Comment by mattfrommars
5 days ago
I don’t know if this warrants a separate thread here but I have to ask…
How can I realistically get involved the AI development space? I feel left out with what’s going on and living in a bubble where AI is forced into by my employer to make use of it (GitHub Copilot), what is a realistic road map to kinda slowly get into AI development, whatever that means
My background is full stack development in Java and React, albeit development is slow.
I’ve only messed with AI on very application side, created a local chat bot for demo purposes to understand what RAG is about to running models locally. But all of this is very superficial and I feel I’m not in the deep with what AI is about. I get I’m too ‘late’ to be on the side of building the next frontier model and makes no sense, what else can I do?
I know Python, next step is maybe do ‘LLM from scratch”? Or I pick up Google machine learning crash course certificate? Or do recently released Nvidia Certification?
I’m open for suggestions
I'm not entirely clear what your goals are but roughly, just figure out an application that holds your interest and build a model for it from scratch. Probably don't start with an LLM though. Same as for anything else really. If you're interest in computer graphics then decide on a small scale project and go build it from scratch. Etc.
The competition for root and branch AI models and infrastructure is intense and skilled.
But if you're adjacent to some leaf use-case for AI, you're likely already as good as anyone else at productizing it.
And that's who is getting hired: people who show they can deliver product-market fit.
Maybe look into model finetuning/distilation. Unsloth [1] has great guides and provides everything you need to get started on Google Colab for free. [1] https://unsloth.ai/
go thru workshops here https://www.youtube.com/@aiDotEngineer/