The Maverick Debugger

(Est. Reading Time: 12 min) Look around. How many tech devices do you see? If you are nerdy like me, it might be a slightly higher than average number of devices. Smart TVs, the phone or pc you are reading this on, smart assistant speakers, gaming consoles, all of those are obvious ones. What about your car? Your lightbulbs? Oven? Smoke detectors? Check and check. Tech is everywhere, and more and more things are getting "smartified" each year. Who builds all that? Hardware people, software people, and a whole slew of supporting cast members. So tell me, when more and more of these tech things are going up, and more and more apps are getting created, and all these billion dollar investment deals keep happening to build the next breakout startup around AI, who is going to do all this? Well, if you said AI and a bunch of "vibe coding" non-developer types, you are very wrong. Those people smell success, but can't see the future problems. They don't understand t...

(Est. Reading Time: 2 min)      So, here I am, tinkering with ChatGPT and other LLMs for over a year now, learning to leverage it for success for software engineering. It needs to be called out. LLMs suck at code. If a chat gets too long, they start to introduce artifacts and oddities into the code. For example, multiple unconditional return statements in a single function, adding a new one every time it "helps". It does have its good moments, though, as it helps discover bugs, sometimes. Ask it how to do something you've never done, and it will give you a basic understanding to work on. But it does not have the same logical mentality to actively participate in writing code. It gets worse. OpenAi researchers found that, beyond face value simple code, LLMs (even the unreleased frontier models) suck at code. Once the code starts to grow beyond a few classes, or a little complexity between different functions, it cannot hold the context in a way to provide meaningful output...