So I’ll be perfectly frank in saying that I’m writing this on the long haul flight back to the USA. Got a bit behind on the blogs, it happens. Also, no huge revelations or intellectual (as my peers have been calling it) ideas, so a bit more objective reporting here. Also, since most of day 10 was travel, and a lot of day 11 was presentations, there’s not a whole lot to report. So I’ll combine.
On Day 10, we started the day off with a long drive from Pohang, south to Ulsan. We spent most of the drive on the main highways, but a bit of it was along a coastal road just a few hundred feet from the sea. Lots of beach vibes; photogenic towns, rocky islands topped with greenery, sea-facing traditionally inspired houses, etc… If I ever come back to South Korea, I will definitely come here for the beach (either this or Jeju Island, although the latter is a lot more inconvenient and expensive).
We reached Ulsan shortly after and ate lunch (and I did a bit of sketching) at the small museum exhibition, “Hyundai culture hall”. After the museum exhibit, we went through a tour of the Ulsan plant in a format very similar to the POSCO plant tour (bus through most of it, then take a walk through one small section). Given that only one of the two factory tours had blast furnaces hundreds of feet tall and glowing thousand-degree slabs of steel, you can imagine which one was more spectacular. However, the Hyundai plant was still a pretty cool sight. I was surprised by how strong the workers benefits are; not only does the campus have free cafeterias, on-site clinics, and transportation services, our tour guide also told me, in response to a question I had about further automation on the factory line, that the labor unions were a big part in impeding the introduction of more automation and robotics in the assembly process. I made the classic argument that you’re supposed to make: “Won’t it open up higher skilled jobs such as robot software programming and robot maintainance?” and got the classic responses; not enough new jobs and “unskilled” workers may not be able to get skilled as easily.

After the Ulsan Plant tour, we took the KTX to Seoul, arriving in time to visit the Namsang tower at night. It’s a tower observatory, similar to ones that exist all around the world, and from it’s position on the top of the hill about Myeong-Dong, it had been watching over us for the entire trip. Good to finally make a visit there, since we had also been talking about it for the entire trip. I’m sure others have shared photos, so I’ll refrain from it. But if you do go to Seoul, it’s certainly a good way to get a sense of the sprawl of the city, especially if you haven’t been to a such a massive metropolis before. To round out the night, we did some light shopping in the night market. A pretty successful day in all respects, even considering the long time devoted to travel.

The final “real” day of our trip was Day 11. The schedule was surprisingly filled out for the final full day of the trip. We started with a visit to TOPIS. TOPIS is a “smart city” system which uses a huge amount of collected data about the city (primarily traffic and metro data) to help make the city more efficient and safe. And in theory it sounds fine, but when we saw exactly how the data is collected and what is done with it, we ended up asking a lot of questions, particularly about privacy. Our presenter showed us a live demonstration of controlling a CCTV camera on the street; she panned around, zoomed in, and was able to get surprisingly clear live footage. A second demonstration showed her tapping into a camera inside a taxicab. When we started watching the live footage, the driver of the cab seemed to realize he was being recorded (there must be some notification), and waved at the camera. Now these aren’t available to the general citizenry (that doesn’t make it okay), but what is available to ordinary people, potentially even to anyone on the internet from any region, is the ability to look at live CCTV footage from the streets (without controlling the cameras), presumably to check traffic status, weather condutions, etc… I assume (hope) that users cannot see past video footage, and the footage is a bit grainy, but those doesn’t make it better either. Our main concern (which was lost in translation to the presenter) was that you could, if you were skilled with the TOPIS website, stalk someone pretty easily. Simply watch one CCTV camera until you see their car leave their house, then switch between cameras to follow the car as it moves from frame to frame. It might be difficult to do as a single person, but with a group of people, or with some predictive AI agent, it might be pretty simple to do. As a matter of curiosity, I’m going to try to see if I can do it at some point. If it is possible, then I’m not sure exactly how TOPIS keeps people from being able to stalk anyone in the city.
Another point of concern from me (but hardly a new one) is that the data is not analyzed directly by TOPIS, it’s managed by a private company. The issue of private companies having access to citizens’ data is not a new concern (see: Palantir), but I imagine it’s a larger concern in Korea, where a company like Samsung, which already wields a huge amount of power in the country, could gain even more power with access to the mobility data of all the citizens of Seoul. The lobbying power of such a conglomerate to influence data privacy rules is dangerous, after obtaining such data, they could change the laws of how they can use it. I already got the feeling that Korea is a bit of a surveillance state, so I’m scared to see what happens when all that data is in the hands of a private company with private interests.

Anyways, after TOPIS, we went out for hottoek and bingsu (Korean snacks), did our final presentations, attended a cooking class (slightly underwhelming), then went out for the last night in South Korea, finally having mastered the public transit enough to get back to the hotel without any navigation apps. It was a bittersweet last full day in the country.