Taro Logo
3

AI blurring lines between FE and BE

Profile picture
Senior Software Engineer at Dda month ago

Hey folks, I am an iOS engineer with little more than 7 years of experience.

I have been thinking of expanding my skillset either to Android (did some 10 years ago) or Backend but haven’t had the time to invest in learning again. (toddler keeps me busy)

With AI being able to do what it could, I am curious how people are thinking about pivoting their careers to add more value and hopefully not be obsolete?

Thanks

37
4

Discussion

(4 comments)
  • 2
    Profile picture
    Tech Lead @ Robinhood, Meta, Course Hero
    a month ago

    I actually don't think which domain you choose will be very relevant in the age of AI as opposed to the depth you have in whatever domain you ended up with. With AI, anybody can technically become an iOS engineer, Android engineer, back-end engineer, etc, but they can't automatically become a good iOS/Android/back-end/etc engineer with only AI at their side.

    In terms of technical depth, there are 3 core areas that come to mind:

    1. Code quality - Take this course to go deeper: Level Up Your Code Quality As A Software Engineer
    2. Debugging - Take this course to go deeper: Become A Debugging Master And Fix Issues Faster
    3. System design - Take this course to go deeper: System Design Masterclass: Shipping Real Features To Production

    More thoughts here: "What technical skills should a senior SWE focus on to stay competitive over the next 5 years?"

    On top of technical skills, the more fundamental "soft" skills become far more important to as I talk about here: "How to AI-proof your career?"

  • 2
    Profile picture
    Software Engineer
    a month ago

    Hm. Instead of what to code, I think how we code will probably evolve.

    I've somewhat gone hybrid AI-assisted coding. Basically I start out a session by manually writing code and then have the auto-complete adapt to my style and suggest improvements. Occasionally I'll open up the chat window and ask for a change, but if the AI can't do it after a few tries I'll just code it myself.

    Technical depth is still super important in this case. Pure vibe coding won't even get you past intern-level code complexity.

    Easy wins:

    • boilerplate code
    • fast variable rename/import changes with auto-completes
    • highly contained logic like "write this 5 line map-filter function for me"
    • if you know exactly what you want but AI can find it for you (ie. some poorly documented pytest hook)

    And if it's regex or CSS, try to get the AI to handle it 😂

  • 2
    Profile picture
    Employee @ Robinhood
    a month ago

    I just write whatever code is needed to move my projects forward. For example, I started taking on iOS since my team's low on iOS bandwidth. I get like 70% of the way there with most commits by myself, and just ask in the public ios engineering channel for help when I'm stuck.

  • 1
    Profile picture
    Senior Software Engineer [OP]
    Dd
    a month ago

    Great insights everyone