The AI graveyard: a running list of projects and startups that didn’t make it
I kind of like that someone is keeping score of the failures. The industry spends so much energy hyping the next model that a running graveyard feels like a small act of honesty. Apple's Siri has been the butt of AI jokes for years, and the repeated delays to a real overhaul have become their own running gag. OpenAI tossing a "super app" at the wall and missing is a different kind of lesson, because it's a company at the top of its game discovering that consumer products have their own brutal physics.
What connects these stories is that most AI failures aren't technical. The models work; the deployment, the timing, the product-market fit fall apart. Siri didn't fail because Apple lacks compute; it failed inside a corporate structure that couldn't ship a coherent vision for a decade. The super app misstep came from a company that built extraordinary tech and then fumbled the mundane business of what people actually want to open every day.
I'd want the list to be more, well, data-rich. How many of these are dead outright versus just delayed or abandoned quietly? Because there's a big difference between a project that's cancelled and one that's parked. Still, the exercise is valuable if only to keep egos in check. Most AI bets will fail, and pretending otherwise is how you end up on this list.