Back in 2000, reading Kent Beck’s book Extreme Programming Explained was an epiphany for me. It’s all about the people! When we work together, have psychological safety and time to learn, manage our own workload, have direct collaboration with our customers, the magic happens! I was able to talk my way onto an XP team, and yes, we experienced the unicorn magic . The whole team approach, putting people first, let us eventually grow into “high-performing teams” with happy customers.
Now we have the tsunami of AI-assisted development. Maybe it’s just me, but what I hear about people now, is about how they’re experiencing cognitive and comprehension debt, feeling stressed out, getting laid off, replaced by AI agents and other tooling.
OK, I do know individuals who are having fun experimenting with AI. There’s research showing that organizations where teams were already performing well are enjoying and benefiting from their AI journey. Yet I don’t see the focus on people that I saw with XP and other flavors of “agile” development. It’s all about Claude, Cursor, Copilot, and other tools. True, the DORA AI Capabilities Model touches on things like psychological safety and user needs. But I don’t see the clear emphasis on people that I saw in the early days of XP.
Jerry Weinberg’s quote “It’s always a people problem” has become a cliché, but I think it is still true. Teams that were struggling before they implemented AI technology struggle even more now. Because they had a people problem, not a tool problem. Of course, the people in question are most likely to be somewhere higher in the chain of command. Perhaps those higher-ups think that if they could just get rid of the pesky people, their businesses would be golden.
Dear reader, am I just not looking in the right places? Have we come to that day when the robots replace us? If so, where the heck are our flying cars? I’d seriously love to get some other peoples’ views of this. Please leave a comment, or DM me on LinkedIn.
8 Responses
In software and in business, there’s a complex tangle of interconnected elements where the relationships are significantly modified by culture, leadership, and batch size. They are the three things to get right… and in that order and I believe it’s as true today as it ever was. Without a generative culture, metrics lose informational power and become motivators of (often strange) behaviour. Introduce AI in the absence of generative culture and it soon becomes a trumpet call for over-production of low quality work. Until organisations fix culture (generative), leadership (transformational), and batches (small… no… even smaller!) nothing else works as expected.
Yes! You explain it so well, Steve. I just hope this important truth does not get lost in the AI chaos.
Having played around with AI a bit, I ran into the same problems as before, just on a more amplified scale: fluffy requirements to begin with, humans rushing and not testing, growing tech debt, and realizing that it’s all about…
…people. People grabbed the AI as if it were a holy grail, yet this just hits Conway’s Law: organizations design systems that mirror their own communication structure, and, well, organizations are frequently pretty bad at… communication, thus what they may build with AI, too, could reflect that.
Thanks for sharing your experiences, Lina! That’s an excellent observation about Conway’s Law. It applies with AI technology, too.
Wow, thanks, Amit, those are such thoughtfully and thoroughly researched posts! I’m happy to discover your website/blog!
One thing that made Agile and XP transformative was the focus on collaboration, communication and shared ownership of quality.
AI can accelerate delivery, but without people, critical thinking and empathy, software quality becomes just mechanical output.
Technology evolves fast, but trust, teamwork and human context are still what make great products possible.
Thanks, Eduardo, I completely agree!