It’s nearly impossible to scroll social media these days without seeing the newest, latest AI assistant that promises to jumpstart, catalyze, or accelerate your business. AI promises to “clear your inbox”, “go to market”, “build a lead magnet”. It promises to automate business creation and bootstrap your startup while you sip on coffee and watch the dollars pour in.
The narrative that AI will “automate” entrepreneurship – or at least the numerous mundane, tedious tasks of building a new venture – is so pervasive that it’s worth examining its premise, getting under the hood of what makes this story so compelling that I nearly pull out my credit card to trial each new promising freemium product.
The underlying assumption here is that building a new business is fundamentally automatable. Why is it so hard to build a business? Well, by this theory, it’s because it’s a lot of work – work that a human is just not ideally-suited for. Why do most startups (90% on a 10-year timeline) fail to ever turn a profit? By this theory, it’s because their founders just aren’t efficient enough, or quick enough, or intelligent enough to scale and succeed.
I’ve built and grown several businesses – at times succeeding and at other times failing magnificently - over the last twenty years, before and after the AI “revolution”. And my experience affirms that starting and scaling a venture is in fact a lot of work. Yet, the core work of founding a business is not automatable. It’s relational and intuitive, creative in the most human (and least artificial) sense. It’s experimental and iterative in complex ways that depend not just upon reams of data and computer feedback – the domain of AI – but upon human trial-and-error and professional mentorship. In short, building a business is a uniquely human venture.
Ironically, the best recent evidence for this comes from the tech industry, the very sector that today drives AI innovation: the great tech founders built their businesses the good ol’ fashioned way. Mark Zuckerberg scraped student photos from his college’s servers to build Facemash (later Facebook) on a scrappy tech stack, iterated quickly and monetized later. Sergey Brin and Larry Page revolutionized internet search (and later advertising, content distribution, and more) on an intuition that PageRank – a calculation of how linked-to a webpage is – was a better way to rank search results on the growing internet. They ran with their idea and iterated on it, testing ferociously and scrappily and (again) monetizing later.
Can a founder shortcut their way to business success using large language models? Would Facemash or PageRank have come out of a ChatGPT prompted conversation? Only if you believe that what stands between an innovator and their next big idea is simply a lack of accumulated historical facts and market research. No, innovation comes not from a dearth of data but from a yearning to solve real human problems.
Let’s not let AI derail our entrepreneurial spirit in the same way that social media derailed our social relationships – commodifying real friendship into a combination of clicks or taps on a screen, to sell back to us with a few addictive games and ads tucked in between DM’s and friendly posts.
Let’s remember that entrepreneurship is expansive, a lifestyle of social problem-solving. A commitment to building things that people want to buy and experience. There’s no shortcut for real entrepreneurs, no quick cheat AI tool to replace the honest hard work that brings hard-won success after many small fails (and some big ones.) I wish you all the joys and challenges of entrepreneurship along the way.
Cover: AI Image generated by ChatGPT (See, I'm not a complete Luddite!)
