Hollywood would have been proud. Elon Musk went full showbiz in his Thursday night unveiling of two new Tesla autonomous vehicles and a deeply unsettling robot. He even took over the Warner Bros. studio lot to have streets to demo the vehicles.

Cheered on by an audience of gushing fans, Musk rode the sleek silver “Cybercab” robotaxi around the set, stopping at traffic lights and avoiding passing cyclists. Next came the Robovan, a 20-person capacity vehicle that Musk claimed would slash the cost of transport. And finally, a gang of Optimus robots walked jerkily onstage, with Musk lauding them as a future babysitter for your kids and likely “the biggest product ever, of any kind.”

As I write this on Friday evening, the reaction hasn’t been good. Tesla stock plunged nearly 9%. Rather than exciting investors, the glitzy but short-on-details event only intensified worries that Tesla is late to the game on autonomous vehicles and at risk of losing out to competitors like Waymo and Cruise. Commentators noted that Musk has a history of being overoptimistic on timelines—something he himself acknowledged with a smile when predicting the Cybercab would be in production by 2027. Suffice to say, few people shared Musk’s optimism about the Optimus’s world-changing impact or its potential to revolutionize childcare.

To be sure, Silicon Valley firms have a long history of exaggeration and fakery in product demos. Commentators and investors almost expect tech companies to juice their stories so they can grab attention and fuel excitement. In his biography of Steve Jobs, Walter Isaacson revealed that the Apple CEO was using a barely functioning prototype of the iPhone during its iconic 2007 debut. Jobs had to memorize an exact sequence of touches and swipes for the demo or the phone would have crashed.

The Tesla product launch was fascinating. But I think the reaction to the launch revealed something even more interesting that the rest of us can learn from. And it’s all about the role that trust plays in innovation.

Trust Is Essential For Innovation

The fact is we just don’t trust Elon as much as we used to—even when he shows us something cool. There’s his erratic online behavior, his promotion of conspiracy theories on X, his far-right-wing dalliances, and his recent decision to go all-in on one side of a divisive presidential election. 

To see why that’s a problem for Tesla (not to mention Starlink, X, and SpaceX), it’s important to understand the difference between an invention and an innovation. An invention is any new solution that allows people to do something new. An innovation is an invention that actually has a socio-economic impact. 

Consider the hovercraft. It is literally a feat of biblical proportions to walk on water. But most of us have never seen a hovercraft in real life, and the vehicles haven’t had a big impact on our world. Compare that to the disposable diaper, a relatively simple product of paper and adhesive strapped to a baby’s bottom. Disposable diapers have had a massive impact on childcare, infant hygiene, the participation of women in the workforce, and even what we do to the environment. The hovercraft is the greater invention. The disposable diaper is the greater innovation.

For something to move from being an invention to an innovation, it needs to have an impact. Which means it needs to be trusted in order to win broad acceptance by society. Musk’s eccentric behavior may make him an outlier, but his growing struggle to win investor and consumer trust contains important lessons for every company trying to introduce something new. Gaining trust is tricky.

Trust Comes from the Form

The first lesson is to make new things seem familiar. At our core, most people really don’t like new things. Psychological research has found that we tend to be suspicious of anything that seems new, focusing more on the potential losses than the gains. In his classic Diffusion of Innovation theory, Everett Rogers found that people are less likely to adopt something new if it doesn’t align well with their prior experience and values. 

That’s why—counterintuitively—the most important thing when introducing something new is to make it look and feel like something old. Compact discs were designed to look like vinyl records to give consumers a sense of trust and safety. It’s also why microwave ovens were designed to look like conventional ovens and why early computers looked like a cross between a  television and a typewriter.

When Tesla’s first mainstream vehicle, the Model S, came out in 2012, it looked a lot like a Porsche Panamera and other luxury sports sedans. People needed to be able to understand what it was. Ten years later, electric vehicles were a lot more commonplace and Tesla could launch something that looked as futuristic as the Cybertruck. If the Model S had looked like the Cybertruck, I’m not sure Tesla would be in business today. 

Trust Comes From The Application

The second lesson is that when you have something new, you need to identify use cases for it that help people feel comfortable. 

The Cybercab and Robovan have merit in this regard. It’s easy for people to see how those vehicles could make their lives easier and more economical without a big upfront investment. It was a mistake to announce those products at the same time Tesla launched a small army of humanoid robots. Musk didn’t help his case when he suggested that the Optimus robots would be great for taking care of your children. Carrying my luggage? Okay. Taking out the trash? Fine. Playing with my babies? Terrifying!  

Trust Comes From The Source

Third, the trust needed for innovation is inherently bound up with people’s perceptions about the source of a new product or service. When I bought a Toyota Prius twenty years ago, hybrid electric cars were something new. People weren’t sure if they were going to last. I bought one anyway, thinking, “It’s a Toyota. They make reliable cars. I’m sure it’ll be fine.” I don’t think I would have bought a Pontiac Prius.

In the early 1950s, Mexico-based pharmaceutical firm Syntex faced a huge trust problem as it sought to introduce their birth control pill to the U.S. market. Contraceptive pills were still a new technology. There was no way that American consumers of that era were going to accept an unfamiliar new drug from a company based in Mexico. Syntex realized this and partnered with medical researchers at Stanford University. Then they moved their headquarters to Palo Alto. Most people don’t know that the birth control pill was a Mexican invention. That was by design.

Trust is such a powerful force that it can often defy logic. Microsoft’s efforts to make Bing competitive with Google’s search engine have long been stymied by people’s overwhelming trust in the latter. 

When Tesla launched the Model S, people vaguely knew Musk as an ambitious tech guy. Now, by contrast, he’s a divisive cultural icon known for his unpredictable and often outlandish behavior. That creates a trust gap. He’s also part of a tech elite in which public trust has steadily eroded over the past decade, largely due to the damaging effects of social media.

The erosion of trust in Silicon Valley is shaping up to be a major barrier to successful AI innovation—a technology that raises fears of the new and unfamiliar like few others. Trusted companies that can roll out safe, highly focused AI applications—such as Microsoft Copilot and Apple’s forthcoming AI applications—have been built heavily on their reputations for privacy.

Of course, some people are able to hold their noses and buy cool products from sources they dislike or distrust—like the guy I saw driving a Tesla Model 3 near my home the other day with a bumper sticker saying “F*** Elon Musk.” But that’s not everyone. And there are more choices than ever for consumers who want to buy high quality electric vehicles from more trusted sources.

The Elon show will go on, and his companies will continue to produce brilliant inventions. But the takeaway from Thursday’s flop is that real innovation hinges not only on creating cool products but on making the unfamiliar feel comfortable and trusted.