My favorite coffee is roasted in small batches by Kus Beham and sold under the brand Giacobean. I’ve visited his facility in Yonkers. He keeps bags of raw beans in burlap bags in plain view, and he can tell you where they come from and how he roasts them to develop their trademark nutty flavor.

To control seasonal allergies, I take a generic, over-the-counter loratadine, Claritin’s key ingredient. In order to reach consumers, the drug had to have secured FDA approval, a process that eliminates 90 percent of all medicines that begin Phase 1, according to this report.

Later today, I’ll fly from New York City to San Francisco on Virgin America. It took me 10 minutes to find its “N-number,” basically the plane’s registration, which gives me access to its age, ownership, and maintenance history.

I could go on and on. The internet’s proliferation has enabled consumers to know where just about everything comes from. As a result, it’s easier than ever to know how things work, as well as how well they work.

In this context of unprecedented corporate transparency, what makes self-driving cars tick seems downright opaque. With the exception of Google/Alphabet Inc., which publishes monthly reports about its nearly decade-long autonomous vehicle experiment, it’s very difficult to get basic information about what any single provider is really doing in this space. Flashy press releases.

Last week, Uber announced that as soon as this month, 100 self-driving Volvo and Ford cars would start picking up passengers in Pittsburgh. Each one will include two Uber employees, including one behind the wheel in case the vehicle malfunctions.

‘m not ready for a ride in one of these Ubers. Before I’d ever consider being a passenger, I’d need basic information, the sort of knowledge we regularly expect of the most important things we consume. For example:

• Who wrote the self-driving software for Uber and its manufacturing partners?

• How long has that software been in development?

• In what conditions has it been tested?

• What are the results of those tests?  E.g. how many accidents have occurred and in what weather and at what time of day?

• What redundancies have Volvo and Ford provided in the event of a system failure?

• How much insurance coverage does Uber provide its passengers in the event of an accident involving one of these test vehicles?

• What training has Uber provided its backup drivers?

Consider the scrutiny that food labels still get today, most recently over whether a product may contain GMOs. If we put that much effort and energy into knowing what we put into our bodies, then we absolutely should know what’s behind an entirely new class of vehicle, one that introduces new and potentially deadly risks.

For the record, I believe self-driving cars are the future, but that future will require that many things we take for granted will have to change. As this article argues, they will need a new type of roadway and infrastructure to work. Providers will need to clearly explain how they operate, and just as they discovered with airbags and anti-lock brakes, some consumers will misuse them. The insurance industry will need to clarify how accident and personal injury liability will work. After all, even when the industry has met all these conditions, fatal accidents will diminish but not disappear.

The technology needed to make a self-driving car will be ready long before everything else is. Google’s has the best software chops and the most experience testing this technology, but even Google isn’t ready for passengers. That’s telling. If Google isn’t prepared, then I’m not convinced anyone else’s tech is ready enough to justify the equivalent of a pharmaceutical human trial.

For all these reasons, I won’t be taking a ride in a self-driving vehicle any time soon.