We sat down with Anna Levitt who writes Anna | how to boss AI , and the founder of Bubble Boss Co., to discuss why AI adoption failures are people problems first and technology problems second, the data chaos hiding inside most marketing organizations that no one talks about, why mid-size companies are better positioned for AI transformation than enterprise, the “don’t become a coder — become tech literate” argument for marketers, why the Chief AI Officer role is now non-negotiable for larger organizations, the hidden talent problem and why the loudest voice in the room is rarely the right one, and how cognitive overload and burnout are quietly undermining AI adoption from the inside out. Plus: why AI layoffs are mostly an excuse, why the CEO job description is about to change fundamentally, and why a leader without a team has nothing to lead.
About the guest
Anna Levitt is the founder of Bubble Boss Co., a company that helps leaders and organizations transform with AI and adapt to the pace of change. Anne works primarily with mid-size and larger organizations navigating AI adoption — with a focus on the human and structural layers that most transformation programmes ignore.
She spent years in corporate leadership before leaving to build on her own terms, and now runs Anna | how to boss AI on Substack, where she writes about AI adoption, organizational design, leadership, and staying human in a disruptive world. She completed the MIT course on organizational transformation using AI and brings both practitioner experience and a coaching approach to her work.
Connect with Anna on Substack, her website, and LinkedIn.
About the host
Peter Benei co-founded AI-Ready CMO the daily intelligence platform for senior marketing leaders. Peter has been serving as a CMO, marketing leader, and consultant to high-growth B2B scaleups for the past 10+ years. He has a background in advertising, working with Fortune 500 brands.
Connect with Peter on LinkedIn, or read his newsletter.
Top 10 Takeaways
AI adoption stalls because of people, not tools. — The tool is the easiest part to implement. The hard part is aligning people so they see this as a value opportunity, not another disruption layered on top of everything else.
Most organizations have a data chaos problem they haven’t admitted yet. — Marketing, especially. Scattered data, siloed processes, undocumented workflows. AI can’t fix what hasn’t been organized first.
Mid-size companies are better suited for AI transformation than enterprises. — Less legacy infrastructure, more flexibility. The mistake is trying to copy enterprise-grade systems instead of designing for who they actually are.
Don’t become a coder. Become tech literate. — Marketers don’t need to build software. They need to understand how AI infrastructure works well enough to make informed decisions for their teams.
The Chief AI Officer role is non-negotiable for larger organizations. — You can’t stack AI transformation on top of existing marketing, HR, or IT roles. Someone has to own it as a full-time strategic responsibility.
Find the hidden talent — not the loudest voice. — The people who will drive AI adoption forward are rarely the ones with the most power in the room. They’re in the middle layer, doing the actual work.
AI layoffs are largely an excuse. — Most mass layoffs attributed to AI are restructuring decisions that would have happened anyway. The more important story is jobs not being created in the first place.
The CEO's job description is about to change fundamentally. — Fixed mindsets and old operating models won’t move the needle. Leaders need to reassess their value proposition — agility and vision are the job now.
Cognitive overload is quietly killing adoption. — People are already maxed out. Asking teams to adopt AI on top of everything else, without slowing the pace of change, leads to burnout, not transformation.
A leader without a team has nothing to lead. — Companies without people aren’t companies. The human layer isn’t a variable in the AI adoption equation. It’s the foundation of the entire thing.
5 Things Worth a CMO’s Attention
1. Why AI adoption keeps stalling — and it’s not the technology
Anne has worked with enough organizations to have a clear diagnosis: AI adoption fails at the people layer, not the technology layer. The tools get bought. The pilots get launched. Then nothing sticks.
The reason isn’t usually resistance in the abstract sense. It’s that people haven’t been given a clear answer to the question, “Where does this take me?” When employees can’t see the personal upside — not just the organizational upside — adoption stalls. It becomes another initiative layered on top of an already overloaded workday.
The second problem Anne consistently finds is data. Most marketing organizations have never been forced to document how they actually work. Processes live in people’s heads, in old email chains, in folders no one has opened in two years. AI doesn’t clean that up — it surfaces it. And when it surfaces, teams realize the infrastructure for AI adoption was never there to begin with.
For CMOs driving AI transformation, this reframes the starting question. Don’t ask “which tools should we adopt?” Ask “Are our people aligned on where this is going?” and “Is our data infrastructure actually ready?” If the answer to either is no, the tool is irrelevant.
2. The case for mid-size companies — and the trap they keep falling into
Anne makes a counterintuitive argument: mid-size companies are better positioned for AI transformation than enterprises. Less legacy infrastructure, fewer entrenched processes, and more organizational flexibility to actually redesign how work gets done.
The trap is that most mid-size companies don’t treat this as an advantage. They look at enterprise and try to replicate enterprise-grade processes — the four-week onboarding, the multi-layer approval workflows, the org chart structures that made sense at 10,000 people but create friction at 200. They hire executives from enterprise backgrounds who bring those models with them.
Anne’s argument: one size doesn’t fit all, and the companies that will navigate AI transformation best are the ones that design for who they are, not who they’re trying to become. That means being honest about which processes are actually necessary and which are just inherited behavior from a bigger playbook.
For marketing leaders in mid-size organizations, this is the permission to build differently. The fact that you don’t have enterprise infrastructure isn’t a disadvantage. It means you can design AI-ready workflows from scratch rather than trying to retrofit them into a structure never built for this.
3. Tech literacy vs. coding — what marketers actually need to develop
There’s a narrative circulating that marketers should learn to code. Anne pushes back on this directly — and so does Peter. The bar is wrong. The right bar is tech literacy.
Tech literacy means being able to evaluate a new tool, understand how it works at a functional level, assess whether it fits your team’s needs, and make informed decisions without needing a developer in the room. It does not mean being able to build something. That’s a different skill set, and it’s not the bottleneck.
The reason this distinction matters: if you set “coding” as the upskilling goal, most marketers opt out. The bar feels too high, the path is unclear, and the ROI for their specific role isn’t obvious. If you set “tech literacy” as the goal, it’s accessible, relevant, and immediately applicable to decisions they’re already making.
Anne adds a second track to this: creative development. In a world where AI handles the volume and efficiency of content production, the skills that differentiate are judgment, taste, and creative thinking. Her advice — take an art course, invest in aesthetic training, and develop the parts of your marketing capability that AI can’t replicate. These aren’t soft skills. They’re the competitive edge.
4. The organizational design question — who owns AI transformation?
One of Anne’s clearest arguments is structural: AI transformation cannot be a side responsibility. It requires someone whose full-time job is owning it.
The current reality in most organizations is that AI transformation gets assigned to whoever is most enthusiastic about it — a marketing director, an IT manager, a digitally fluent HR lead. These people have existing jobs. They have teams to manage, results to deliver, and leaders asking them questions every day. Adding “drive company-wide AI transformation” to that job description doesn’t work.
Anne’s view: the Chief AI Officer role is non-negotiable for organizations of a certain size. Not because the title matters, but because someone needs to sit at the intersection of legal, HR, risk, technology, and operations — with the authority and bandwidth to actually drive change. The role is already emerging, but the job descriptions are still confused bundles of responsibilities borrowed from other functions. They’ll clarify. Companies that create this role intentionally now will be ahead of those that wait.
For CMOs, this has a practical implication. If your organization doesn’t have someone in this role, you are probably absorbing parts of it by default. That’s worth naming explicitly — both to protect your own capacity and to make the case for a dedicated resource.
5. The burnout problem that no one is putting in the AI adoption model
Anne’s most underrated point is about cognitive load. Organizations are asking their people to adopt AI, learn new tools, redesign workflows, and deliver on existing targets — simultaneously, at pace, in an environment where something new drops every week.
The human nervous system isn’t built for perpetual high-speed adaptation. Anne is blunt about this. Burnout isn’t a fringe concern in the AI adoption conversation — it’s a central variable that almost never appears in the model. Demotivated, cognitively overloaded teams don’t adopt new technologies effectively. They comply minimally and wait for the next initiative to replace it.
Her prescription isn’t a productivity framework. It’s a design question for leaders: are you pacing this in a way that humans can actually absorb? Are you giving your teams permission to go deep on one thing instead of chasing every new capability? Are you building in the kind of recovery time that sprint-based work requires?
The companies that will get AI adoption right won’t be the ones that moved fastest. They’ll be the ones that moved intentionally — that understood their people’s capacity as a real constraint, not just an HR talking point. Anne’s argument, ultimately, is that staying human through this transition isn’t a soft goal. It’s the whole ballgame.
