Another week, another headline about a Tesla on Autopilot doing something no human would defend.
This time it's the familiar script: a driver's eyes off the road, a system that was never fully self-driving, and a crash that a family says should never have happened.
The lawsuits keep stacking up, and each one chips away at the story Tesla has told for a decade.
The technology is genuinely impressive in a lab and on a clear highway at noon.
They are construction zones, faded lane lines, sudden brake lights, and a guy in a pickup drifting into your lane while eating a burrito.
A system marketed with the word "Autopilot" invites drivers to stop paying attention, and human beings are spectacularly good at accepting that invitation.
The naming itself is the original sin. "Autopilot" in aviation means something narrow and heavily supervised, but in a showroom it sounds like the car drives itself.
Tesla's own manuals and warnings say otherwise, buried in fine print most buyers never read.
That gap between the marketing and the manual is where the lawsuits live.
It's also where regulators are finally poking around.
Meanwhile, the rest of the industry is racing to ship hands-off systems of its own, which makes this less a Tesla problem and more an American problem.
We are handing over the most dangerous thing we do every day to software that isn't ready, then acting shocked when it fails.
Every automaker wants the credit for autonomy without the liability.
We've quietly decided that watching a screen at 70 mph is acceptable because the car has a camera watching us back.
That's a bet that a distracted person will behave better when monitored, which is a bet we keep losing.
We love the idea of the future arriving early, and we're willing to be the test subjects to get it.
But dreams don't have to face a jury, and the families in these cases do.
None of this means the tech is worthless.
It means the honest version of the pitch should be small and boring: it helps on highways, it does not replace you, and it will absolutely get confused.
Buyers deserve that truth before they hand over sixty grand, not after a crash makes it public.
A car that drives itself is coming, and it may well save thousands of lives someday.
But selling the promise before the product is ready isn't innovation, it's marketing with a body count.
Final Thoughts
Until the software truly earns the name, the name should change, and the drivers should keep their hands on the wheel.