The hardest part of building software is not coding, it’s requirements
The hardest part of building software is not coding, it’s requirements
Why replacing programmers with AI won’t be so easy.
You're viewing a single thread.
Someone tell my boss this, they don’t understand agile. They think we can “start the process” of developing a solution before we’ve understood a single thing about what the customer needs.
And it’s not that we don’t have 100% of the requirements either. It’s basically a we don’t talk to the customer or perform market research to know where we should take the product, so I’m going to make up features at an absurdly abstract level and no you don’t need to meet to talk about it, just start working. “The requirements will come later”, they say. From whom exactly? 🤔
I feel this pain. My last assignment we went through all the process of grooming and story pointing, but the documentation was voluminous and we were all expected to be 100% conversant in each story. Then a story would take hours to groom and the testers weren't exactly sure how to test it, so the developers would have to basically write the end-to-end test cases along with detailed explanations of what could and couldn't be tested (arbitrary xml format, so some things you had to change independently to test, while other things would never be changed independently in prod so they weren't valid to test. Also, sometimes the correct thing to do would be determined by the database schema (is nullable?) which would lead to vastly different behaviors.
And lastly, we had to commit to delivering these features on time and to let everyone know ASAP if a timeline was under threat. Well, sometimes I'd tell them the timeline was under threat before the sprint was even started. That pissed them off something fierce because then it made us look incompetent to the customer - well... if the truth hurts, fix something in leadership. Also we were expected to do "whatever it takes" to deliver stories on time. Whatever it takes means skipping the grooming stories and reading the documentation for the next feature because I have to get this feature out the door.
I've never worked in the game development industry, but stories like this are what has kept me out of it. "Crunch time" that lasted months? Fuck that. I'm fifty years old with 25 years experience. I have kids at home. I work my ass off for about 9 hours a day and every once in a while a little more, but I'm not about that fucking lifestyle over shitty business software to make a shitty business more shitty money.
So needless to say I'm between jobs at the moment... I'm so over terrible leadership.
I can sympathize with a lot of the struggles you’ve endured. I deal with a lot of the same things and I’ve started to look elsewhere because for a company of 15, this place is run like a train wreck.
Just got off my stand up meeting too, we have to cancel todays grooming session because the project manager is on vacation. How will we survive without them repeating buzzwords and nodding like they understand what we’re talking about??? No shared responsibility for this team. It’s back to the waterfall method; and my boss likes throwing rocks off of the cliff. Every man for themself!
okay, I've finished your software, it meets your zero requirements!
“Add support for another vendor”
“Okay I understood what you wanted perfectly without any context, I’ve also compiled, pushed, and upgraded your customers”
Some people think it will really be that easy. Maybe 50 years from now, but not with these generative models lmao.
Do you work with me? My boss refuses to do a market research, refuses to check our website usage, change features that users actually care about, because he doesn't seem value on it.
And good lord, i have to create a dashboard to show our analytics, but he doesn't talk exactly what is needed. I spoke multiple times, said what i needed from him and he still doesn't know
I might as well! Our website hasn’t been updated since 2016 and we don’t know what customers already have a solution to the one we sell. They’re running on decade old market research, and have pretty much abandoned the idea of getting large customers since our sales team (which is also my bosses, we don’t have a dedicated sales position) doesn’t perform. Instead my bosses are chasing after smaller customers that provide smaller profit margins?? And this is all on the assumption that they’ll be interested in our product when federal subsidies start to be handed out to our industries customers.
In short, we’re fucked.
Analytics? Customer empathy gathering? Market research? Why bother? They just saw a post on LinkedIn about a Blockchain ChatGPT AI Machine Learning NFT. You really need to keep your eye on the ball on how we can work together to shoehorn any of that in the product so I can seem smart posting it on LinkedIn too. I'm never gonna hit CEO gaining market insight. Gotta fleece everyone with the most fancy sounding thing I have no clue about today. Don't worry. I'll forget by the time you have it released and complain about the new maintenance burden though. We need more features!!!
You joke but one of my bosses did request we implement ChatGPT on our platform. It's a recruitment platform, and he wanted for ChatGPT to write the job descriptions...
And it's not like it has zero application. But vaguely gesturing at a trendy technology and saying "we sure gotta leverage this guy's" is not a feature. Why are you doing this? What is the downside? Should we be doing this? Let's do a ChatGPT is not a strategy.
"We shall make solutions to problems people don't even know they have!"
"What people?"
"All people! Including us!"
Exactly. We’re not inventing an iPhone or anything. Extremely niche software that is losing market space every month while the PM’s sit around and prioritize the wrong things, there’s no strategy.
THIS
Ah, yes. The problem commonly known as PM-who's-also-the-BPC-but-also-doesn't-understand-what-either-role-actually-does, or "PMWATBPCBADUWERAD" for short.