After an onslaught of announcements about self-driving cars at CES, it’s almost hard to believe that even if you want to buy an autonomous vehicle, you can’t.

It’s a strange era, as car companies try their hand at something new: selling an idea, rather than a product that doesn’t yet exist, and won’t for some time. But throughout 2017, automakers, suppliers, startups, and giant tech companies spent billions of dollars on R&D, followed by millions of dollars marketing a good chunk of their R&D advancement at CES this week. But fully autonomous self-driving cars are not yet available, or even desired en masse by the general public.

We’ve said it before: your ability to buy a self-driving car is years away, and is not guaranteed. What you can buy today are cars with some rudimentary autonomous driving software, by Tesla and Mercedes-Benz, and vehicles with advanced safety systems, like the Volvo XC60, that employ automatic braking and steering.

But after much speculation, more automakers are talking publicly about their testing efforts and partnerships. Self-driving news dominated the transportation sector during CES. Ford made several separate announcements about advancements in its autonomous efforts, and was followed by GM’s Friday reveal of an autonomous concept with no steering wheel or pedals. Aurora, headed by Google’s former self-driving car boss, said it will support Volkswagen and Hyundai’s autonomous programs. Toyota showed new and improved LiDAR abilities and emphasized its partnerships with delivery services as a possible way to roll out autonomous pods. Among those companies is Pizza Hut, which appeared as an answer to Ford and Domino’s self-driving delivery program.

But the push to get more people inside of these cars is starting to happen. Some startups are beginning to test out autonomous vehicles in geofenced areas, like Voyage, which will test vehicles in a Florida retirement community. This is likely to push car companies and suppliers to follow suit and move beyond dedicated facilities like Mcity to move around actual passengers. The take away from CES and for 2018: the self-driving revolution will start with senior citizens and pepperoni pizzas.

Some automakers used CES to present a wild view of their future products, which are tied to self-driving visions. Mercedes-Benz teased the distant year 2030, by bringing along the Smart EQ Concept that we rode in after midnight on the Las Vegas strip. The truth is that no one knows exactly how self-driving deployment is actually going to work, which is why the amount of time spent hyping it up by a traditionally conservative industry is striking.

But the effort to win the game of public perception is real. It’s why companies give journalists rides in autonomous car prototypes including our test rides in both a Lyft car and the remote-controlled Smart concept, as the cameras rolled. It’s why we sift through dozens of self-driving pitches per week at The Verge.

For companies that rely on keeping their investors and shareholders happy, the push for autonomous development is not an obvious business strategy. These efforts seem more theoretical, and still very much in the academic research phase. As part of its CES announcements, Ford made a big to-do about its interest in developing open-source code for the industry, and positioned itself as a public partner, rather than a commercial car company.

”Ford’s vision for the smart city is an interesting premise, but at this point it’s not much more than that,” said Jessica Caldwell, Edmunds executive director of industry analysis. “Bringing this smart city to reality will require significant municipal cooperation and investment, and it remains to be seen if local governments share Ford’s ambitions. It’s admirable that Ford is taking a philosophical approach and is looking at how connected vehicles can change society for the better, however, this future is a long way off.”

Ford may describe itself as a mobility company and want to move in that direction, but today it makes its money selling vehicles, and answers to shareholders that are more consumed with profits on the F150 in the next quarter than making the world a better place.

One possible reason for the accelerated push to talk about their AV plans: panic. There’s an acute fear that if car companies don’t show their efforts to move forward, when the big change does come they won’t be ready. If they ignore autonomy, they could be bypassed by Tesla or Google, which has clocked the most self-driving miles in its spinoff company Waymo.

But there’s another, more nuanced possible strategy for the push to play up the self-driving: that someone has to do it. Or as they say in drag racing circles, “Scared money don’t win.” Legislators will not lead the effort to convince consumers that self-driving is something they can trust. Currently, there are no official self-driving regulations on the books. As autonomous efforts rise,they will be forced to sift through the complex set of new safety regulations on the horizon. It’s essentially a chicken-egg cliche. It’s impossible to govern that what does not yet exist.

So it’s up to companies to take the plunge and take the risk that self-driving will happen sooner, than later. It’s why so many startups are pumping money into this category, and why automakers want to jump in early on investments. Perhaps car companies have learned from past mistakes (see Who Killed the Electric Car) and are actually beginning to change how they do business.

But what all this hoopla hasn’t quite reckoned is how much will people want it? Is it the future or a speculative gold rush? The social benefits for special-needs communities and potential to make cities more efficient present a compelling argument. But will everyone else trade in their steering wheels and take the risks associated with new technology? Surveys so far say most people are not in favor of full autonomy — what the industry designates as level five. And the harder, messier part isn’t the existence of full autonomy, but the ensuing chaos of roads shared by drivers and autonomous vehicles.

So if car companies want to ensure that their investment is worthwhile, changing people’s perception of self-driving is essential. Security risks and safety questions fade when the familiar and friendly Domino’s self-driving delivery car pulls in the driveway. And by doing so, automakers can keep close tabs on how customers feel about welcoming self-driving technology into their personal driveways, or in fleet services they access. Automotive market research firms will have guaranteed work trying to analyze the data of perception as new business units are formed.

But as CES ends, automakers, tech companies, and the media that cover them will board planes headed for Detroit to kick off the North American International Auto Show, where they must shift gears and focus on selling SUVs, crossovers, and trucks this year. After several years of records sales, 2017 was a year when vehicle sales declined. All this talk about future, vision, and imaginative partnerships won’t matter if these companies don’t pay attention to what real people want now, and stay afloat until then.

“Last week at CES automakers tempted with visions of the future of transportation, but the Detroit show is all about the here and now,” Caldwell said. “While 2018 will be another strong year, it will be a down year — so automakers are using the Detroit show to serve up more of the trucks and SUVs shoppers crave. Funding mobility startups and investing in EV and autonomous technology isn’t a cheap proposition, and automakers have to keep the cash flowing in now if they want to build the strongest foundation for the future world they’re envisioning.”

Not to be outdone by the barrage of CES self-driving announcements, particularly by arch-crosstown rival Ford, GM chose to reveal its new autonomous vehicle early Friday, wedged between CES and the NAIAS media days, which officially begin with another GM reveal: a pickup truck event on Saturday evening. From what we know, the new Silverado will have nothing to do with autonomous driving, but it will be just as important to ensuring a pathway to the future for GM.