omenti

manpo-kei, gold plated teaspoons and home runs

A friend had been struggling to find regular physical activity for a few years.  She complained enough about a lack of motivation that her brother gave her a low-end Fitbit for Christmas.  Somehow the pedometer function clicked for her.  Being able to see the numbers of steps was as motivating for her as filling in circles on an Apple Watch is for other people.  On top of this she'd heard about the magic 10,000 step number.  Initially the goal was beyond her, but she was motivated and within a couple of weeks she was there.  She's kept it up and regularly does 15,000 a day along with increasing her speed.  Her blood pressure is down from last year's number and she feels healthier.

 

In the 60s a Japanese company marketed a pedometer known as the Manpo-kei - which translates to the ten thousand step meter.  The pedometer became a fad and somehow the 10,000 step goal was enshrined in popular culture.  There's no reason why the number means anything other than being something like five miles of walking for most people.  That's over an hour of walking and good exercise. So people who aim for the number are getting a good benefit because the general direction of walking a lot is healthy rather than the number being the "right" target. 

 

The NY Times recently ran an article on some inaccuracies in smart watches.  It doesn't surprise me.  I've been around Olympic programs and athletes for a few years and know some horror stories about the over-reliance of data when you don't know it's accuracy or the context of the measurement or its use.  Like anything else with data, you need to know why, what and how you're measuring, manipulating, and finally interpreting and using the information.  In sport that turns out to be difficult.  There are a number of interesting approaches to deal with the issues, but that's another post. 

 

The NY Times piece and elite training experience suggest broad trends over time are much more important than local accuracy and precision  If you're trying to improve your fitness level or play amateur level sports, the motivation from a smartwatch can make a big difference.  If you're an elite athlete there are many other things to consider and it's likely you're taking advantage of them.

 

The notion of broad trend over local accuracy has utility in many areas.  One that frustrates me is building  carbon dioxide removal technology.  It's something of a delaying tactic being cheerleaded by the petrochemical and coal industries.  In reality it's extremely expensive, requires a large amount of energy and diverts funding from much more effective technologies and behaviors.  Currently four 1 million ton  class plants are on the drawing board.  With luck each could remove a million tons of carbon from the atmosphere by 2030 or so with a projected cost of about a billion dollars each.  So how much of an impact is that?

 

A million tons of carbon dioxide is about 1/40,000 of current yearly emissions.  That's roughly the same ratio as a teaspoon to a bathtub full of water.  Imagine running the tub and in the time it takes to fill it, you can remove a teaspoon of water.  The bathtub keeps running..  in the time another tub of water comes out of the spigot, you can remove a second teaspoon.  And then a third, forth and so on.  

 

We currently have many deployable technologies and behaviors that will turn the tap down much more than a teaspoon worth for far less money.  When we're finally down to the point where we're only putting a gallon or so of water into the tub each year, then it might make sense to deploy a number of teaspoons to make a difference for where it's difficult to turn the tap (aviation and shipping for example) 

 

Spend some money and talent learning how to improve the process, but don't count on it in the near term as it isn't, and will never be,  a silver bullet.  Nothing it.  all we can hope for is silver buckshot and using as much as we can possibly afford.  It makes sense to use the buckshot with the greatest cost/benefit ratio - things like wind, solar, better power grids, efficient transportation (full sized EVs don't count!), and any number of energy thrifty behaviors. 

 

And finally home runs and climate change.  Recently a paper made the major news outlets with projections that climate change will have an impact on Major League Baseball home runs: "Several hundred additional home runs per season are projected due to future warming."  The paper is poorly done .. basically a sensitivity analysis based on a single variable. I won't give it the credibility of a link.  Something that doesn't work in the real world with something as complex as baseball.  I have a some expertise in the fluid dynamics of balls flying through the air and would argue the effect is small.  In fact they note the same thing, but word the conclusion to make it appear there is an effect and climate change impacts everything.  This is pure clickbait for news organizations and probably has the impact of diverting attention from much more serious issues associated with global warming.  It's trivial to see the home run impact is insignificant, but a poorly done analysis can lead to fuzzy thinking and even harm.

 

revisiting eliza the next generation

a minipost

With all the talk about large language models and applications like ChatGPT,  I recalled a post from about two and a half years ago:  eliza the next generation.  I'd say much the same now - some of the predictions have come about.  Of course it's only been a short while and I played with prototypes, so there's that:-)

 

It was high pandemic and I was spending quite a bit of time talking with friends who were working on interesting things.  Among other things I started to become aware of progress in large language models.  While many who work in AI (a poor label, but you get the idea without me having to talk about a variety of fields and subfields as well as what's unrelated), believed them a diversion, there was progress in LLMs and some interesting applications were emerging.  We chatted about where it could go and what techno-social issues might arise. (disclaimer: I'm not a techno-solutionist)  About the time I posted, Emily Bender at the University of Washington wrote a terrific paper on issues with natural language understanding and if we can believe what comes out of LLMs.  Recently an article on her work appeared in The Intelligencer.  It's a necessary read if only to understand her octopus analogy. 

 

This morning Gregg Vesonder and I exchanged some text messages. He's a serious AI guy with decades of experience and is someone who thinks deeply about social technical issues.  One of his comments:

We’ve seen this scene before, blinded by the hype. My concern this time is that explicit or inadvertent bias in the data can mask some nasty behavior that can subtly or not so subtly mask really tragic manipulations.

Aka speech acts. 

I hadn't thought about speech acts much before.  The definition and examples given in the link are excellent tools for thinking about these things. 

 

I'm not against LLMs and there are some useful applications beyond writing horoscopes in white bro-speak, but the dimensionality of the space for abuse strikes me as large.  As Suw Charman-Anderson notes, we're already seeing it in publishing. 

 

'AI' is a very large and diffuse collection.  Some areas are useful and potentially revolutionary.  Others are potentially dangerous.  We really need to think about these things rather than just trying them out and seeing what breaks.  The things that break may be far too valuable. 

 

 

peak oil - really?

Once a year I read BP's Energy Outlook.   They have a record of serious greenwashing, so anything from them (or the other fossil fuel companies) requires a bit of skepticism, but this year's report took a turn from the normal path. They suggest global oil demand peaked in 2019.  There have been any number of 'peak oil' predictions, but most refer to extractable supply - a bit more on that later.  While the BP predictions are more scenarios than rigorous predictions, it's significant that a large oil company believes oil demand is, and will continue to, drop.  

 

BP outlines three scenarios: Accelerated, Net Zero and New Momentum.  I'm guessing they'd prefer the New Momentum scenario, and will probably lobby and use other tactics to move in that direction, but the bottom line is they see less oil coming from the ground.

 

I think they're transportation modeling is too conservative. Electrification will probably take place much more rapidly than predicted.  Granted there's a lot of inertia in fleet replacement given how long cars last, but the assumptions strike me as too  slow.  They don't seem to be taking into account the likely unbundling of transportation. Horace Dediu has been a big proponent of unbundling and points out most trips are local and short. [side note: you really need to follow Horace as well as well as David Levinson if you have any interest in the future of ground transportation] It makes no economic or environment sense to use a two ton car, gas or electric, to make a three mile trip. While much of North America has saddled itself with legacy transportation infrastructure and little imagination of what is possible, change is coming to Europe and it seems likely the developing world will favor very inexpensive electric micromobility solutions. 

 

Back to peak oil. M. King Hubbert was a geophysicist with the Shell Oil Company in the 50s when he started looking at long term oil production predictions.  All tended to make use of variables that weren't well pinned down.  Hubbert's model was simple assuming once a discovery was made, production would increase exponentially as more resources and efficiencies are brought in. At some point a peak is reached and an exponential decline begins. The model can apply to any region, or groups of regions,  assuming all of the discoveries are taken into account.  It also assumes the impact of new technologies occurs at a constant rate.

 

On the extraction side the so-called Hubbert curve has accurately modeled many oil fields and regions - notably the US through about 2005.¹  When dramatic new technologies appear - like practical fracking, shale and tar sands production, and  computer guided drilling - you effectively add new sources.   It's an interesting resource mining curve that's been used in many areas outside of oil production.  Hopefully we'll move to a point where most extractable oil is left in the ground and oil exploration becomes a thing of the past. 

 

__________

¹ It may look like a Gaussian or Bell curve, but it doesn't die off as quickly.  It takes the form

 

 

fusion!

The short version is it’s a remarkable technical and scientific achievement. Controlled human-made power from fission is easy - last week marked the 80th anniversary of the first reactor. Fusion - what powers the stars - is much more difficult on Earth. The necessary temperatures are far greater than those found in most stars. A star like the Sun gets away with a very slow and low power density reaction that would be impractical for generating power on Earth. It turns out the power density of the core of the Sun is about a quarter of your metabolic density and close to that of a lizard. To get around this different fuels are needed along with much higher temperatures. Still, if you can only figure it out, it sounds like the perfect goal - fuel from water and no radioactive waste.

 

The first serious attempt at magnetic confinement look place about 60 years ago .. it, along with every other effect until this last week failed. Those that achieved fusion required more input power than the power they produced. The figure of merit is Q: the ratio of the power of the fusion reaction divided by the power required to ignite and sustain it.

 

The laser confinement experiment announced focused a short pulse of light - about a billionth of a second - from 192 lasers onto a small gold container that vaporized to generate x-rays that collapsed a diamond coated fuel pellet of deuterium and tritium. The power of the laser light was 2.05 MJ (megajoules — a gallon of gasoline is 121 MJ, a hot dog in a bun is about 1.5 MJ) and the fusion explosion yielded 3.15 MJ. A Q of about 1.5. What isn’t mentioned is the efficiency of the lasers or the facility isn’t included - or the efficiency of extraction power from the miniblast. The lasers are less than one percent efficient so you can see there’s a long way to go.

 

This facility has been running for about a dozen years. It has been solving enormous engineering and applied science problems along the way as well as contributing to the pure science of understanding plasmas at this scale and temperature. Even the fuel pellets are very difficult to make and test. They have to be extraordinary symmetrical and most are rejected in testing - usable pellets are very expensive.  What has been achieved is not a prototype, but rather a triumph of instrumentation technology along with the control of some difficult to manage parameters. 

 

Moving this line of fusion forward seems like a long shot. A factor of five improvement may be possible with a redesign, but they need more than a thousand. Not impossible, but there’s a long path ahead.

 

There are other design types that confine microwave heated plasmas. Getting it right and working will take many years - probably decades and then there’s the issue of what does a facility cost compared to other types of power stations.

 

I hope work goes forward, but not at the cost of workable technologies that address global warming *now*. There’s so much that can be done, but very little political will.

 

28 years!

This is the original holiday card I sketched back in 1994 along with a bit of midi. I borrowed a Wacom tablet and made a few strokes with the pen - a minute's worth of work.  The music took longer.  The web was new so I posted it on a server under my desk at work.  I didn’t have enough time or imagination to make some new the next year so it returned unchanged.  The third year rolled about and I would have forgotten about it, but there were a few dozen requests.  From then on it went up every year on or soon after Thanksgiving in honor of the Winter Solstice holidays.  It’s a survivor having been hosted on five different servers over the years.  For the silver anniversary I though about drawing something new and held a vote.  The old card stays - it was unanimous.

Now it’s 28 years old.

the original holiday card

Autoplay didn’t exist then so you’ll have to start the music yourself.

Happy Holidays! - peace, friendship, and health to all.

Steve

 

 

chindōgu flavored technologies

Chindōgu is the curious Japanese art form of inventing gadgets that seem to a clever solutions to  problems, but which end up causing more problems than they solve.  You may have seen examples like:

toilet paper dispenser hats: you never run out of tissue when you have to sneeze

baby mops: a baby outfit that also mops the floor

umbrella ties

many more .. search the Internet

 

To keep the art form pure you're not supposed to make money - almost guaranteed as the public will see the gadget as useless and won't buy it.  (there was a tragic mistake - the selfie stick started as chindōgu)  I know of an engineering professor who taught a course on creative engineering using it as a vehicle for student projects.

 

For a few years I've found it useful to extend the term to cover technical and social technical inventions that are seen as solving problems by the maker, but are clearly clueless.  A few examples from the past month:

Flat panel displays in cars are a bit of automotive chindōgu.

Blockchains are an example of computational chindōgu.

Meta's Horizon Worlds is pure techno-social chindōgu.

 

Perhaps you might find hunting for this kind of chindōgu useful.  I find it to be a nice way of thinking about what might represent meaningful and useful longterm change.  And of course sometimes you label things incorrectly and have to come back and rethink.  That may be even more useful.

 

And then there are potlatches..  the destruction of wealth to impress and gain status.  Fully autonomous (class five) vehicles, NFTs, Facebook's version of the meta verse, hydrogen powered cars ...  

 

 

 

 

 

where nobody is average

The first military airplanes designed after WWII began to arrive in the late 1940s.  A number of these early jets were more complicated to fly and the number of incidents and accidents alarmed the military.  Incidents where something went wrong, but the pilot and aircraft came back safely, and accidents were more serious.  As the jets were mechanically more reliable than their piston powered WWII era cousins, it only seemed natural to blame pilot error.

 

A  few remarkable pilots would find themselves in a bad spot, but always manage to get themselves out of trouble - often saving the aircraft.  Chuck Yeager was the legendary master  and was sent to airbases to show pilots how to fly correctly.  But the fact was this was happening to very skilled pilots.  With pilot error and aircraft malfunction seen as less likely root causes, eventually cockpit design was questioned.

 

Military airplane cockpit measurements were based on an average of a few measurements of about a hundred airmen in the mid 1920s. Improved nutrition and larger pilots was suspected as changing the average, so a massive study was undertaken that took 140 measurements of nearly 4,100 pilots.  In theory this would lead to a better average cockpit.

 

Gilbert Daniels joined the medical division at Wright Patterson out of college and was one of the people tasked with making these measurements and coming up with averages.  It turned out he was biased.  In college he did some research trying to find an average hand size and found it impossible.  Sifting through the data he looked at ten measurements deemed most important for the cockpit and calculated an average range for each that included 30 percent of the measurements - a nice wide average.  The surprise was than none of the pilots fell within all ten of the wide averages.  Gilbert found if you picked any three of the ten, less than four percent of the pilots would be average.  No one was average.

 

It shouldn't have been a surprise. Women's clothing designers had known it for awhile - something attributed to women being "different."  Eventually the military believed Gilbert's work and mandated highly adjustable cockpits.  All new aircraft had to be adjustable for males in the 5 to 95 percentiles for each of about ten critical dimensions.  Airplane manufactures balked - "it can't be done",  but finally came around because .. well .. contracts.

 

The fifties saw more adjustability come to automotive interiors, but at a much slower and more incomplete pace than military aviation.  About ten years ago I found myself driving a friend's Japanese sports car on the Pacific Coast Highway.  A lovely car considered one of the best handling in the world, but I didn't fit - there was no way to adjust the seat or steering wheel enough.  I'm not particularly tall, but have short legs and a long torso.  I also have very long feet (in theory the right dimensions for a swimmer, but not my thing).  A few months later I was helping a friend move in a rented truck.  The same problem.  There weren't enough adjustments to prevent my feet from getting caught near the pedals.  And then there's automobile safety.  The instrumented dummies used in crash tests cover less than a quarter of adults bringing into question the tests themselves.  

 

And it's not just physical differences.  'What color is the dress?' became a popular question as people legitimately came to different conclusions.  Tonal language native speakers are much more likely than other populations to have "perfect pitch."  Some people are purely visual thinkers, but most are a mixture of audio and visual.  There are people with very poor stereo vision - I know one who is an animator.   Increasingly "on the spectrum" is a meaningless construct.  Our brains have so many differences - so many dimensions - that it's likely none of us are truly "normal."  This has become an issue in fMRI studies recently. For example studies showing differences between men and women are often in the noise and meaningless. The field will eventually get more rigor.  You really need to know what dimension you're measuring.

 

Finally there's neuroplasticity.  We know that kids undergo huge changes in neural connectivity as they grow and some areas of the brain aren't stable until the mid 20s.  (I have a neurological condition that is believed to have resulted from an incomplete severing of neural connections when I was a baby.  It isn't a big thing, but is different enough to be worthy of study).  One seen as stopping around 30, neuroplasticity is seen throughout life, albeit at different rates.  There are examples of dramatic repurposing of regions of the brain.  Helen Santoro's missing left temporal lobe is one of the more dramatic examples.

 

Differences are a feature.

 

 

 

clearing the air

a mini-post

A year ago I managed to help out getting a small grant from the state to build a number of Corsi-Rosenthal boxes with the help of two teachers (science and art) and about twenty high school kids.  It took a Saturday morning to build about eighty of the air cleaners.  All but two are still working and, even though the filters are visually ugly and dirty, they're working just as well.   I won't list all of the learnings - get in touch if you want to know more - but a big benefit besides cutting the viral load is dramatically cutting PM 2.5.  Kids and teachers report fewer allergy and allergy problems and toxic particulate from a nearby highway has been cut.  I suspect these would be great for fires and dusty areas (thinking Arizona and Salt Lake City).   In the past year more clever variants have appeared so check out YouTube if you're curious.  Oh - and filter prices drop if you buy 100 at a time:-)  The only downside was I never want to see duct tape again,

 

The motivation for the post was a fun discussion Andy Slavitt had with Richard Corsi and Jim Rosenthal on his In the Bubble Podcast.   Such a fantastic project for kids and adults!  (there are a number of commercials - Andy donates all profits to charities).

<br.

couplings and friction

a quick minipost

 

Beginning physics students learn about simple coupled systems.  Things like pendulums with a weak spring between them and the like.  Next you look at somewhat more complex systems like a group of pendulums.  Put them on a solid surface and start them in different oscillations and they do their pendulum thing unaffected by their neighbors.  Now add a little communication by putting them on a floor that can transmit a bit of vibration, something interesting happens.

 

The coupling of systems is common in nature covering a very wide range of complexity inducing our brain.  Many of these systems are mathematically fascinating.  Turn up the amount of coupling and they behave as you might expect - at least to a point.  Then, as the comments of the system impose themselves on each other,  there's a dramatic change.  Sometimes these are good from our point of view and sometimes they're disasters.  The point of inflection is often marked by a perturbation - lightning strikes, hurricanes, pandemics, big lies.   The couplings are too strong to allow for anything but a change to a new state.

 

One feature of our society is an increased coupling of systems.  Social media and other targeted messaging, GPS, the Internet of Things, globalization, just in time everything, religion, democracy....  More systems are being connected and the amount of communication is increasing at a terrific rate. These couplings are often introduced to decrease cost and the "friction" parts of the system have with each other.   Many are  too complex to understand well, but the basic math of over coupled systems is clear - these systems can be very unstable. 

 

Building alternate pathways and redundancy into these systems is an answer, but that costs money and clever design.  Another alternative is reducing the coupling of certain important subsystems.  It's something I find myself paying more careful attention to.  Highly coupled or low-friction systems are usually brittle at some point.  I dislike the term "black swan" as real black swans (the bird) are not uncommon in nature.  But we continue to build serious and increasingly effective black swam bait...

 

 

 

31 is the new 35

Andrew Revkin and I had a brief exchange about the power of labels.  Both of us object to the phrase "climate emergency."  Even though it is an emergency that could escalate into an existential threat, people don't respond to labeling longer term threats as emergencies.  I've been suggesting we've entered an adaptation epoch.  Andrew, a much better communicator, is still searching.

 

The past two or three years have taught us short term temperature excursions can be dramatic.  Last year the Pacific Northwest recorded previously unheard of temperatures -  the highest in the area in at least 7,000 years and probably much longer.  This Spring it was South Asia.  Last week record heat came to England and parts of Europe.  These events are probably going to be more common and time goes on and greenhouse gas levels continue increasing. 

 

Wet bulb readings of 35°C (95°F at 100% humidity)  have long been considered unsurvivable.  A person in good physical shape would probably die in a few hours. Such events are extremely rare. Until the past decade they've only been recorded a few times and then only for an hour or so.  They're increasing in frequency and length - mostly in the Indian subcontinent.  Finding cooler spaces for people is a matter of life and death.  

 

It's become clear the 35° number was something of a guess.  Recent careful work shows it's more like 31° for young people in good physical shape.   Older and compromised people have lower limits.  The Tokyo Olympics had periods with wet-bulb temperatures in the upper twenties. Athletic performance during these periods was certainly compromised (I saw it happen to a friend). 

 

What can be done?  Cities suffer from heat island effects.  All of the asphalt and concrete increases the local daytime temperature and holds them at higher levels at night. (hot nights may be harder on the body than hot days).  Air conditioning helps, but it costs money and a power grid that not everyone has access to.  Reducing the amount of asphalt and concrete while planting trees can make a big difference.  It's a great justification for reducing the number of cars and roadways in cities (green spaces are a key motivating reason for the dramatic change currently underway in Paris).  Unfortunately green spaces are usually linked to wealth.  In many places racism and redlining have produced some of the worst heat islands.

 

There are other tricks that are easily implemented.  White paint  on roofs can make a difference. Paint that holds up to temperature and monsoons for the better part of a decade isn't cheap, but is becoming more popular.  In Western countries more reflective shingles are appearing. Some of these look fairly normal, but are highly reflective in the infrared and can make a difference in cooling bills. 

 

And you can send some of the excess heat directly to space.  

 

If you're in a very dry area go outside on a clear night and look up.  Feel the cooling on your face.  Now hold a piece of cardboard between your face and the stars.  You'll notice a warmer feeling.  Remove the cardboard and your face gets cooler. 

 

Your body glows - you've undoubtedly seen infrared photos and videos of people.  We radiate photons at a variety of wavelengths, but it peaks around 9.5 microns - about twenty times longer than visible light.  The  atmosphere is transparent to visible light and also from 8 to 13 microns.  Photons from our bodies can radiate to space.   Deep space is really cold (a bit over 2° above absolute zero) so your face is radiating some of it's heat directly to space - enough that you can feel the cooling.

 

Clever material engineering has produced materials that are good at absorbing the heat around them and then radiating it into space. It's enough to reduce cooling requirements - at least in places with very dry atmospheres.  There's still work lowering costs and making the surfaces last longer, but it's something else to try.

 

Back to the idea of an adaptation epoch. Many things will need to be tried and efforts will stretch for generations.  There's enormous opportunity.  It's both sad and dangerous there's so much 19^th century political inertia, but I'm optimistic more clever people, organizations and governments will finally emerge.  As a species we don't have a choice if we want to survive.