#but it’s this constant internal debate of not being productive enough versus trying to rest | Explore Tumblr Posts and Blogs

hopefulfestivaltastemaker · 4 years

Text

May 24, 2020

Here is my weekly collection of thoughts and things that happened.

Speed vs. Space

One of the basic tradeoffs in transportation planning is speed, which can be thought of as how far you can get in a time budget, versus how much space you take up during your trip.

Jarrett Walker has “the photo that explains almost everything”. Anyone who has spent any length of time in the urbanist community has seen this photo or some variant of it.

In the article, Walker explains that “In cities, urban space is the ultimate currency.” However, the photo that explains “almost everything” only tells half the story.

This is an isochrone map. You can make your own with the TravelTime app, which is a lot of fun. The image is centered at Orenco Station, Hillsboro, Oregon, where I live, a transit-oriented development that was built up around the MAX station starting in the 1990s. The blue region is all the territory I can reach in 30 minutes by public transport. The red region is all the territory I can reach in 30 minutes by car.

This illustrates the fundamental tradeoff. Contrary to what Walker claims, urban land is not a fixed commodity, but rather is based on what can be reached in a reasonable time. The 30 minute standard is Marchetti’s Constant, which is the observation that cities tend to be structured to afford most citizens a 30 minute one-way commute on average. The geography of the city is thus dependent on the primary transportation mode.

You can go the car-dependent route, and have a geographically large, low-density city, or a transit-dependent city with Manhattan density, or a super-dense arcology where internal transportation is all walking and biking. In each case, I estimate that the city can support a unified labor market of about 7-10 million people before either travel time or congestion get out of control. That is, uncoincidentally, about how large cities get today. Some metropolitan areas can grow to 20-30 million by relaxing the condition that the workforce has to thoroughly mix throughout the city, and by using hybrid transportation systems that simultaneously capture the benefits of each.

There are, of course, other factors that come into play: land use, energy consumption and CO2 emissions, infrastructure and building costs, safety, and aesthetics, among others. Our discussion of land use and transportation policy should be based on such tradeoffs and avoid simplistic “ban cars” slogans.

Oregon Metro’s Transportation Planning

Speaking of transportation, the subject is on my mind now, as Oregon Metro is currently preparing a bond measure for the 2020 ballot and had a work session on the topic this week. Last fall I submitted a comment that the process should be better tied with land use. If a neighborhood or city is to get regional funding for transportation improvements, such as expanded roadways or light rail stops, the region should insist that the city increase its zoned capacity to insure that the transportation improvement is fully utilized. Otherwise it is unfair to taxpayers in the rest of the region. I would like to see the federal Department of Transportation adopt a similar standard, as HUD has done with affordable housing grants.

Aside from that, I would like to see Metro show a bit more transparency on how they costed and estimated the benefits of proposed projects. Perhaps the information is there, but I haven’t been successful in finding it yet.

Otherwise, I don’t have much more to say now on the merits of proposed projects. What Metro is proposing now seems like a reasonable across modes and parts of the region. Said balance might be more the result of the political process than dispassionate analysis, but I wouldn’t expect otherwise.

Inclusive Wealth

Over the years I occasionally see calls for “better measures of progress” (than GDP) or measures of wealth that take well-being and the environment into account, or some variant thereof. Last week I wrote about how “growth at all costs” is a strawman, not an accurate representation of how economic policy is actually made. This week I took some time into trying to better understand these alternative measures.

One such measure is Inclusive Wealth, which the UN Environment Program tracks. It is a composite measure that adds GDP, changes in the quality of the environment and stocks of natural resources, and some measures of human well-being such as education. The index is meant to show that, with a broader definition of wealth, much of the world is not actually improving or even getting worse, in contrast to what the GDP metric tells us.

To start with the positive, I do appreciate the efforts made by ecological economics to quantify the physical basis of wealth. I think this effort is necessary and should be further developed. One of my early influences in energy is work by Ayres and Warr to quantify the role that energy plays in economic growth. They found that it can explain most of the change in total factor productivity in the United States from 1900 to 1970 (beyond which the explanatory power weakens). I do generally like the ecosystem service pricing model, though it has its limitations. And yes, it is important to ask the deeper questions about what well-being is and what causes it.

The Inclusive Wealth metric has several major problems, however, which render the numbers meaningless as a representation of “inclusive” wealth. The most glaring is that scientific and technological knowledge is omitted, except insofar as they are represented in the traditional GDP metric. This renders the calculations on natural capital stock useless, since natural resources are a function both of physical material and technological capability. Technology is the difference between a usable and unusable oil or mineral deposit, and technology is what allows us to harvest useful energy from sunlight, wind, uranium ore, waves, tides, the crust’s heat, or deuterium in the ocean. Contrast to the Simon Abundance Index, which shows real declining commodity prices despite (or because of) growing population.

I don’t understand the material on well-being well enough to judge its usage in Inclusive Wealth, though what I have seen elsewhere makes me suspicious. That might the subject for a future post.

In the end, Inclusive Wealth is something that I think could have been a valuable contribution to our understanding of the economy, but it fails in that regard. Behind positive rhetoric about natural capital and well-being, the report carries Malthusian arguments and betrays population anxiety. It makes me wonder if there ever could be a super-metric of wealth that is coherent enough to be used for policy and free of obvious ideological bias. I doubt it, and it should be recognized that policymaking is a pluralistic exercise. An attempt to aggregate mutually incommensurate values into a single metric necessarily masks value judgments with the illusion of objective measurement.

Interstellar Flight Concepts

Marc Millis has a presentation online now from the Tennessee Valley Interstellar Workshop last year. With a grant from NASA, he is building a database of proposed mission architectures for unmanned interstellar probes.

First, I find it wild that NASA is funding such a thing. Projects involving speculative physics, such as warp drives, are expressly included in the scope. It is good to see a small portion of our research efforts going to far-reaching projects, few if any of which are likely to come to fruition in this century.

Second, I like the idea of a public-facing database of proposed solutions. I would like to see much more of this across fields to help the public better find what is available and place it in context. The IAEA maintains a similar database for advanced nuclear reactors, which I have found helpful.

Infrastructure for Artificial Intelligence

I’ve also been thinking about AI this week. Last year, Eric Drexler published Reframing Superintelligence: Comprehensive AI Services as General Intelligence, which I found to be a useful corrective to the agent-based notion of artificial general/super-intelligence that is prevalent. If I understand the argument (which, I must say, is not the easiest to understand), he envisions that AGI can be stitched together by decomposing economically useful tasks into a tiled task space, where a narrow intelligence can be developed to perform each task, and AGI can be stitched together across tiles.

This is all quite debatable, but Reframing Superintelligence does point to the possibility of an AI infrastructure which, even if it is not structured as Drexler imagines, may be an important concept. To analogize to road networks, agents, such as DeepMind, might be the cars and trucks; what then, are the roads?

The best thought out answer I know of is the Semantic Web, an idea developed by Tim Berners-Lee (developer of the World Wide Web) as a machine-readable extension. It is, in effect, a method of encoding knowledge in a graph database of propositions in subject-predicate-object format. It is a very intriguing idea but one that has not yet gained traction in the way that advocates had hoped.

I can imagine the Semantic Web, or some analogous concept, playing an important role in the technology stack that ultimately leads us to AGI. It could also be instrumental in realizing the potential of expert systems and automated reasoning, techniques in AI that for years have languished with limited productive results.

0 notes