In the industrial revolution you suddenly had all this technology that could replace manual labor. But it actually took a while for companies to adapt and make

In the industrial revolution you suddenly had all this technology that could replace manual labor. But it actually took a while for companies to adapt and make use of that. Marc Andreeson talks about this idea that new technologies follow a cycle. First they're ignored, then people fight them, then they settle on calling people names for using it. You can look back on the industrial revolution, or more recently the internet for an idea of what that pattern looks like. Some companies might adapt fairly fast but I suspect that will be the rarity. Instead what you'll have are small groups of individuals, highly leveraged by AI come in and make new products that wholesale replace non-ai companies. Some old companies will acquire new ones in time and survive, many won't. One early example might be Lensa[0]. They use stable diffusion on their backend against a paid iPhone app. Pretty simple stack there, not even training any models themselves. And yet - they're now doing $1M/day in revenue. We're going to see a lot more of these. Big companies will "try" too but they'll mostly just have meetings and powergrabs about trying. The 2020s are the decade of the startup. [0] https://apps.apple.com/us/app/lensa-ai-photo-video-editor/id1436732536 https://apps.apple.com/us/app/lensa-ai-photo-video-editor/id...