It's been a year since Honeycomb issued its AI mandate. Charity reflects on what that produced, why AI isn't special (it just amplifies what's already there), and shares the first of three new documents on Honeycomb's AI norms and values: how we do business.
By: Charity Majors
Has AI killed the SDLC as We Know it? - O11yCon 2026 Panel
Boris Tane, David Poll, and Charity Majors squared off on whether AI is upending the software development lifecycle, moderated by Jessica Kerr. This O11yCon 2026 panel is an hour of productive disagreement between three people who know the space.
[Watch Now](/resources/conference-talks/has-ai-killed-the-sdlc-as-we-know-it-o11ycon-2026-panel)
It's been almost exactly one year since we issued our [AI mandate](https://charity.wtf/p/in-defense-of-ai-mandates) here at Honeycomb, and we've been doing some reflection.
When we issued our mandate, it's not like we hadn't been using AI. We were the first in the industry to bake a feature powered by AI into our product, way back in May of 2024. Many of us had been experimenting and using these tools in our spare time.
But we believe that software is the killer app for AI. We believe that the way we build and maintain software is in the middle of a generational upheaval. We believe this is existential for us, and for our users. We had to take AI out of our spare time and put it front and center. We had to fund the change and make time for all of us to level up.
And level up we have.
What a year of this actually produced #
You are starting to see us release quite a lot of material about what we have learned and what has happened as a result. Just in the past month, this includes:
30 to 70 PRs a Day: How We Managed Not to Wreck Our Systems, by Liz Fong-JonesAI Amplifies Your Existing Practices: Lessons from Our Shift to an AI-First Strategy, by Liz Fong-JonesEmbracing the Code Review Bottleneck, by Fred HebertWhat Comes After Observability?, by Austin ParkerSpend More Time Talking to Humans, by Doug SooHow I Support Humans in the AI Era, by Ileanell Perez
... with many more sociotechnical topics in the pipeline to come.
Get your free copy and learn the foundations of observability,
right from the experts.
AI is not special #
Every one of those posts arrives at the same place from a different direction, and it's the thing we have returned to so many times that it stopped feeling like an insight and started feeling like a law of physics.
AI is not special, it is just a tool. AI amplifies what is already there.
Teams with strong learning culture and real ownership get faster and better. Teams that were already shipping slop now ship slop at a truly magnificent scale. Orgs with clear values apply them harder, and orgs without them find out. It doesn't fix you, and it doesn't ruin you. It turns the volume up on whatever you already were, which is a great deal less exciting than either the doomers or the boosters would like, and considerably more useful.
We sat down to write about AI and wrote about ourselves #
Honeycomb has always been a company founded on values, first principles, and the stubborn conviction that technology can be so much better than what people are used to.
After many intense internal debates about the ethics of using AI, we decided to write down a set of AI norms and values for ourselves. And again and again we found ourselves circling back to the same discovery: this isn't about AI, it's about us. How are our values being stressed in the era of AI? How do our standards apply, how must our principles adapt?
Three documents emerged from this process, and we'll share all of them with you over the next few days: one on how we use AI together at the company level, one for engineering that lays out our reasons for adopting AI along with north star goal and what support people can expect, and this one.
The one I'm publishing first is called “How Honeycomb Does Business,” written by Manny Alves, SVP of Sales and long time honeybee. It was written for our go-to-market (GTM) teams to sum up our approach to customers and business, but it applies to everyone. “We treat customers like people, not pipeline.” I could not love it more.
Values are not supposed to be aspirational slogans on the wall. Values should codify how things actually work at your company, when people are working well together, which should be most of the time. This is what that looks like.
How Honeycomb Does Business #
Honeycomb has ambitious goals, and we want to win. We also care deeply about how we win.
Our reputation as a thoughtful, technical, customer-focused company is one of our greatest assets. Every interaction with a prospect or customer either reinforces that reputation or takes something away from it.
This isn't about being less commercial. We should compete hard, negotiate well, create urgency, ask for commitments, and get paid fairly for the value we create. We want to build a business where Honeycomb wins because our customers win too.
AI doesn't create a new standard for how we conduct business. It raises the stakes on standards we've always had. These tools make it easier to communicate, personalize, analyze, and produce at scale. They also make it easier to be careless, impersonal, overwhelming, or deceptive at scale. We want the former without becoming the latter.
These are the principles we expect our GTM (go-to-market) teams to operate by. We create value before we extract value. We want Honeycomb to win because the customer is winning too. We should charge fairly for the value we create, but we don't optimize for extracting every possible dollar from a customer. A great commercial relationship should feel like a good deal on both sides.We treat customers like people, not pipeline. We respect their time, understand what they're trying to accomplish, listen before prescribing, and communicate like humans. This starts with the very first interaction. Personalization isn't inserting someone's name and company into an AI-generated email. There should be a reason we're reaching out and something useful behind it.We tell the truth, even when it makes the deal harder. We don't misrepresent what Honeycomb can do, hide meaningful limitations, manufacture false urgency, or tell customers what we think they want to hear. Sometimes the right answer is “we don't do that today,” “we're probably not the right solution,” or “I don't know, let me find out.” Trust matters more than any individual transaction.We earn the right to ask. Selling requires asking people to do things: take a meeting, bring in an executive, spend time on a POC, commit to a timeline, sign a contract, introduce us to another team. We shouldn't be afraid to ask, but there should be value and reciprocity behind the ask. If we're asking a customer for their time, money, or political capital, we should understand what they're getting in return.We compete hard without compromising who we are. Being customer-friendly doesn't mean being commercially passive. We should negotiate confidently, defend our value, challenge customers when appropriate, create urgency, hold people accountable to commitments, and ask for the business. We want to win, but we don't need to manipulate, mislead, or take advantage of customers to do it.We respect the cost of other people's attention. AI has made producing things cheap, but it hasn't madeconsumingthem cheap. Don't send a five-page document because it took five minutes to generate. Don't send 100 “personalized” emails simply because a tool makes it possible. Before sending something, ask: Is this useful? Have I read it? Would I want to receive it? We don't make our efficiency someone else's burden.AI doesn't lower our standards. We should use AI to research, prepare, analyze, summarize, learn, improve our writing, understand accounts, and remove repetitive work. But we're still responsible for the output.If your name is on it, you own it. Read what you send, check important facts, apply your judgment, and make sure it represents Honeycomb well.We optimize for durable relationships, not the next dollar. We want customers who stay with Honeycomb because we continue to create value for them, not because we've made leaving painful. This doesn't mean leaving money on the table. It means building commercial relationships proportional to the value we're creating and allowing those relationships to grow as the value grows. We want both sides to feel like they got a good deal.
Finally, we value good judgment over hard-and-fast rules. No document can prescribe every interaction, and we don't want one that tries. We hire adults. When you're unsure, come back to the principles: create value, tell the truth, respect people's time, compete hard, make reasonable asks, and think long term. Every interaction teaches our customers, prospects, and partners the kind of company Honeycomb aspires to be.