Insurance agents aren’t dead yet, but even underwriters participating in the personal automobile insurance market may be looking for another line of work within the next two decades, industry observers said at a recent conference.

“The most radical view of personal lines, given advances in digitization and advanced analytics, is that the need for people in the personal lines business is going away,” said Michael Pritula, director of the Global Insurance Practice of McKinsey & Co. “There are plenty of attackers that actually believe that they can do the whole business system with zero people,” he said during a session of the Property/Casualty Insurance Joint Industry Forum earlier this month.

“The technology impact in personal lines is far more profound today than in commercial lines. This is a really big deal in personal lines,” Pritula said, identifying “the two primary edges of that impact” as digitization and advanced analytics.

“In commercial lines, you have the impact of data, advanced analytics, but it is far less significant than in personal lines. And with respect to digitization, it is moving much more slowly than in personal lines,” he said.

“It’s quite a profound moment in the personal lines business. And I think so profound that this industry is going to change significantly,” he said leading in to his extreme assessment that people might ultimately disappear from the personal lines insurance business.

Pritula later warned carriers to look outward to technology startups in Silicon Valley and to technology giants like Google and Facebook. “There are others now outside of the traditional definition of carriers and reinsurers who look at every piece of the value-added that we’ve stewarded in the industry and publicly saying they would like to come after it.”

Another Forum panelist, Brian Sullivan, editor for Risk Information Inc., has been tracking the progress of one of those disrupters—Google—and said that Google’s two most potentially impactful technologies are moving forward slowly.

Inviting personal auto insurers in the audience to be “a little smug,” he reported that Google’s online comparison tool, Google Compare, has only launched in four states in the nine months since they set out to tackle the U.S. market. They had expected to be in about two dozen states, Sullivan said.

“They were going to X amount of traffic and they had 10 percent of the traffic they thought,” he continued.

“Agents are not dead yet and will not be for quite some time,” Sullivan asserted.

“Google’s entrance is very important, and ultimately it’s going to be very impactful. But they’re not doing well. They’re going very slowly because they underestimated the complexity of personal auto. It’s really, really hard to do,” he said, noting that a conversation he had with a Google Compare executive confirmed this.

Sullivan provided his take on the situation. “The [insurance] industry is often unfairly criticized for being slow. I don’t think we’re slow. I just think it’s hard,” he said. “It’ going to take a long time” for Google to meet its goals, he predicted.

But when they do catch up to incumbents in the online insurance world—the likes of CoverHound and Gogi, which “had a terrific 2015″—Google’s platform “will be transformative,” Sullivan said. It’s not only going to allow people to go online and get eight or more quotes in 10 seconds, but Google Compare will also make carrier recommendations “in the context of who you are,” he said. In other words, Google aims to understand what things are important to each online shopper in addition to price. “They’ll present to me the companies that they think will win my business because of the brand nature, the package of coverage, etc. Already we’re seeing in the very early days that it’s not commoditization,” he said.

Sullivan and Pritula also discussed a potential death blow looming for the auto insurance industry—from accident avoidance technology and ultimately driverless cars being developed by Google and others. But both said that extreme scenario is 20 years away.

“There’s no question that this stuff works, Sullivan said, giving the example of backup sensors. They “have pretty much wiped out in that car class people backing up into telephone poles. There’s a whole swath of the collision repair industry that’s lost,” he said.

“It hasn’t changed the [insurance] industry [yet], but you start adding a thousand of these little teeny changes and it’s going to” do so, he said. “Auto insurance is already smaller than it was on an inflation-adjusted basis 10 years ago.

Noting that there was an upward bump in auto insurance last year, Sullivan said the industry will see more surges from time to time in a recovering economy. “But there’s no way anyone can tell me that we’re going to have more accidents in 10 years than we have today. It just doesn’t seem credible,” he said.

“2040 feels like a good time for that changed world. And if I told you that the auto insurance industry would be cut in half in 20 years, that doesn’t feel crazy,” he said. “We’re talking about almost $100 billion in premium disappearing,” he said.

Agreeing that things are moving more slowly than insurers may fear, Pritula nonetheless said that “from a management standpoint, the management of a shrinking line of business requires a particular mindset. When you throw on top of that the consolidation and increased concentration—in personal auto in particular—it’s going to be very difficult as a smaller carrier to participate in a shrinking business in which the larger carriers gain certain scale [and] brand advantages.”

The good news for personal lines insurers is that while auto is shrinking, homeowners is increasing. There will be a point in time when homeowners will be larger than personal auto. “So we’re not losing personal lines. We’re losing personal auto,” he said, noting that lines like personal liability and cellular protection could grow.

Sullivan disagreed with Pritula’s suggestion that big carriers will win. “I do believe there’s going to be a lot of good small companies…There’s a lot of technology. There’s a lot of data. So a small, specialized company that learns to leverage the other parts of the market will do just fine—even without the billion-dollar ad campaign,” Sullivan argued.

“There will be fewer companies for sure. But there are companies that are really large that are going to fail because they’re just not smart enough. There are small companies that will be very smart, and they will thrive.”