September 7, 2026
TechCrunch published a glossary this week covering terms like "opaque recurrence" and other AI jargon that's crept into everyday conversation. On its face, that's a small service piece. But it says something bigger: the language around AI is outpacing most people's ability to keep up with it, including plenty of businesspeople who are being asked to make purchasing decisions based on it.
I've sat in meetings where a vendor throws around terms like "agentic," "grounding," or yes, "opaque recurrence," and half the room nods along without a clue what's actually being promised. That's not a knock on anyone's intelligence -- it's a sign the industry is moving faster than its own vocabulary can be explained. When language gets ahead of understanding, it becomes a tool for obscuring rather than clarifying, whether intentionally or not. A vendor who can't explain what their product does in plain English is a vendor whose product you should question.
This is exactly why we think plain-language reference material matters, and it's part of why we maintain our own glossary for the terms that actually come up when teams evaluate business software. The goal isn't to compete with TechCrunch's list -- it's to make sure that when someone is deciding whether an AI feature is worth paying for, they can understand what it does before they sign the contract, not after.
Travis Kalanick, the Uber co-founder, is reportedly steering his company Atoms toward the robotaxi business, describing it as a chance to finish what he started. It's a striking move, given that Uber itself walked away from building its own self-driving cars years ago and now mostly partners with autonomous vehicle companies instead of building the tech in-house.
Whether Atoms succeeds isn't really the interesting part yet -- there's no product, no fleet, no regulatory approval on the table according to what's been reported. What's interesting is what it says about founder psychology in this AI cycle: the people who built the first wave of consumer platforms keep circling back to the hardest, most capital-intensive AI problems, as if the first attempt was just a warm-up. Autonomous vehicles are one of the few AI applications where the gap between demo and deployment is measured in years and billions of dollars, not weeks and a product launch. Waymo and Tesla have both learned that the hard way, and Kalanick knows the terrain better than almost anyone, having lived through Uber's own retreat from it.
For business readers, the takeaway isn't about robotaxis specifically -- it's a reminder of how uneven AI's timelines really are. Some categories, like AI-assisted coding or customer support automation, are delivering usable results today. Others, like full self-driving, keep being "a few years away" no matter who's building it or how much they've raised. If you're deciding where to place your own bets, it's worth being honest with yourself about which category you're actually in. That same instinct -- separating hype from what's deployable now -- is why we built our workflow automation tools around problems businesses can solve this quarter, not moonshots that might pay off in a decade.
Both stories point at the same tension: the AI industry is generating new language and new billion-dollar ambitions faster than most organizations can absorb either one. My honest take is that the winners over the next few years won't be whoever coins the most jargon or announces the boldest moonshot -- they'll be whoever makes the complexity legible enough that an ordinary team can actually use it.
If Kalanick does jump into robotaxis, do you think Atoms has a real shot at succeeding where Uber's own self-driving unit couldn't -- or is this a case of a founder chasing unfinished business for its own sake?
Sources