« BackAsk HN: Is fast.ai's "Deep Learning for Coders" still relevant in 2025?undefinedSubmitted by hedgehog0 3 days ago
  • distracted_boy 8 hours ago

    What if your goal is to transition into the role of an ”AI Engineer”, would the course be valuable for this?

    • thiago_fm a day ago

      Yes it's still very good if you want to toy around with it. I did it and in the end I feel like I could do simple models and understand how it works.

      If you want a job, it may not be what you need.

      If you want to work with ML you likely need the PhD + get lucky.

      Remember that the skillset for AI jobs != ML.

      If you are interested on it because of AI, it's better to focus on LLMs and more NLP related fields.

      • hedgehog0 21 hours ago

        Thank you for your suggestions!

        Originally I was more interested in ML/DL theory and mech interp, so you can see I was more into theory. Recently, I am also curious and leaning towards learning more about how to build products with foundational models, such as LLM; for instance, recently I got the "LLM's Engineer Handbook [1]".

        > If you want to work with ML you likely need the PhD + get lucky.

        I do am interested in doing PhD in theoretical computer science, but I'm not too sure about AI PhD as I heard many (not-so-good) things about it.

        > If you are interested on it because of AI, it's better to focus on LLMs and more NLP related fields.

        Do you have any recomenndations that you have gone through that you think are helpful?

        [1]: https://www.oreilly.com/library/view/llm-engineers-handbook/...