Bloomberg recently interviewed the outgoing head of San Francisco's transportation agency -- Jeffrey Tumlin -- about the impact that self-driving cars have had on the city. Along with maybe Phoenix, San Francisco has the most direct experience. Robotaxis have already been operating in the city for four years.
It's an interesting interview. On the one hand, robotaxis have, according to Tumlin, gotten better than most humans at "seeing" and predicting the behaviours of pedestrians. They offer slow and steady law-abiding rides, which is arguably not how must humans drive. This is a safety improvement.
But on the other hand, robotaxis still represent a fundamentally inefficient use of roadway space. They take up just as much space as human-operated cars, but importantly, they offer a less frustrating driving experience. Meaning they tend to induce demand, much like ride-hailing platforms.
In a 2018 study by San Francisco County, they found that roughly 50% of the increase in vehicle miles traveled in the region was due to Uber and Lyft. So not surprisingly, there are important things that will need to be figured out as robotaxis continue to spread across our cities.
I also find the comparison in the interview between San Francisco and Phoenix to be particularly interesting. The former is walkable. The latter is not. And this seems to be creating a different experience with self-driving cars because robo or not, in Phoenix, traveling by car is pretty much the only option.
For the full interview, click here.
This is Map 2 from Toronto's Official Plan, dated January 2024. It shows the city's "urban structure", and one of the main things it tells us is where the city is anticipating growth. You have the downtown and central waterfront, urban "Centres", and then "Avenues." All three of these colored areas are where growth and intensification are encouraged. The problem, however, is that it still resembles a mono-centric city, with a dominant core and then mostly low-rise housing surrounding it. The Centres in the inner suburbs are relatively small and many of the Avenues are broken up without a lot of connectivity back to the rest of the city's grid. It sets us up for a divided city -- core versus the rest -- because different built forms naturally create different political priorities.
Now let's look at what Toronto has just proposed in terms of new Avenues (shown above in purple). This is the kind of thing that immediately gets me excited because, as proposed, it implies a significant upzoning for a large portion of the city and it creates a much more uniform urban structure. Here, we have a blanket of intersecting Avenues, which will open up a ton of new housing opportunities and make it far more feasible to build efficient transit and other mobility solutions across the city. In fact, I'd argue that this is one of the most important land use discussions taking place in Toronto today. It's foundational to moving us away from the anachronistic model of the car-oriented North American city.
Now we just need to make it happen, and then empower developers and builders of all scales to build housing and a mix of uses all along these purple lines. For more on this, check out the City of Toronto's Housing Action Plan.
Cover photo by Adam Vradenburg on Unsplash
Also on Conde Nast Traveler's list of the best places to go this year is Ho Chi Minh City. I've never been, but I would love to go. More importantly, though, check out the photo they used, which is of the city's first Michelin-starred restaurant, Anan Saigon:
From what I can tell, the building (and lot) is maybe 3-4m wide. And it looks to be 7 storeys (the top floor is partially concealed because of the angle of the camera).
On the first two floors you have what I am presuming to be the main restaurant spaces. On the third floor there is a bar. Maybe this is also part of the restaurant. And on the fourth floor you have Anan's pho restaurant (I checked their website).
Above that it looks to flip to residential. And then again above that it looks to be residential. Finally, on the top floor it appears to be more commercial in nature, so I'm guessing it's a rooftop bar for Anan. If anyone has been here, please let me know if I'm getting this right.
Regardless, this is a terrific photo and a powerful example of what's possible on small urban lots if you just let people build.
Hotels are an important city-building tool. Let me give you an example. Each year, Conde Nast Traveler publishes a list of the best places to visit right now. Here's the 2025 list. If you scroll through it, one of the 25 places you'll find is Marseille. And if you look at the first picture, you'll find a hotel/restaurant called Tuba Club.
Long-time readers of this blog will know that I'm a big fan of Marseille and that, back in 2021, Neat B and I stayed at this hotel. In fact, the whole reason we even decided to stop in the city for the first time was so that we could check out this exact place. It opened the summer before and I had seen it written up somewhere.
I can also tell you that if you flip through Monocle's new handbook about France and go to the hotel section, you will likewise find Tuba Club. The place is seemingly everywhere and now has a loyal following.
But here's the important thing: it only has 5 rooms! And the room we stayed in was best accessed via a window and a large rock outcropping beside it. There's no way this place would meet code anywhere in Canada. But boy was it cool. And because we loved the experience so much, we decided to go back to Marseille the following summer and check out more of the city. It didn't disappoint.
Now I tell people and write blog posts about how much I love Marseille. And it all started with 5 non-code-compliant rooms perched on a bunch of rocks in a sleepy fishing port on the south end of the city. A great hotel can help you build your city.
Last month, the Berkeley National Laboratory, under contract with the US Department of Energy, published this report estimating total data center energy usage across the country. It also forecasted future demand out to 2028.
As you can see, in 2018, total electricity consumption by US data centers was estimated at approximately 76 TWh or 1.9% of the US total. In 2023, consumption more than doubled to 176 TWh or 4.4% of the US total. And by 2028, this is expected to further jump to somewhere between 6.7-12% of the US total.
Here's some commentary from the report:
With significant changes observed in the data center sector in recent years, owing to the rapid emergence of AI hardware, total data center energy use after 2023 is presented as a range to reflect various scenarios. These scenarios capture ranges of future equipment shipments and operational practices, as well as variations in cooling energy use. The equipment variations are based on the assumed number of GPUs shipped each year, which depends on the future GPU demand and the ability of manufacturers to meet those demands. Average operational practices for GPU-accelerated servers represent how much computational power, and how often AI hardware in the installed base is used, to meet AI workload demand. Cooling energy use variations are based on scenarios in cooling system selection type and efficiency of those cooling systems, such as shifting to liquid base cooling or moving away from evaporative cooling. Together, the scenario variations provide a range of total data center energy estimates, with the low and high end of roughly 325 and 580 TWh in 2028, as shown in Figure ES-1.
This strikes me as being an important macro trend and a big deal. All signs point to more data centers being needed. And before we know it, they're going to represent a meaningful chunk of total electricity usage.
Note: TWh = terawatt hour = one trillion watt hours
I think that lots of people would like to live in multiple places around the world. I know I would. That's why when people get rich and have the means, they often start to buy second homes. To that end, here's an interesting concept out of Japan called Not a Hotel.
Their model is fairly simply. What they do is build incredible design-forward vacation homes across Japan and then sell fractional shares, while at the same time offering full concierge and management services.
The typical fraction is for 30 days (1/12th), but if you'd like, you can buy up to the entire year. Ownership gives you access to the property for the amount of days you've purchased, or you can trade your days and stay at other homes within the Not a Hotel network.
On the nights you don't use, the company operates the home like a hotel and the owner gets the benefit of reduced management fees.
Most of their homes are already completely sold out. But 1/12th of this home in Kitakaruizawa is available for US$490,000. And on the other end of the spectrum, 1/36th of this Bjarke Ingels-designed home is available for US$2,460,000.
Here's a video of the home:
Fractional ownership is not a new model, but it is still relatively niche. I also think that the way Not a Hotel is going about it -- with their focus on over-the-top design and architecture -- is pretty unique.
Today is the third day of New York's new congestion relief zone. And if you're curious to see how it's already impacting traffic conditions, here is a website run by Joshua Moshes and Benjamin Moshes, under the supervision of Brown University Professor Emily Oster.
The site collects Google Maps traffic data every 15 minutes for 19 routes leading into Manhattan (some of which are directly affected by the new relief zone and some of which are not). It then calculates average traffic times for each day of the week, both before and after the congestion charge.
Here is the Holland Tunnel on Sunday (which was day number one):
And here is the Holland Tunnel on Monday (which was the first weekday):
Already, we are seeing a meaningful reduction in average traffic times. Maybe demand is more elastic than I suggested yesterday. But obviously we're only looking at two days. So I'll check back in later once we have more data points. In the meantime, if you'd like to follow along, you now have a website.
Good morning. Well, it finally happened.
After decades of delay and negotiations, New York City finally implemented congestion pricing for the area of Manhattan south of 60th Street. This is a first for the United States, and so it's a big deal not just for the city, but for this part of the world. It went into effect yesterday, on Sunday at midnight, so that the MTA could work out any kinks before this morning's rush hour. And apparently everything went smoothly. Drivers are now required to pay $9 to enter the zone during peak hours (5am to 9pm during weekdays). The charge is also expected to rise to $15 by 2031. Of course, this is a highly contested initiative. Trump is still vowing to kill the program as one of his first acts in office and, as soon as the pricing came into effect, suburban drivers started protesting it in Manhattan. I thought Jarrett Walker had a clever response to this:
One of the common rebuttals when it comes to things like road and congestion pricing is this one: "yeah, this might work in cities like London which have great transit systems, but it doesn't work in our city because we don't have that and it will unfairly disadvantage those who have no other alternative but to drive." In fact, this exact excuse was recently raised by local politicians here in Toronto. But this is New York fucking City. It has the highest annual transit ridership in North America (beating out Mexico City by nearly 2x) and it has the largest system by total length. According to the 2012-2016 American Community Survey, 85% of people traveling to Manhattan's CBD (I'm assuming lower Manhattan here) also take transit. And only 11% drive a car. So what exactly is the problem here?
This objection also ignores the fact that, generally speaking, congestion pricing has two main goals: (1) to, of course, reduce traffic congestion and (2) to generate money for more efficient modes of transport. In this case, the MTA is hoping this new congestion relief zone will generate up to $15 billion that can then be reinvested in transit and other infrastructure. Demand for roads can also be relatively inelastic in the short term, meaning demand doesn't change all that much as the price moves up and down. This makes it a good place to find money for public infrastructure, but it might mean that $9 is too low to have a dramatic impact on traffic congestion. We will see; I'm sure we'll get some data soon enough.
My prediction is that this will ultimately have an impact on congestion and that people in New York will get over the $9 charge. They'll also come to appreciate the reduced traffic congestion within the zone. So I think this road pricing will stick, and my hope is that it will become an example for other cities in the US and across North America. Congratulations on finally getting this over the line, NYC. It was certainly a hard-fought battle.
Cover photo by Veronika Galkina on Unsplash
I've been reading this book before bed over the last few days. It's about the Wasatch Mountains in Utah, but more broadly it's about how snow works, why and where in the world epic storms happen, how not to get killed in avalanches, and how global warming is impacting our climate. It's written by Jim Steenburgh, who is professor of atmospheric science at the University of Utah and author of the blog Wasatch Weather Weenies. And I'm finding it really fascinating, even if there are limits to how granular I want to get on the science behind stellar dendrite snowflakes.
One of the main questions he answers is, of course, the title of the book: Why is the powder skiing and snowboarding so good in Utah? Is it in fact, the greatest snow on earth? (The State of Utah started using this very successful slogan in 1962 and it made its way onto license plates starting in 1985). It turns out that there's lots of science to support this claim, particularly when it comes to the Big and Little Cottonwood Canyons on the east side of the Salt Lake Valley. This is where you'll find resorts like Snowbird, Alta, Brighton, and Solitude.
Little Cottonwood Canyon has one of the most dramatic snowfall contrasts in the world. At its entrance in the valley, the average annual snowfall is about 100 inches. But drive 7 or so miles into the canyon to Snowbird (which during a snowstorm can be super treacherous), and the average annual snowfall increases to over 500 inches. Generally speaking, the average annual snowfall in the canyon increases by about 100 inches per 1,000 feet of elevation gain.
One of the reasons for this is that the terrain surrounding the Cottonwoods is both high and broad, which means that it is exposed to storm flows coming from nearly any direction. During "stable storms", this can also create a blocking front, where storms get caught on the windward side and continue to dump in one place, instead of passing over the mountain. This is one of the reasons why there's this saying: "It doesn't need a reason to snow in Little Cottonwood Canyon; it needs a reason to stop."
Another very snowy place in the world is Japan's Hokkaido Island. In fact, Jim argues that if there's any place that could give Utah a run for its money with the claim of the greatest snow on earth, it's here. Based on historical data, Hokkaido Island has a 90% chance of at least 100 inches of snow in the month of January. This is a lot more than Utah, though snowfall falls off more quickly outside of the peak months.
So statistically, if you want the highest probability of powder snow during the month of January, Hokkaido is the place for you. It also happens to be where we're going for our annual ski and snowboard trip this year. I'll be sure to report back and confirm whether this is true or not.
The last chapter in the book is on climate change and he starts by stating the obvious: global warming is real. The climate of the Wasatch Mountains today is demonstrably warmer than it was when they were mining silver in Little Cottonwood Canyon in the late 19th century. That is bad news for skiing and snowboarding, and already in the Western US, declines in the average snowpack below elevations of 8,000 feet have been observed.
The good news is that there's yet to be a clear and consistent trend above 8,000 feet (at least according to Jim when he wrote the first edition of this book about a decade ago). So stay high up, my friends. If you're interested in this topic, or climate in general, I would highly recommend you check out Jim's book. Or at the very least, his blog. I'm going to take my copy of the book and leave it at Parkview Mountain House the next time I'm there so it's available to everyone who visits. It's fascinating stuff.
Cover photo by Alex Moliski on Unsplash
I was just introduced to a development firm based in Los Angeles called SuperLA. They are focused on sustainably built infill apartments and last year they completed what is the first ever mass-timber multifamily housing project in Southern California. Located at 3520 Marathon Street in Silver Lake and called the "Bungalows on Marathon", the project is 3-storeys (with parking underneath that takes advantage of the slope of the site) and has 9 homes.
Six of the homes are one-bedroom (~660 sf of interior space) and three of the homes are two-bedroom (~1,320 sf of interior space). It's a beautiful project. And because it's LA, all of the building's circulation is outside and tucked toward the back of the site. It's also not a huge site -- my rough Google Map take-offs have it at approximately 12m x 40m. So let's call it the equivalent of two Toronto single-family lots.
Based on suite count, this is similar to the kind of density that you could get in single-family neighborhoods throughout Toronto. However, the built form here on Marathon is decidedly more urban. Despite its horizontality, LA is not as low-density as many might think. For more on SuperLA, here's their website. They've also done a great job with their brand and identity.
Photos via SuperLA