What The Amish Can Teach Us About AI
Silicon Valley has a pervasive mindset of new-ism. If it's new, of course you should embrace it. Smartphones? Social Media? Crypto? AI? Whatever the latest cool new thing is. Anything that hints at providing some marginal benefit has to be wholeheartedly embraced.
Contrast with the Amish. It's a misconception that they don't use any technology invented after a cut-off date in the 1800s. Instead, each new thing is evaluated not only on its purported benefits but also on whether it helps (or harms) the community as a whole. That's why they decided against cars (great for travel but makes you spend the weekends away from the community), but don't mind solar panels.
What we can take away from this is that technology needs to be evaluated on multiple dimensions, including long-term effects and whether it's beneficial on the whole. Too often, the only question is: "Does this make it cheaper?" or "Does this make it faster?"
Well, spamming everyone in your reach with an AI-generated outreach message is cheaper and faster than finding people worth reaching out to, getting to know their problem space and crafting a message that they'll be happy to receive. Is it long-term better?
Outsourcing critical thinking to AI might feel great in the moment, but you'll trade away the deep wrestling with the issue for the AI's performative certainty.
Skipping customer interviews in favour of chatting with an AI that you told to assume a persona is faster, cheaper, and less emotionally taxing. But before you know it, your assumptions are no longer grounded in reality.
On the positive side, using it inside a thought-out development process lets you go faster, cheaper, and safer, getting products into your customers' hands faster, and getting that all-important feedback sooner. So it always comes down to this question: Does it, ultimately, help?
