Waymo’s model makes the case for AI, with measurable consumer benefits

Before launching an AI initiative, write the reader-facing sentence: ‘This helps you by ...’ If that sentence is weak, the strategy is weak.”

Guy Tasaka, managing partner at Tasaka Digital and Editor & Publisher’s technology columnist

A few weeks ago, my daughter and I took a Waymo in San Francisco. It was my first ride.

The first reaction was obvious: there is no driver! The steering wheel turns by itself. The car waits, merges, pauses for pedestrians and handles the strange little moments that make city driving city driving.

But after a few minutes, the novelty fades. Then the more important thing hits you.

Waymo is not asking passengers to trust AI because it lowers operating cost. It is asking them to trust AI because the experience can be safer — statistically, significantly safer. That is a lesson publishers should pay attention to.

Most media conversations about AI begin in the wrong place. Inside the company, AI quickly becomes an expense-elimination discussion: fewer people, cheaper workflows, more output with less labor. That may be an operational reality, but it is a terrible public argument. Readers, subscribers and advertisers do not wake up wondering how a publisher can reduce cost. They listen through the filter everyone uses: what is in it for me?

Waymo’s strategic move is that it turned AI from a labor-replacement story into a user-benefit story. The promise is not “we removed the driver,” but “the ride can be safer, because the system never gets tired, never gets distracted, measures what happens and keeps learning.”

Publishers need their version of that argument. Not: AI lets us cut newsroom cost. But: AI lets us serve you better, with more useful stories, deeper context, better personalization, more relevant newsletters, smarter alerts, and advertising that is more useful to the consumer and more effective for the business. That is the trust frame.

Waymo measures what matters

Through March 2026, Waymo reported 220.6 million rider-only miles without a human driver. On its public Safety Impact dashboard, the company says the Waymo Driver has had 94% fewer serious-injury-or-worse crashes; 82% fewer injury-causing crashes; and 82% fewer airbag-deployment crashes than human benchmarks over comparable roads in its operating cities.

Those numbers matter. But the more important part is how Waymo presents them. The dashboard breaks out miles, crash types, comparison benchmarks, city-level data, confidence intervals, methodology and safety research. Waymo does not simply say, “Our AI is safer.” It says, “Here is what we measured; here is where we measured it, and here is the claim the data supports.”

That discipline is what makes the user-benefit argument credible.

Publishers need the same structure. If AI helps produce more local coverage, show where coverage increased. If personalization improves newsletter usefulness, measure opens, clicks, saves, return visits and retention. If AI-assisted ad targeting improves relevance, show campaign performance and user response. If a newsroom uses AI to analyze public records or meeting transcripts, explain how that creates more accountability coverage for the community. The benefit has to be visible, measurable and connected to the reader. Otherwise, the audience will assume the benefit belongs only to the company.

The system is the driver, simulator and critic

Waymo describes its AI system as three connected pieces: the Driver, the Simulator and the Critic.

The Driver makes decisions. The Simulator tests controlled scenarios. The Critic evaluates performance, flags failures and creates signals for improvement. Real-world driving feeds simulation. Simulation improves the model. The model goes back on the road and generates more data.

Now translate that to publishing.

A publisher’s Driver is the daily operating system: story selection, newsletter packaging, homepage placement, audience segmentation, ad targeting, paywall decisions and customer support.

The Simulator is where you test before you damage trust: historical data, small audience segments, internal pilots, past campaign performance and previous subscriber cohorts.

The Critic is the part most publishers are missing. Every day, it asks: What failed? Why did it fail? What rule should change, and how do we know the fix worked? This matters because consumer benefit cannot just be a marketing claim. It has to be engineered into the operation.

If a publisher says AI will create better reader experiences, the organization needs a feedback loop that proves it. Did the story help? Did the alert matter? Did the recommendation serve the user or merely chase a click? Did the ad match intent or just follow someone around the internet?

A loop requires memory — a place where decisions, failures, fixes and outcomes accumulate so the system gets smarter. That is where publishers need to stop thinking about AI as a tool and start thinking about AI as an operating layer.

The failure is part of the trust argument

The best Waymo example is not a perfect ride. It is a failure.

In Austin, Waymo vehicles were caught on school-bus cameras illegally passing stopped buses with flashing red lights and extended stop arms. The incidents triggered an NHTSA investigation and public criticism.

The company acknowledged a software issue, said it had implemented updates and publicly framed the problem as part of the learning process. Vishay Nihalani, Waymo’s director of product management operations, told ABC News: “I don’t think people should expect perfection. What’s really important, though, is that we’re learning from all the different scenarios that we encounter.”

That sentence matters. He did not say the system was perfect. He did not hide behind the algorithm. He said the company found a scenario, analyzed it and changed the system.

Publishers will have their own school-bus moments with AI. A generative summary will miss context. A recommendation module will surface the wrong story. An ad unit will appear next to something inappropriate. A personalization system will optimize for clicks while damaging trust. The question is whether the publisher has a Critic.

Can you detect the failure? Can you explain it? Can you fix it? Can you update the rule? Can you tell readers, advertisers or staff what changed? If the answer is no, then you do not have an AI strategy. You have a risk surface.

Three things publishers should do Monday morning

First, reframe the AI conversation around user benefit. Before launching an AI initiative, write the reader-facing sentence: “This helps you by ...” If that sentence is weak, the strategy is weak.

Second, build a first-party event layer. Start smaller than a giant CDP project. Track the events that prove usefulness: article starts, scroll depth, newsletter clicks, subscription prompts, ad interactions, return visits, saves, shares and churn signals.

Third, create a Critic function. Every week, review what the system got wrong: failed headlines, weak newsletter segments, underperforming ad campaigns, AI outputs that needed human correction. Capture the issue, cause, fix and rule for next time.

Waymo is not interesting because a car can drive without a person behind the wheel. Waymo is interesting because the company built a system that connects AI to a consumer benefit people understand: safer transportation. Publishers need to make the same move.

The AI story cannot be, “We found a cheaper way to make content.” That confirms every fear readers already have. The story has to be, “We can serve you better than before.” More relevant information. More useful coverage. More context. Better recommendations. Better ads. Better products. Better accountability when the system gets something wrong.

The robotaxi is not the model. The user benefit is.

Guy Tasaka is a seasoned media professional with a 35-year track record of leading change in the industry. He has collaborated with renowned organizations, such as Macworld Magazine, Ziff-Davis and The New York Times, where he honed his expertise in research, strategy, marketing and product management. As the former chief digital officer at Calkins Media, he was acknowledged as the Local Media Association's Innovator of the Year for his work in advancing OTT and digital video platforms for local news organizations. He is also the founder and managing partner of Tasaka Digital, specializing in helping media and technology companies navigate business transformations using his extensive experience and forward-thinking approach. He can be reached at guy@tasakadigital.com.