Updating examples to new AI APIs and LLM versions

[bmerkle]

the AI train runs at a high speed and meanwhile LLMs and APIs have changed. Some of the examples are already a bit outdated and we should decided if we update them

Observation e.g. from chapter 06 python examples:

• LLM gpt-3.5-turbo

  • this LLM is outdated and we should update to some gpt-5.x version
  • the performance and costs are even better for newer LLMs that for the old ones
  • some environment (e.g. my university OpenAI proxy) does not even offer gpt-3.5 anymore

• OpenAPI calls

  • some samples uses deprecated APIs, e.g. the client.chat.completions is considered legacy by OpenAI
  • https://platform.openai.com/docs/api-reference/completions
  • we should update to openai.ChatCompletion.create

• Approach

  • we should decide if we update
  • I can help with python and the js code examples
  • I can also review the docs and update them
  • I do not know if the videos have references to specific model names/version or api calls and this would be probably more work

please advise or lets discuss 😃

[github-actions[bot]]

👋 Thanks for contributing @bmerkle! We will review the issue and get back to you soon.

[bmerkle]

maybe we could use GitHub (Next) Agentic Workflow https://github.com/githubnext/gh-aw and let the agents do the work for us 😺

[koreyspace]

@bmerkle yes, willing to see what the agentic workflow can bring us. I have in mind to do a major refresh in the new year!