Go are being implemented at https://github.com/TannerKvarfordt/gopenai