Is Python really the first step in learning AI?

Data Productivity Cloud
,
1

Watching a youtube video last night, the speaker suggested that the first step towards learning AI is to set up your local development environment to be able to learn Python. Traditionally, this has been my biggest stumbling block to learning how to code. There always seems to be some config file, registry setting or system path that gets skipped and makes the entire effort too painful to continue.

Can the high code editor in DPC be my “local development environment”? Can i skip all the messy setup and just run python inside of DPC?

Or am I conflating the application of Python to learning AI with the purpose of Python applied in the DPC for building and supplementing pipelines?

2 Likes
2

Pushdown Python within the local dev environment will indeed handle a lot of the setup for you.
You’ll still need import packages within the code as with any python script, but package installation, system paths and the like should be much less of a headache.

Primarily its designed to supplement pipelines but you can do a lot with it - for example building traditional predictive models or working with LLMs.

For learning to use generative AI of course we do have a lot of native components to allow you to work with vector databases and LLMs directly, without any python code needed at all.

3 Likes
3

Thank you @JulianWiffen ! That feels like a pragmatic answer.

Since I posted, I had a conversations with an “AI expert” and the conclusion we’ve begun to formulate is that AI has evolved at such a rate and in a way that concerning yourself, as an analyst, on how the “evolved modules” work (eg., LLMs / foundational models, etc) is probably not a good use of time. Better to concentrate on “how to leverage” these as components than worrying about the wiring.

Appreciate the thoughts! I think this confirms for me that the first step in learning AI is leveraging Matillion to handle all the wiring. I can concern myself with how to use the native components and chat-oriented programming for edge cases that could benefit from custom coding in python.

2 Likes