Interesting Problems

[xthread]

About twenty years ago, I sat down to have a good think about what I wanted to do with my professional life. What was an interesting and valuable enough artifact to spend a life building? I settled on helping build the Internet, which was a bit ironic, because twenty years ago most of the core technology was already in place, it just wasn't evenly distributed yet[1]. But I was thinking about this question the other day. None of my new answers are in the field of Computing; I'll touch on that at the end.

But without further ado, here are a bunch of things I'd like to see smart people thinking very hard about how to accomplish in commercially viable ways[2]:

Carbon Scrubbing - we've been dumping a lot of carbon into the air. Our ecosystem is poorly prepared for how much carbon we're dumping. Addressing this would be good.

Soil desalination/recovery - a lot of the world's most arable land has been destroyed over the centuries by inadvertently irrigating (or eroding) salt into the soil. Learning how to time- and cost-effectively reclaim that soil would be good.

Desert recovery - a permutation of the above. At the very least, if we can report 'recovering desert reliably costs $$$/hectare,' that becomes an available, rational choice, and perhaps people will then optimize it further.

Low-water agriculture - if we use a lot less water on agriculture, we can spend it on other things instead. This would be good.

Improved water reclamation / desalination - so long as cities are heat islands, we'll be drying microclimates as we put people into them. Being able to reclaim water more cheaply would make the world a lot better, both in cities and out; a lot of the water cost of climate change is that we're finding ourselves wanting to pump water from lower and lower quality water sources.

Biodiesel - this is getting a *lot* of attention at the moment, it's part of the generalized 'fossil petrochemicals are a very poor energy strategy over time, let's do something better' problem. There are better ways to do this (algae that can grow in hydroponic beds in the desert to produce biodiesel) and worse ways (refining biodiesel from corn - yeesh, at least use palm instead, please?!) to do this, but getting really good at it could improve the world dramatically.

High efficiency / low-cost solar power - this is one of the short list of ways out of our coal problem. Although it's getting a lot of attention, which is good.

High-efficiency / low-toxicity batteries - this is *huge.* It's what could free us from the catch-22 that the only ways to avoid coal involve scary amounts of toxics. Also getting a lot of attention, because small consumer devices have even more demanding battery requirements than transportation, so they're driving the costs down. Which is good.

Further out there...

Very high tensile strength carbon fibers - this way lies cheap access to space, and much lighter laptops, cars, industrial machinery, ships... This would be good.

Ironically, *none* of these improvements are in industries that I touch directly. Interesting things are happening in computing, but we've basically got the core communications and computing infrastructure we need to do a lot of interesting things, and the most compelling problems in front of us are not going to succeed or fail on the back of the next high-power processor or high-density memory core. Although I really want a bunch of the flexible screen / electric paper people to be really successful, because that would *personally* improve my life.

Notes:

[1]: The future is here, it's just not evenly distributed - William Gibson. People are working on many of these projects, which is a very good thing. The next hard part (which is the current hard part for some of them, I suspect) will be making them sufficiently inexpensive/accessible/available that they can become broadly distributed, and finding business models that can work well enough that it's possible for people to make money distributing them.
[2]: A standard caveat across all of these: When I say 'I'd like high-density, low-toxicity batteries,' I intend to imply that these are manufacturable in industrially significant volumes, at relatively low cost, using manufacturing methods that are not themselves hopelessly toxic, and that can be improved over time so that they eventually become absolutely low cost as well.

Comments

[xthread.livejournal.com]

You're not actually talking about the problem I asked about - what you're describing is Gentrification, where, for example, the bohemian culture of Northwest Portland is priced out of the market and flees to Hawthorne. The problem I'm trying to look at is the problem that San Francisco and the Bay Area have, where the housing supply is so far below the housing demand that people are building houses 75 miles away, and people are moving to them because that's the closest that they can afford to live. An off-the-cuff answer to that is 'great, remove the zoning and land-use limitations that are preventing more construction,' but that immediately leads to the question 'why will anyone ever build a new structure that is priced at less than half a million dollars a unit in the resulting construction boom?'

I'd like a good answer to the second question. I don't actually mind that Gentrification happens - even though it can suck to be the long term tenant who has to move because their income hasn't kept up with the value of the property that they're renting, I'd rather have that than have the hollowing out of the tax base that happens if you don't let that happen, and I'd rather have the dynamism of a neighborhood that comes from having new tenants move in over the years as different economic endeavors make sense and as the tenants move through the financial stages of their lives. I don't think casting neighborhoods in amber is actually any better than casting people in amber.

[xthread.livejournal.com]

Reclaiming in-use water is definitely higher value, but reclamation/desalination techniques and technologies should be a strict superset of in-use techniques.

Greening cities - there's been a lot of work done on this over the twentieth century. For references, look into the work of Jane Jacobs, the EcoCity Berkeley (http://www.ecocitybuilders.org/) folks, and the Rocky Mountain Institute (http://www.rmi.org). Also, the Permaculture folks have done some work on that subject; src or permiegirl might be able to elaborate. High green density (cf Portland, Berkeley) lets you get higher urban density and more greenspace than current planning strategies. You only have to give up 1) large lawns per lot and 2) rampant height restrictions, and 3) adopt more mixed residential/retail zoning. Forcing people to drive to the corner store is not desirable! (And current land-use policies, that make the corner store economically unviable in many parts of the country are even worse!)

One Econ question I'd like some grad student somewhere to spend some time thinking about: Where does low-price housing come from? Current thinking is 'compel developers to build and offer some minimum number of lower-priced units within each development plan;' I'd like to see some good data showing how successful that thinking is actually being, and evaluating enough alternatives that we have some idea at least of how badly the different options we have so far actually work, relative to each other. And if some of them actually work well, then lets adopt them more broadly and start spending more time talking about the fact that they work. (http://en.wikipedia.org/wiki/Jane_Jacobs) ()

[gomijacogeo.livejournal.com]

Why do you think I keep flirting with biotech?

[gomijacogeo.livejournal.com]

Oh, and on that topic, a new course starts on Wednesday.

Happy dance.

[xthread.livejournal.com]

Pray elaborate - what Interesting Problems do you hope to acheive some personal traction on?

[gomijacogeo.livejournal.com]

Most of my career has been working on enabling technologies in high performance computing so that people with a stronger science background and more patience can do their thing in computational chemistry, bioinformatics, drug design, CFD, weather/climate, things that go boom, things that shall not be named, etc. It's kinda indirect, but it's cool when you come up with a speedup that allows someone to run efficiently on a 128 node cluster when before the scaling was dropping off at 16 or 32 nodes.

I don't know what's going to come of the bioinformatics stuff I'm doing. Right now, I'm just doing it because I get off on learning new stuff. Maybe job N+1 will be working in the guts of one of the many tools out there. It's a crazy time right now. The impossible of a few years ago is now routine. I haven't watched that kind of change since computing/networking 25 years ago. But I expect I'll continue to labor in the background trying to make the revolution a little faster, easier, or a little less buggy.

As for biotech itself, I can easily see it being targeted towards solving many of the things you mentioned - for example, creating more salt tolerant plants, or even better, plants that sequester the salt. Ditto for low water irrigation, desert recovery, better biodiesel crops, heavy metal removal, etc.

[rmd.livejournal.com]

some of these are the things i think about when i feel stalled out and wonder what i'm doing that actually, you know, DOES ANYTHING. i mean, yeah, okay, we help people save money. that's really not that impressive.

[plantgirl.livejournal.com]

If you have not read Water Wars: Drought, Flood, Folly, & the Politics of Thirst, by Diane Raines Ward, I highly highly recommend it. Of the many Ohmygodwaterissuesareaproblem!!! books, it's the best I've read so far. She doesn't have Answers^TM, but she does lay out the problems quite clearly, along with current attempts to solve them. And there may be answers somewhere in that.

[jonwa.livejournal.com]

I so want to land a job helping to create the next nifty green technology.
Or at least working in the industry some way.
Building electric cars would work.
Installing solar panels.

Interestingly solar actually makes some sense in Washington, as the summer days, when we have sun, are so long. We wouldn't generate much power in the winter, but heaps in the summer.

[strspn.livejournal.com]

Hey Jonwa. I love battery electric vehicles, too, but I like Richard's suggestion to look into carbon scrubbing. I bet there are big machines we can build to remove carbon from the atmosphere, but are any of them going to be a net-negative, and how many of them would need to be from wind power first? The U.S. could power herself with wind turbines on 3% of her farmland. There would have to be a lot of storage, and we might need to double grid capacity.

If the lumber shortage has been addressed by sustainable forestry, then I'm sure we can solve this very similar problem, as long as you smart people keep thinking about it.

[jeffpaulsen.livejournal.com]

Desert- and city-related water factoid: I was at Casa Grande in Arizona a couple months ago, looking at the 12th century ruins there, and the tour guide mentioned that since 1890, well depths in the Phoenix area have increased from 20 feet to over 100 feet.

Being a northwest guy, it never occurred to me that you could draw down an aquifer that big, that much, that fast. I'd always thought about retiring down there in 40 years or so, but now that doesn't seem like a responsible choice.

[jeffpaulsen.livejournal.com]

Also: with your high-efficiency / low-toxicity batteries, I'd like to see a safe way to quickly charge an electric car. By "quickly charge" I mean that I can put gasoline in my car for 10 minutes (including hookup and pay time) and not have to do it again for 350 miles, and I'd like to see electric cars that can do that.

I imagine that if it took 40 minutes, you'd get truck-stop like things where you can hook up and get lunch while your car charges. I can imagine electrical connections capable of putting 350 miles worth of electrons into my car, but I'm not sure I want to be in the car when that is going on.

[auros]

Amory Lovins actually argues that if battery tech doesn't do amazing things soon*, distributed H2 production may kick in. Your utility will lease you solar cells for your roof and a box for your garage containing an electrolyzer, fuel cell, and H2 storage tank. You turn sunlight and water into H2 during the day, then put that into your car at night**. Any excess capacity in the solar array during the day can be used to power your house or sent to the grid, as can any excess H2 left over at night after your tank is full. This largely solves the two biggest problem with the Hydrogen Economy concept -- the fact that H2 is really difficult to store for long periods and distribute to many individual sites over long distances.

* ...which it might. Certainly Tesla's betting a lot of capital on that proposition, and we already have the tech to confine toxicity to the production site (where it can be captured) and makes the batteries themselves fully recyclable -- cradle-to-cradle engineering.

** Another fault in the fuel-cell car idea is that H2 has great energy per unit mass compared to batteries, but crummy energy per unit volume. Compressing it costs energy, making the overall results less efficient. There are some speculative improvements for H2 storage, like using metal-hydrides (related to the stuff in NiMH batteries).

[auros]

Incidentally, one side benefit is that the box in your garage can double as your water heater, since the electrolyzer and fuel cell both already have waste heat that can be dumped in the water. :-)

[xthread.livejournal.com]

My private bet, by the way, is that there will be impressive battery improvements and there will be some number of fuel-based vehicles as well, because there are times when you really need to be able to internally combust something - whether those are H2-based or hydrocarbon-based or some other solution is probably going to be dictated by market evolution.

[a-steep-hill.livejournal.com]

I was going to mention Richard Register's work, but you already know about them. I see you're already turned on to permaculture as well. I assume you know about Bill McDonough and Cradle to Cradle.

One good, highly relevant project for the computer geek set would be developing really good, easy to use, low-cost building simulation software. The commonly used free package (eQuest) is quite old; more recent software is hard to use and expensive. I would love to see someone take the free EnergyPlus analysis engine and marry it to Sketchup.

As for carbon fibers: the shortcoming there is not the performance of the fibers, but their cost and, more importantly, their usefulness for mass manufacturing. FiberForge, a spinoff from RMI, is working on this but don't seem to be making alot of headway towards commercial application.