I think the robots are winning
And what you can do about it
Every founder I know lives a double life.
A professional Dr. Jekyll and Mr. Hyde.
Dr. Jekyll is probably the reason you started your company. He’s the main strategist, reads non-fiction, connects disparate ideas, and has deep conversations with customers. He wants to build something real, something with value. He does the hard work. The homework.
Mr. Hyde is the creature you become to survive the internet.
My job demands I spend time where my customers are. For me, as a founder at a small AI startup, that usually means LinkedIn, X, various communities, and other digital black holes.
And the moment I log in, Mr. Hyde takes over.
He sees the game and knows he has to play it, to chase vanity metrics, say contrarian things for engagement, and simplify complex ideas (because those are easier to digest by the masses).
He’s not doing the homework, but lets the robots do it for him.
This is the central disease of modern marketing and strategy. We’ve stopped thinking and started choosing costumes. And once you see it, you can’t unsee it. The entire “world” is a maaaasive Halloween costume party.
My Mr. Hyde problem comes from a simple, uncomfortable truth I've observed in myself and others:
You’re probably NOT making as many independent strategic decisions as you’d like.
If I can predict your entire content strategy from knowing you read one specific newsletter or follow one specific guru, you're not thinking strategically. You’re LARPing.
We’ve stopped building from scratch and started subscribing to pre-packaged identities. Each tribe has its own uniform, its own creed, its own playbook.
Here are just two tribe examples:
★ The "Thought Leadership" Package: Standard features include daily LinkedIn posting, contrarian hot takes, and a fundamentalist rejection of anything that looks like traditional marketing. Your personal brand is the product.
★ The "SEO-First" Package: This bundle comes with keyword-stuffed blogs, a quasi-religious obsession with “topical authority,” and treating search volume as gospel. If you can’t measure it with a keyword tool, it certainly doesn’t exist.
Each tribe offers a sense of belonging and a clear path forward. They quiet the anxiety of the blank page. The problem is, they're not your path.
The playbook was written by the people selling the game.
The PLG crowd pushes product-led content because that's what their analytics and onboarding tools need you to create.
The thought leadership tribe promotes personal branding because that’s the content that keeps LinkedIn’s lights on and sells personal branding courses.
The SEO tribe worships keywords because that’s what their expensive software suites are built to track.
What you need to do here is start building your own machine.
Become a scavenger. Walk through the scrapyard of every tribe. Grab a gear from the SEO folks, a piston from the PLG crew, a chassis from the brand marketers. Bolt together a machine that works for your specific situations.
How do you do that?
You ask better questions:
↳ What do my best customers actually read in the dark, when no one is watching?
↳ Where do they go (a subreddit, a Slack group, a trusted colleague) when they have the exact problem I solve?
↳ What specific conversations, phrases, or ideas consistently turn browsers into buyers? (Go ask your sales team. They know.)
Your strategy should be built from the answers to those questions.
It should feel like it came from someone who intimately understands your customer, not someone who read the same five blog posts as your competitors.
So what is this place?
This newsletter is my lab for Dr. Jekyll.
It’s called ROBOTS ATE MY HOMEWORK because blaming the system, the algorithm, or the playbook is the ultimate excuse for not doing the hard, human work of thinking for yourself.
Here, I’ll be documenting the real work: what we’re building at Yahini, what’s working, what’s blowing up in my face. I'll share the workflows, the mental models, and the insights from the trenches of building a marketing function inside a growing AI company. Plus, what’s working for other founders, because I’m always happy to steal some techniques (after all, we can’t 100% make independent decisions in business, right?)
If that sounds useful, stick around.