Oct. 17, 2022
"It's so important to realize you are planning a place for people to live, not just drive," says Zach Katz.
A Brooklyn-based musician and safe streets activist, an "armchair interest in urbanism" led Katz to Amsterdam, where he lived last year, absorbing the biking culture and infrastructure. Now, back in the U.S., Katz is bringing those lessons to Twitter with @betterstreetsai. Using OpenAI's artificial intelligence platform DALL-E 2, he transforms images of real roadways into pedestrian friendly renderings, complete with bike lanes, wide sidewalks, and green spaces. The tool allows users to easily "create realistic images and art from a description in natural language."
It didn't take long for Katz's designs to go viral; in less than a month, @betterstreetsai amassed over 15,000 followers. He attributes the reception to a simple fact: urbanism has gone mainstream. "The sheer amount of people who want better streets in their city is amazing," he says.
As his follower count climbed ever higher, Planning caught up with Katz to discuss his work, the power of AI, and how to change minds. This interview has been edited for length and clarity.
PLANNING: You use the term "better streets" in your handle. What does that mean to you?
KATZ: Streets that are designed for people instead of cars. Nearly all streets in the U.S. are, by design, dedicated almost entirely to the movement and storage of cars — as many cars as possible, going as fast as possible. This is horrible for the safety, mobility, health, and wellbeing of everyone, even those who drive.
A "better street" is a street that prioritizes the safety, movement, and happiness of people. Pedestrian promenades. Arterials with wide, smooth, well-lit, separated bike paths. U.S. planners should just copy what works in the Netherlands. It's a style of planning that's so wildly different from the way we have operated since Robert Moses sunk his teeth in decades ago.
But it's so important to realize you are planning a place for people to live, not just drive. The two paths are fundamentally at odds — the former unequivocally leads to a more pleasant world, while the latter leads to a less pleasant one. Get yourself a copy of the CROW and watch your city flourish in ways you never thought possible.
PLANNING: How does DALL-E 2 work?
KATZ: It works by uploading a picture, erasing the part you want to replace, then typing in a prompt for what you want to replace it with. This sounds simple — and it is, compared to the "old-fashioned" way of making a rendering — but there's some skill involved in knowing exactly what to erase and how to prompt it to yield good results. It helps to have a clear vision for what you want the street to look like.
PLANNING: What inspired you to use it for this purpose?
KATZ: There's a street in my neighborhood — Irving Ave in Bushwick, Brooklyn — that I've always thought needed a protected bike lane or to be completely pedestrianized. I was even thinking about launching a campaign last fall, but the need to pay someone to make renderings was a roadblock. So when I finally got access to DALL-E, after a few days of playing with it, I eventually thought, "what if I used this to transform Irving?" And that was kind of a mind-blowing moment, when I realized how well it could do that.
PLANNING: What goes into your transformations? How do you pick the locations, or elements to include or exclude?
KATZ: It's art. I just go with my gut: what looks good, what feels good. I pick a location that sounds fun to reimagine (with a slight bias toward cities and streets that I know) and try to envision what the best possible version of that street would be. If there's a bus, I think, "what if this was a streetcar, or an underground subway line? Then the street above could be a pedestrian promenade." Or, "Let's make the best possible bike lane — wide and well-protected — and only then we'll see if there's any space for cars." Again, just copying the Dutch.
PLANNING: What role do you think AI can play in urban design?
KATZ: Whatever it takes to plan better streets! AI's speed, creativity, and affordability will save quite a bit of time for planners, which in turn will save cities money, allowing them to transform more streets faster.
The designs will also help build consensus, but I don't think that's as important for planners to do, since most people don't know what they want. Like Henry Ford said, "If I had asked people what they wanted, they would have said faster horses." Your job as a planner is to plan good cities, so it's important to do that and not worry too much about disapproval. Think about how mad people were when Facebook came out with the news feed. Are those people still mad? Of course not. They love it now.
PLANNING: Why do you think your designs are going viral?
KATZ: Urbanism is becoming mainstream, which I've been aware of for a few years, but I think a lot of safe streets activists and politicians haven't realized or internalized yet. That being said, I'm still shocked at how much attention the account's been getting. The sheer amount of people who want better streets in their city — with an essentially religious fervor — is amazing.
PLANNING: To some, car-less streets may seem radical or even impossible. Do you think using this tech in such a visual way can change minds??
KATZ: Not only is it effective for changing minds, but in my opinion, it's the most effective way. "Seeing is believing" is a cliche for a reason. Most people don't know what a protected bike lane is. But if you show them a picture of what it would look like, they're like, "wow, that looks amazing. I would love that."
PLANNING: Do you have recommendations for anyone trying to leverage your designs to spark change?
KATZ: Just show them to people! Share them on Twitter, post them on reddit, put flyers on telephone poles. The designs are so powerful for showing people what's possible and getting them excited about it.
Ultimately, there's a person (the mayor, the head of the DOT, et cetera) who has the power to make something happen. And the more people demand something, the more likely it is to happen. So spread the word and demand better streets from politicians.
Also, we've just launched a new platform called transformyour.city, which is a way for anyone, anywhere to advocate for a better street in their city. Check that out.
PLANNING: Have you heard from designers or decision makers looking for ideas they could propose in their communities? Any advice for planners working to make their streets less car centric, particularly when it comes to gaining consensus?
KATZ: If you think a street would be better pedestrianized or with protected bike lanes, plan it. Your fear of disapproval is killing people and destroying the planet. We need better streets, you know we need them, so do whatever it takes to make them a reality. You'll end up getting consensus anyway, because people love car-free streets.
PLANNING: Which of your designs would you most like to see in real life? Which do you think would have the largest impact?
KATZ: Personally, I'd love nothing more than a protected bike lane on DeKalb Ave in Brooklyn, so I could feel safe biking to Prospect Park.
I think you could make an argument that any transformations in New York City would have the largest impact on a global scale. It takes so much political energy to make something happen here, and the scale is so vast, that the "if you can make it here, you can make it anywhere" principle applies. Robert Moses's highway-centric planning style spawned a global movement that continues to this day — and it started in New York. The same could be true for Jane Jacobs-style planning, too.
PLANNING: Overall, what do you hope the impact of this project will be, and have your expectations changed since you got it started?
KATZ: I'd like cities to build more protected bike lanes and car-free streets, and to build them faster. That's it.
PLANNING: Any questions you'd like to answer that we haven't asked?