We invited a room of marketing operators to talk about AI, and the conversation kept circling back to everything AI can’t do for them. To be fair, no one there had AI fully figured out, from VC platform leads to a startup CEO who used to run global marketing at a SaaS company. We found just saying the statement above out loud to peers was a relief.

Small teams rarely get a space to convey what’s hard about their work, so Edgar Allan and Krista Peters of Premji Invest hosted a wine tasting at The Roaming Goat in San Francisco for marketing operators running big marketing programs with small teams. We called it an anti-networking event, asked everyone to leave the pitch at the door, and pulled our chairs into a circle to talk honestly about how we organize the work and decide where to focus. The wine was a lovely (and welcome) bonus.

Intimate rooms make it safe to say what everyone is thinking

Intimacy turned out to be the event’s biggest strength because in a circle like that everyone talks. As soon as one person named what they were afraid of, others followed, and the conversation went somewhere a large conference panel never reaches. 

An anti-networking event is a small, no-pitch gathering of peers who understand your job, built so candor becomes the main thing attendees take home.

The need for rooms like this keeps growing. A few years of remote work has left a lot of operators pretty lonely, and it's easy to spend a whole day alone with Claude or ChatGPT and forget you have a team and a community to lean on. A group where it's safe to admit what isn’t working gives you honest feedback from people who have done your job.

Other operators are seeing the same need. For example, Profound’s Nick Lafferty, whom we featured in our Building the Next Web series, runs a monthly dinner club for his peers. Our guests asked for the same thing on their way out: a standing community where they can speak candidly about what they’re struggling with.

The fear of being replaced by AI and a cheaper hire feels closer than it really is

The shared worry in the room took a while to surface: that a younger, cheaper hire paired with AI could do our jobs. Once someone said it out loud, shoulders dropped all around the circle.

One attendee described the pressure she feels to be technical. Her job is events and partnerships, yet she feels she should be building apps and automations too, because she worries someone who can do all of that will come in at a lower salary. The group pushed back  on that because what she brings is her history with people and her read on a room, and the tools can be learned or handed to someone else.

I see the same split inside Edgar Allan and across our clients: some people have gone all in on AI, and others resist it to keep their own thinking sharp. Both groups are protecting something real, but the tools exist because people spent decades figuring things out without them, and that hard-won context is what makes AI output worth trusting.

The room’s conclusion was that years of context and the judgment that comes with them are what an AI-savvy 22-year-old can’t replicate yet.

AI earns its place by clearing drudgery so people can spend their time on wisdom work

The AI uses people shared were practical and often personal. One guest built a tracker that alerts her when tables open at restaurants that book out almost instantly. Another prototypes ideas in Lovable to win buy-in before designers and developers build them for real. In every case, the person still decided what the tool was for and what to do with its output.

At Edgar Allan, we describe this shift as moving from knowledge work to wisdom work. 

Wisdom work is the part of a job that relies on what you've lived through: the context, relationships, and judgment that tell you which of the 10 reasonable ideas AI will give you to pursue. When AI takes on the repeatable retrieval and assembly of knowledge work, people get more of their week back for that judgment.

The hours AI gives us back are only worth something if we spend them on that judgment. Automating your morning recap and follow-up list can make you look like you're crushing it, but you're still accountable for what goes out the door. Deciding what's worth sending and how to frame it is the wisdom work those saved hours are for.

Human connection is what makes AI visibility count

Wisdom work matters just as much to how buyers find you. As buyers move to ChatGPT, Perplexity, and Gemini, it's tempting to hand content production entirely to AI, and audiences can tell when a brand has done that.

Here’s the point I keep making to our team:

The goal of AEO is to get humans to connect with other humans. A person still types the prompt, reads the answer, and decides whether to click.

If what the model surfaces about your brand sounds generic or wrong, ranking first in that answer won’t bring them to your site.

That's why Edgar Allan's Visibility Engineering and Optimization (VEO) program starts with brand clarity. A Profound-backed AEO audit shows where you appear in AI answers today, and VEO connects brand, SEO, AEO and CRO to close the gaps. Our AEO playbook walks through the first steps if you want to start on your own.

Almost every operator challenge comes down to how to spend scarce human hours

We asked every guest to share the challenge they were working through before they arrived. Most teams in the room were small, so nearly every challenge came back to the same question: which work deserves a person’s time. A few came up naturally. Here are three of them, with our take on each.

• How do I turn founder-led distribution into a repeatable system without losing authenticity?
  Systematize how the founder’s ideas get captured. Record their calls and conversations, then have someone on the team turn that raw material into posts. We do this at Edgar Allan with our Building the Next Web series conversations, and with an internal Claude skill we call the spark miner: we collect notes from Fathom or a call transcript, and it surfaces the moments worth posting, with hook options for each. That way, a human stays the source of every idea, which keeps the voice intact while the volume goes up.‍
• How much of our outreach can we automate?
  ‍AI can handle the research, list-building, and sequencing behind outreach. The message itself should come from a person, because prospects can tell when nobody human wrote to them, and a generic note rarely earns a reply.‍
• How does a VC firm build a brand and PR strategy that stands apart from other firms?
  Own a point of view and build community around it. This event is a live example. Premji Invest co-hosting an intimate gathering for operators does more for its brand than another panel appearance would.

Build the team first, then clear its plate

Most of the teams in the room were working with two or three people, so every hire, partner, and tool had to earn its place. These are the three moves we recommend for getting there.

1. Build your team like a chessboard. Each piece, whether an employee, a freelancer, an agency partner with a program like Edgar Allan’s Customer Success Program, or a tool, should cover a gap the others can’t.‍
2. Hire for what someone brings. Look past the perfect résumé for skills that add a new perspective and get the job done, because the right freelancer or hire may come from a completely different field.‍
3. Automate the low-strategy work. List every recurring task that needs no judgment, from report pulls to first-pass document reviews, and hand it to AI so the people you hired can focus on the decisions that matter most.

The teams that pull ahead will keep testing and comparing notes with peers, with their own judgment at the center of the work. Find a room where you can do that honestly, or keep an eye on our LinkedIn feed so you can join our next event.