If you're a non-technical founder in AI trying to hire engineers, what you need are solid, old-school software engineers with a great background in software architecture, design, and communication skills. There is no mythical "AI expert" engineer 😅 #ai #claudecode #softwareengineer #swe

I asked Claude to move my Node.js and SuperBase project into Python and MongoDB and it basically created this hot mess where it used some APIs from Node.js, some from Python. It had some tables in SuperBase, some in MongoDB, some overlapping tables even and basically an unusable project. I talked a lot more about this in my last video and a whole bunch of comments appeared around it. I want to start addressing some of these. In this video, I want to talk about this one comment, which basically said, "When do you decide it's time to hire an expert AI engineer?" So that stuff like this doesn't happen again. A word of advice, when you talk about this mythical AI expert engineer, just know that you sound like you don't know what you're talking about. People may even sell themselves as this AI expert engineer, but that's because there are other people that exist that don't know what it is. Essentially, people who are exceptionally good at using LLMs for coding are just exceptional software engineers with great software architecture and system design background and excellent communicators. So if you know data structures and algorithms really well and you can write and review LLDs or low-level design documents and translate product requirements documents and high-level design documents into LLDs, which is the implementation level software architecture stuff. And if you can understand the product requirements and make sure that you're guiding the LLD in that direction, and you have good communication skills that you can be understood by the LLMs, very clearly and crisply, you're going to do great. So engineers who are really good at using AI coding tools and getting to the right outcomes are none other than good old-school software engineers with solid fundamentals and good communication skills. Unfortunately though, the median software engineer is not good at some of these things. So for mid to senior level engineering hires at places like Google, there is a mandatory system design interview. It is one of the hardest interviews for software engineers to pass. And that's the interview that requires no coding, zero coding skills required. People may be excellent at lead code and they'll fail the system design interview just because they're not system-level thinkers. So yeah, if you're a non-technical AI founder and you're looking to hire engineers for your company, don't look for this mythical AI expert engineer. What you need are solid software engineers. Did the emphasis being on system-level thinking, architecture, communication skills, etc.