How do you balance using AI at home for experimentation versus at work within company processes and SOPs?
ASKED BY · Unknown · 14 SEP 2026 · ▶ 00:29:18
John explained that software lifecycle management (STLC) and lean manufacturing principles (via agile/Six Sigma) remain essential at enterprise scale to ensure software is trustworthy, auditable, and maintainable beyond short-term point products. Pairing someone with system thinking and AI expertise with deep business knowledge yields outsized productivity gains.
Summarised by AI from what the panel said. Watch the moment for the full answer.
From the transcript.
Machine-transcribed, so it may contain errors. Every timestamp opens the episode at that second.
29:17with a big development team and infrastructure and all this and SOPs and you know like there's all these processes that you have to go through is it is it hard to kind of what are they overlapping at all is it you know is it people that are messing around as much as you are that are trying to do that for work related tasks yeah i think that um
29:39everything that we do um at work as as a core process of what is software development like STLC still needs to live and operate to have successful software life cycles. And if you're going to generate AI slop and just create some point product and abandon it in three weeks, okay, you don't need a STLC. But if you're going to create software that you want to live and extend past, you know,
30:09a small cycle and other people use it and trust it, then you're going to have to establish these principles that have been long, way before we had AI, even back when we had machine code, right? These people were doing very similar processes than what we do today. If you look at how software is created today, guess what? It's just lean manufacturing. The Japanese, you know, they learned this a long time ago,
30:32and it's the same principles. I mean, now we call it agile. But go read, you know, your Six Sigma's books and your lean manufacturing. It's all the same. interesting so like i came out of manufacturing and now you know heavy into a software and finance company um but i was like oh i i know all these principles i know this is all it all related to
30:52me right away because it's the same how do you produce and uh support a good that you're putting to market it's you know that's all it is that's yeah it's about making something that's repeatable that people can trust, that you can use, that can withstand an audit. All those things are important if you are someone who wants to create software. If you want to have fun, do whatever you want.
31:18Don't use GitHub. Don't use any source control. Just put stuff out there. But if you want to actually create something that you're proud of and you can stand behind, then learn how to create software. Learn what a software release cycle is. Learn what DevOps is. Make sure you're doing testing. you know like the reason why the ai is now doing this for you is because they're like okay people
31:41don't trust this so i'm just going to now add a step to my process that's going to validate my output and it's going to cost more tokens but if the humans aren't willing to do that and then they're going to make fun of me for this then i'm going to just go ahead and address this yeah courtney we were talking about this before because you had made software before for your own
32:03use, you know, in your own business use. And it was like, I'm not going to release this because I didn't have to service it and, you know, do customer service. I have all these, I mean, I've released commercial software before and almost every tech support question I'd get, I would, you know, go, okay. And I would put a link to the operator manual, which goes out with the software and here's
32:27page 25. Here's step-by-step instructions on how to do what you're asking for. it's in the operator's manual i just send them a link to that but yeah and you know uh i find the first thing that corporations are doing is handing off their tech support to ai and more and more like the behringers you know they got that uly bot now to handle your tech
32:51support you can't necessarily speak to a live person you got to deal with the chat bot when you give it a problem and they're relying on that chatbot to provide answers to all the stupid questions at least and um you know the question is how do you get past the chatbot to elevate something to real programmers that are can you know elevate a software bug or something to
33:15those that can actually fix it right right yeah it is an interesting thing with uh with customer of support now because i've seen i've seen some of it can be really good if it if it knows the product really well then you can ask it particular questions and it'll give you really good answers that it can also be completely wrong um very easily um and send you on a wild goose chase