I recently attended the 4th annual transportation technology conference hosted by the Texas A&M Transportation Institute and had the opportunity to hear firsthand where autonomous vehicles (AVs) are headed from some of the foremost thought leaders in the space. I left the conference inspired but also struck by how wide the spectrum of thought was – with believers, naysayers and even deniers (those who believe AVs will never happen) all equally confident in their positions.

If the experts were so divided, I thought, how must those with a passing interest in the field feel? With such diversity of thought, coupled with the many misconceptions and misunderstandings out there, I thought I’d summarize the conversations and meetings I’d had into a “cheat sheet” on AVs.

This guide answers some of the most frequently asked questions about the subject, presenting both the optimistic and more skeptical viewpoints to arm you with enough knowledge and insight to hold your own in your next conversation about AVs.

When will cars truly be autonomous?

Fully autonomous firmware and software? It could happen by 2025, according to the National Highway Transportation and Safety Administration, which produced a helpful timeline charting the course from level 1 (basic driver assistance features) to level 5 (full autonomy).

There are even some optimists that believe we might see full level 5 cars being purchased or leased before 2025, based on the logic that technological advances tend to accelerate.

However, artificial intelligence has its limitations. While self-parking and lane-tracking features are useful add-ons, full autonomy might only be achieved after a lengthy hybrid period. The hybrid world, in which human-operated and autonomous cars coexist, could expose the limitations of the algorithms as they struggle to make the right decisions with full context awareness.

What can cities do to be ready for AVs?

Much needs to be done by cities and states to be AV-ready. City planners will have to take the following steps, in this order, to prepare for an increase in autonomous vehicles on the road:

1. Re-evaluate road markings to enable easy and unambiguous reading by radars, lidars, and other car sensors.

2. Upgrade urban furniture with advanced roadside units that continuously communicate information to cars about the status of the streets, upcoming events, and other relevant data.

3. Amend city regulations to reflect what autonomous vehicle should and should not do.

Some argue that cities might not need to do any of these things. A fully autonomous vehicle might be able to understand its surroundings regardless of road conditions, cleanliness, or configuration. After all, what’s the point of calling AVs autonomous if they have to rely on external inputs? Sophisticated onboard algorithms combined with advanced sensors might ultimately be all AVs need to detect and interpret roadway realities better than humans.

Will AVs be connected vehicles?

Undoubtedly, yes. An AV cannot be efficiently autonomous without the help of other agents, and for that it needs to be connected. Take traffic jams, for example: The car will need to be connected to a service that would alert it to upcoming traffic congestion, and from there the car could decide on a more optimal route to take.

A caveat to this argument is that if we rely on vehicle connectivity, there is a huge liability and dependency that comes with it. Just imagine a scenario in which a cyber-attack shuts down this connectivity and paralyzes every car on the road. Today, if a cyber-attack shuts the internet down, we would still be able to drive our cars.

Will AVs create more traffic congestion on our roadways?

If AVs deliver on their promise of taking you from A to B without you knowing how to drive, everyone who can afford it would use this mode of transportation. Since our roads were not designed for such high volumes of vehicles, you don’t need to be a hardened skeptic to predict that this would translate to more congestion, no matter how efficient the AVs are.

A more optimistic view of the future, however, is one with less congestion. AVs are designed to drive at the optimal speed to get you to your destination. And unlike humans, AVs don’t get distracted and won’t slow down or speed up unnecessarily. With AVs, some say we will have fewer car crashes and traffic incidents. As a result, congestion on the streets could be a thing of the past.

Is safety really going to improve?

Since the inception of the cars, safety has improved almost every year. From seat belts to anti-lock brakes to newer features like back-up cameras and lane-change warnings, all have contributed to better safety. With AVs, this pace could accelerate and see safety continuing to improve.

But as the cars themselves continue to improve safety, the sad reality is that pedestrian injuries and fatalities are on the rise. Our reliance on machines means we are more distracted than ever. And with more people around the world moving to cities and using all modes of transportation – walking, biking and scooting – the combination of personal and vehicular technology might be at odds for a transitional period.

Is AV a revolution or an evolution?

AV enthusiasts like to classify their advancement as a “revolution,” not least because of the wide-ranging impact on how and where people live and travel. They’ll predict the reversal of people living in cities and the end of frustrated commuters behind the wheel, as people move out to rural areas and start to use their travel time more efficiently. This brave new near-future is why AVs are often considered to be something of a revolution.

In reality though, research into the intelligent automated logic needed for autonomous cars began in the early to mid-1970s. In the 1980s to 1990s, Mercedez-Benz and EUREKA embarked on the PROMETHEUS project, which established the foundations for networked mobility. So to anyone following the story from its 20th-century beginnings, AVs can only be described as the next phase in its evolution.

Whether you call it a revolution or an evolution, there’s no denying the millions of driving jobs that will disappear. This includes Uber and Lyft drivers, who might die by the same sword they disrupted the taxi industry with.

Joe Boissy is CMO at Iteris, a global provider of informatics for transportation and agriculture.