017 AI for Operators

Mikayla LaRosa, Sr. AI Programs Lead at Airtable, X links

Hi there, 

Welcome back to AI for Operators. Here’s what we’ve got for you this week:

  • The Operator: Mikayla LaRosa, Sr. AI Programs Lead at Airtable (listen to the episode here)

  • The Essay: Operators and the Post-AI Org Chart

  • The Links: 10 links, including a playbook for eliminating operational debt with AI, a look at how Brex has integrated AI, and insights on how consumers are using AI right now.

The Operator

Mikayla LaRosa, Sr. AI Programs Lead at Airtable

Mikayla LaRosa, AI Programs Lead at Airtable, shares how she drives AI adoption across product and go-to-market and demos a workflow that doubled her team’s efficiency.

  • No technical background? No problem. Mikayla started out as a TV writer, turned into an Airtable power user, and now builds AI workflows, proving that non-technical operators can become AI experts, too.

  • Deliberate AI adoption program: company-wide hack days, weekly enablement, and market-aligned programming move teams from curiosity to usage.

  • Focused enablement sessions drive behavior change: after an 80-person live session, active usage jumped ~300% one month later.

  • Reframing AI as colleague: treat models like an intern you coach - be specific about outputs, but preserve your authentic voice.

  • AI-enabled customer service chat now assists the vast majority of incoming cases (90%+ are supported with AI!).

  • Airtable’s new Omni builder accelerates the creation of proof-of-concepts so anyone can build and test.

  • Newsletter-creation workflow with Airtable: 20-hour monthly process drops to 10 hours with same quality, less manual drafting and coordination.

  • Possible to profitable: known, defined processes and automate them to make them ‘profitable’ - aim AI at repetitive, high-leverage loops your team runs every month.

The Essay: Operators and the Post-AI Org Chart

What will the post-AI org chat look like, and what will that mean for operators? VC Tomasz Tunguz published a provocative essay recently outlining his view on what the post-AI org chat will look like.

His take? Most companies will evolve from an org chart that looks like a pyramid (one exec managing seven managers who each manage seven direct reports) to a ‘short pyramid’ (one exec just managing seven managers, who each manage a team of agents) or a ‘rocket ship’ (one exec managing seven managers who collectively manage fourteen direct reports plus a slew of agents).

In both scenarios, the company’s headcount shrinks dramatically. Does that mean operators’ roles will be ‘shrunk’, as well?

I doubt it.

First, there’s the transition to think about. Unless you’re building a company from scratch, teaching execs and managers to be AI-first is - and if you’ve tried it, you’ll know - no easy task, even when you’re dealing with smart, motivated people. They have to change their mindset, learn new tools, and create new ways of working, all while the technology is evolving every day. Operators are perfectly positioned to propose, facilitate, and measure that transition, identify the departments that should change first, figure out the tools they should be using, map processes, set up dashboards, and keep the exec team in the loop on the changes.

Second, if you make it past the transition, there’s the status quo running of the business to consider. Imagine a world where you’ve got seven managers who are each managing a couple of human direct reports and a handful of agents (each of the direct reports is probably working with a couple of agents, too). Suddenly, you’ve got more systems to monitor, performance benchmarks to set, and each of those manager is more crucial than ever, because there isn’t a person that can easily step into their shoes if they leave. Great operators’ combination of analytical acumen and EQ is exactly what’s needed for such a system to function effectively.

When it comes down to it, the fundamental things that operators are good at: translating ambitious leaders’ ideas into actionable plans, keeping a team pointed in the same direction, navigating technological and process complexity, driving rapid change while keeping the commercial needs of the business in mind, are not going away.

Now, if an operator is a glorified paper-pusher, only there to steward processes, put up roadblocks, play telephone between execs, and protect the exec’s ego, then it’s going to be increasingly apparent that they serve no real purpose. It can be easy to fall into that trap, so stay on your guard.

However, if you remain attuned to the needs of the business, use your analytical and social skills to gather as much company context as you can, and stay up to date with the latest in AI, you’ll have an exciting opportunity to increase your span of control, help your company make the leap from AI-curious to AI-native, and maybe grow the company - and your career - faster than you ever thought possible. Figuring out what your company’s post-AI org chat should look like is a great place to start.

Join Chief of Staff Network and Lindy

At a hand’s-on workshop on building agents on October 15 at 2pm ET

The Links

Practical

  • How consumers are using AI right now: Helpful deck full of insights from Dan Frommer, a leading consumer analyst (free with email signup).

  • Brex’s AI push: How the fintech unicorn is upending the way it works by combining the best of AI with its best people.

  • The Post-AI Org Chart: The traditions of 6-7 direct reports per manager, pyramid structures, and weekly 1-on-1s no longer hold true in a world where companies can scale to $100m in a few months and managers are as likely to be managing agents as they are to be managing humans. What comes next? VC Tomasz Tunguz has some ideas.

  • Eliminating Operational Debt with AI: How one operator automated a key contracting process with AI, saving hundreds of hours (with step-by-step instructions).

Perspectives

  • What if AI never replaces us? Radiology was seen as one of the easiest disciplines to automate with AI. Reality has proven to be more complicated. Now, radiologists are more in demand (and more highly paid) than ever. How did that happen?

News

  • ChatGPT Pulse: OpenAI makes a bid to make its product more of a daily habit, along with creating a new surface area for advertising revenue.

  • Notion 3.0: The productivity ‘startup’ makes its big AI play with the launch of agents that can create docs, build databases, and execute multi-step workflows. Will this be enough to take market share from Google and Microsoft, or will this turn out to be too little for Notion to remain competitive?

  • GDPval: OpenAI released a new set of benchmarks that measures model performance on a variety of real-world tasks. In other words, they’re trying to prove that the “AI will have no impact” crowd is already wrong. One of their findings: frontier models can complete the benchmark tasks 100x faster and 100x cheaper than industry experts. Uh oh.

  • Nvidia has a lot of cash: how are they going to use it? Apparently, partially by investing $100b in OpenAI and $5b in Intel (if I was Intel, I’d be a little offended).

Thanks for reading,

Tom Guthrie

If you liked this edition of AI for Operators, share it with another operator you know!