​Read online​

Welcome to Unpacking Meaning. If you received this from a friend and enjoy it, ​subscribe here.

AI won't steal your job. It'll steal your agency.

Here’s something I’ve been feeling lately that I think a lot of you are feeling too.

Every morning I open my feed and there’s a new wave. OpenClaws that can run entire marketing campaigns. Vibe coding that lets any 13 year old spin up a SaaS over lunch. Someone even said apps are becoming the new info products (in the scammy sense), which is both funny and terrifying if you sit with it for more than five seconds.

And then, mixed in with all the excitement, the layoffs. Jack Dorsey’s Block just cut around 40-50% of their workforce. Thousands of people without a thanks to/because of AI efficiency.

So you’ve got this weird cocktail: boundless possibility on one side, existential dread on the other. And right in the middle, the vague feeling that you should be doing more with AI, faster, before someone else does.

I sat with that feeling for a while, and then I read something that reframed the whole thing for me.

When everything is “automatable,” the real scarcity isn’t code or copy.

It’s agency.

The motivational spell that doesn’t work

You’ve probably seen the meme: “You can just do things.”

It’s everywhere. LinkedIn, Twitter, tech podcasts. And for a certain kind of person, it makes sense, it can feel empowering.

But I can see how for a lot of people it can also feel confusing, empty, and honestly, a little insulting.

Because the map keeps changing. The tools keep changing. What felt like a solid career path six months ago now has a question mark over it. “You can just do things” sounds a lot like if I, as someone who’s played drums for 25+ years, asked a complete newbie to pick up the sticks and have a go at it.

I was inspired reading this piece by Octopusyarn and it made something click.

The core argument: agency isn’t a meme you adopt. It’s a worldview you participate in. And it runs deeper than most of us realize.

The author uses a great metaphor. Some people move through life like jazz musicians, improvising, responding to the room, making it up as they go. The majority are more like karaoke singers. Picking from the setlist and performing sanitized versions of what’s already been done.

I think it’s just what happens when the world feels non-malleable. When it feels like the options are already set and your job is to pick the best one.

Which, if you think about it, is exactly how most people use AI right now.

But let me not get ahead of myself.

Where agency actually shows up

The article goes “four layers deep” on what agency really is, using the philosopher John Vervaeke’s framework of “four ways of knowing”. I think this is the part that matters most for us as marketers and operators. So let me walk through it quickly, because each layer maps to something very practical.

Propositional knowing is the surface layer. It’s what you can say. Facts, concepts, mental models. “AI can help with messaging.” “Best practices say X.” “Here’s a framework for Y.”

Useful, but knowing about agency and having agency are completely different things. This is the meme layer. And it’s where most AI marketing advice looks like right now.

Procedural knowing is what you can do. Skills, workflows, muscle memory after lots of reps with a tool or process. You know how to run a research sprint, how to write a positioning statement, how to structure a landing page. This is deeper than propositional, because it’s embodied. But it’s still about executing known moves.

Perspectival knowing is where things start changing. This is what you can see. It’s your worldview shaping which possibilities even register as possible. When a great strategist looks at a market, they don’t just see the data everyone else sees. They see the gaps, they have a different view because they empathize with their ICPs, and they can create messaging from that perspective. That’s perspectival knowing, where you’re not picking from a menu anymore. You’re noticing doors that other people walk right past.

Participatory knowing is the deepest layer. This is what you become through acting with reality. It’s the jazz musician in the jam. Your identity, your way of moving through the world, evolves through the feedback loop of doing, learning, adapting, doing again. Your thinking → your market → AI tools → your thinking → your market… and on and on again.

The key is, agency shows up late, at the perspectival and participatory levels.

In other words, just because you know something, or you know how to do it, it doesn’t mean you can actually “just do it”.

So here’s the uncomfortable question: if agency is developed at these deeper layers, what happens when we hand most of our thinking to a tool that operates at the surface?

This is exactly how people are using AI

Most people are trying to get agency from AI by staying at the propositional layer.

That same article I mentioned actually uses the metaphor of a compass vs. a GPS, and I think it’s the most useful way to think about how we relate to AI, too.

GPS mode is how most people use AI today.

“Tell me the right answer.”

“Pick the best headline.”

“Generate the landing page.”

GPS mode assumes the destination is known, and the map is accurate. You plug in where you want to go and follow the turn-by-turn directions.

It works great when you’re driving to the grocery store. It doesn’t when the terrain keeps shifting, when you’re not sure the destination is even correct, and when the map was drawn by someone who’s never been where you’re trying to go.

Compass mode is different.

“Here’s my direction. Ask me questions, one by one to clarify my thinking”

“What am I NOT seeing that would make this fail?”

Compass mode assumes the map is incomplete and the path is made by walking it. You know roughly which direction you’re headed, but you’re reading the actual terrain as you move. You’re adjusting and learning as you go.

And here’s why this matters so much for the work we do.

Positioning strategy. Messaging strategy. Understanding your customer deeply enough to join the conversation they’re having in their heads.

This is all compass work.

It’s the work that still requires a human with judgment, with context, with the ability to sit in ambiguity and make a call.

It’s the work I do every day with clients, and I’ll be honest: I think it’s the work that will still need to be done by humans for a long time. Because the strategic decisions behind those words, the ones about direction, emphasis, trade-offs, what to say and what to leave out, those require the kind of perspectival and participatory knowing that AI doesn’t have.

To stay on top of the abundance of slop, we need to use AI as a compass not as a GPS.

So what does that actually look like for our positioning, messaging and copy work?

I’ll dive deep into it next week. Stay tuned.

In the meantime, I’m curious. How are you navigating this uncertainty? What are the patterns you’re seeing?

DISCOVERY

If you have a process, AI produces good stuff

I recently sat down with Caden Damiano on The Way of Product for a conversation that went deeper than the usual AI hype cycle.

We talked about research-first copywriting, the difference between archetypes and personas, why 70% of the work happens before you ever “write,” and how AI should amplify your own process rather than replacing it.

The episode drops soon. If you care about message-market fit, systems thinking, and using AI without outsourcing your brain, you’ll want to catch this one. In the meantime here's a quick writeup.​

Surviving the abundance era

Speaking of slop abundance, in this newsletter, Samir goes over his 4 principles for surviving as creators (and I think we all are, even as marketers and founders now). I particularly love number #2 The audience is the boos and #4 Make writing a daily practice.

The future of writing with AI

If you want to get a glimpse into what writing might look like soon, this piece by Katie at Every is a great starting place because she shares her thought process while building a new system. Like for most innovations, combining existing mental models with novel frameworks is often the most effective way to build.

• Related to this is the way new math research agents are starting to work, as explained by Chamath here. What if we saw writing as a communication problem to solve?

RESONANCE

Ryan Holiday, "The Definition of Success Is Autonomy"

Have a great weekend!

Cheers,

Chris

Chris Silvestri Founder & conversion alchemist

🙌🏻 Let’s be friends (unless you’re a stalker)​

When you're ready, here's a few ways I can help

Not sure where to start? Take our free message-market fit scorecard.​