2 min read

What's after mimicry?

What's after mimicry?

I’m a sucker for a Steve Jobs keynote. And this speech from the International Design Conference in Aspen in 1983 is a classic of the form and eerily relevant for this exact moment in time. In the speech, he talks about the personal computer as a new medium of communication, comparing it to the radio and the television. He observes that “each medium shapes not only the communication that goes through it, but…the process of communication” itself. When a new medium arrives, Jobs notes, our first instinct is to copy what came before. Early television, he points out, was essentially just radio with pictures. But over time, as we understand the unique affordances of the new medium, and as people get used to what’s possible, the communication evolves.

This observation neatly ties into Marshall McLuhan’s famous phrase, "the medium is the message." While often interpreted to mean that the medium of communication influences how the message is received, I think it carries a broader implication: that, over time, the medium dictates the message. The content that succeeds in one medium often fails in another, and it takes experimentation and invention to figure out the message that will thrive. Indeed, to Jobs’ point, as television matured, new approaches were tested - multi-camera filming, live broadcast events, and new show formats (sitcoms, late-night talk shows, etc.) - that leveraged the power of the screen to engage viewers.

Today, we stand at the cusp of another potentially transformative shift in communication with the rise of AI. And we’re all desperately trying to figure out what the dominant “message” of this new medium might be. But I wonder if, just like Jobs observed, we’re still stuck copying what came before.

That’s certainly the thought that comes to mind to me when paging through the deck of yet-another-accounting-copilot, or reading the 50th funding announcement for an AI SDR in so many months. This can’t be it, can it?

Indeed, through this lens, even the somewhat radical idea of replacing an employee with an anthropomorphic “AI agent” and “selling the work” strikes me as hopelessly shortsighted. Not because the underlying technology isn’t impressive. Not because “agents” don’t have massive potential for impact. But because they’re so obviously mimicking the way the world used to work, rather than charting a truly new path.

Look, I genuinely hope that 2025 is the “year of the agent.” I’d love for my friend Gemini or ChatGPT to book my next flight for me. But my advice for startups is to leave the mimicry to the incumbents. Not to say mimicking what came before is necessarily bad. It’s a natural stepping stone. But the real opportunity is to take full advantage of the unique, weird, surprising capabilities of artificial intelligence. Lean into what makes AI different and new.

If you watched that talk from Steve Jobs linked above, you may have noticed that he was speaking to a room full of product and industrial designers. And he spent a large portion of his time pleading with the attendees to forgo jobs designing cars or buildings and come join the personal computer revolution instead. I think what he realized is that computers were hard to grok and good design could help the average person understand what made computers uniquely useful. Just like how the computer, with its invisible electrons whizzing about, was intimidating to people whose mental model of a machine had gears and pistons they could touch and feel, AI is similarly alien. And it will require just as much thoughtful design and purposeful invention to reach the mainstream.