Scaling with AI: Cutting through the noise
What I’m seeing as a coach—and where I’ve seen it work
If you’re building or scaling a business right now, it’s almost impossible to ignore the AI chatter. Some say it’ll 10x your output. Others whisper that your job is next. Between Substack posts, conference decks, and “solo founder with $2M ARR” stories, the noise is deafening.
Tired of it? You’re not alone.
Even Mary Meeker’s recent 340-slide report - her first major publication in five years, acknowledges this paradox. AI is everywhere, with generative tools seeing record-breaking adoption. OpenAI’s ChatGPT hit 800 million weekly users in 17 months. Infrastructure spending is ballooning. Everyone is investing. But here's the catch: value isn’t keeping pace with velocity.
As Nate Eliason rightly puts it in his excellent Meeker recap: “Adoption is real. Enthusiasm isn’t. Rollout plans need to bridge that gap.” And from what I’m seeing in coaching sessions, that’s the gap we’re stuck in.
So where is AI actually landing? In theory? Everywhere. In reality? Mostly nowhere deep. At best, companies are scratching the surface: playing with tools, developing an ai feature here or there, collecting inspiration, maybe buying a few seats of something. But many scaling or enterprise orgs face real blockers: no time or space to experiment, budgets stalled at “we should look into this,” learning expected to happen “on the side,” infrastructure not ready (if you’re still on paper forms, AI isn’t your first priority).
And underneath all this? Product people are excited but deeply anxious.
These are the four personas I see most often when it comes to AI:
🧍♂️ The Skeptic: “AI isn’t ready or relevant for my work.”
🤔 The Worrier: “I’m trying… but I feel like I’m behind or doing it wrong.”
📚 The Collector: “I want to understand everything, just in case.”
💻 The Integrator: “I learn what I need, when I need it and make it work for me.”
Most leaders aren’t resisting AI. They’re just stuck in the middle - between curiosity and conviction, under pressure to move faster than they feel confident doing.
How scaling AI can work
At Multiverse 2.5 years ago, at the beginning of the GenAI hype cycle, we leaned into AI with intent. We didn’t wait for perfect alignment, we started with a hackathon. That event created the time, energy, and visibility needed to go from “AI is interesting” to “AI is valuable.” From that seed came Atlas, an AI-powered coaching tool, now a major product differentiator. It worked because we anchored AI to a real problem, had cross-functional ownership, not just hype, and gave it space to evolve, rather than treating it like a one-off sprint.
You can’t shortcut that process. But you can design for it. At Shopify, CEO Tobi Lütke made it clear from the top: AI wasn’t just a tool: it was part of every employee’s job to learn and use it. Employees are now required to justify why a task cannot be completed using AI before requesting additional resources. It wasn’t a nice-to-have, it was a cultural norm, reinforced by leadership. Duolingo’s CEO and CTO set a directive from the top that they were an AI-first company and what that meant for it’s employees. Yeah it might have backlashed a little, but the message is clear: this is important, and we’ll back you to figure it out.
Meanwhile, in my own work, AI has become my daily work companion:
ChatGPT refines my marketing, strategy and coaching content.
MagicPost helps me create LinkedIn content that resonates.
Granola helps me document notes and drive insights.
Gemini is answering all my Google questions.
I’ve experimented creating automated ways to onboard clients, and creating a new website using Bolt or Loveable, but to be honest have had really mixed results. When my squarespace (for now) does the job, and onboarding is a 5 minute task, I have to ask myself this: what’s the ROI - and how much time can I afford fu*king around with this?
What I’m SURE of (because product people are driving AI products) is what adoption will look like going forward is this: frictionless, ambient, inevitable. You won’t need to “learn AI.” It’ll come to you. These companies have every incentive to make this tech so intuitive it feels silly not to use it. And when that happens, the adoption gap won’t be about tooling. It’ll be about mindset, comfort, and culture.
So what can leaders of scaling companies actually do?
Lead by example. Show your work. Talk about your prompts. Normalise iteration and learning.
Create time to experiment. Whether it’s a hackathon or an hour a week or 20% time, make it visible and sanctioned.
Look for real use cases. Don’t chase shiny features, solve actual pain.
Be honest about readiness. If your systems are chaos, AI will just amplify the mess. Walk before you can run.
But most importantly: meet people where they are. You don’t need everyone to become a prompt engineer. You just need to keep the door open. Because like Meeker’s report shows: the tools are coming fast. And the companies who integrate them thoughtfully, not just quickly, will unlock the next wave of compounding value.
The hype is loud. The change is real. But what still matters most? Good leadership. Focused execution. Space to learn. That’s the part AI can’t do for us.
If you’re curious to hear more from product leaders, I run monthly group coaching for free. I create a safe space to share what’s really going on for them. Want to join the next one? Get in touch at emma@product-leadership-coaching.com