Charity Majors on AI, Determinism, Instrumentation, & Eating Your Broccoli – RedMonk

You are using an outdated browser. Please upgrade your browser to improve your experience.

Skip to Content

James Governor talks with Charity Majors, Co-founder and CTO of Honeycomb, about what AI is doing to engineering practices. Majors argues that non-determinism takes more discipline, not less, and every CEO who wants their AI cookies has to eat their broccoli first. Agents in production are finally forcing teams to do the instrumentation work they should have done years ago. They get into Chad Fowler’s Phoenix architectures and the idea that the system, not the code, holds the truth about how software behaves. Also: which problems AI has genuinely solved and which it hasn’t, why bragging about token spend is like driving a Hummer, and the last year of AI hype has been aimed almost entirely at individual productivity when the hard useful work is always about software development as a team sport. They discuss Honeycomb’s Canvas, the company’s AI-guided Observabiolity workspace. And finally they discuss the difference between performative outrage and actually artists in the AI era – and make some recommendations accordingly!

Honeycomb is a RedMonk client, but this video is free an independent content.

Links

LinkedIn: Charity Majors

X: @mipsytipsy

Transcript

James Governor (00:04)

hey this is James Governor from RedMonk i am excited to be here because i have a buddy Charity Majors from honeycomb and Charity always has lots of great ideas many of which he even puts into production because test in prod or live a lie

But yeah, you have been writing a lot. You’ve obviously been thinking a lot about the impact of AI. And in fact, so much so Charity that we’ll sort of get to this. But I had a thought today in another conversation, I was like, you know what? Charity may, well, we’re not gonna get to it. I’m gonna do it right now. Thinking about the nature of the harness and where…

one lives sort of in the stack. I kind of thinking that Honeycomb, arguably is going to be moving up the stack and we’ll have sort of like an operational excellence harness rather than just being an observability platform. Because if we think about all of one thing that you and your organization have done, I think over the years, he’s just done such a good job of like codifying engineering, knowledge and culture.

take something like testing in production. That is something that you have written about, have made very clear, have made as a set of virtues and values that frankly could end up being something that a machine could read and use as the basis of inputs and outputs and run your systems in production.

Charity Majors (01:51)

I like where you’re going with this. I think we have increasingly come to the realization that people’s change budgets are so low because it’s all the change budget is being sucked up by AI. And so most people have just decided observability is a solved problem. Yeah, we know our solution is shitty, we don’t have any cycles to try and make it better. It’s three pillars, it’s metrics, logs, and traces. That’s what we’re doing. that’s so…

As you know, we have never really thought that that’s the right way to make sense of your systems. But you can only, you you need to go where people are willing to accept change. And I think that the next great frontier is knitting together determinism and non determinism. What does even mean? What is it? You know, what I love about AI is that it is

finally forcing everyone to do all the things that they needed to do all along but couldn’t get the investment time or what like instrumentation with traces the trace is the source of truth the trace is part of the product which is something we’ve been saying for How long is that this needs to be a product decision not an infrastructure decision? Yeah, and I really feel like I feel like agents in production are forcing the point on so many of these. It’s so exciting.

James Governor (03:15)

That is exactly right. think so your argument about production excellence. I one of the things is there are lots of dimensions to this. And certainly when I’ve spoken to you about progressive delivery, it’s been on the notion of like, there are so many things that you need to do. There’s actually a set of associated disciplines being rather than just one thing. And they are possibly multiple entry points, but there are so many things you need to do. But there is no doubt that AI is a forcing factor where

Look, if you’ve got shitty processes, poor documentation, all the knowledge is in people’s heads, and then you… And it’s very hard to get the benefits if you haven’t done all of that groundwork. And this is groundwork that there are many reasons that we should have done this. there are every few years, I think a forcing fact comes along, but it’s pretty clear.

Charity Majors (03:53)

to your AI if it’s all locked.

James Governor (04:11)

that...