How Google can save itself - sell privacy.

It's been 3 years since the Google-Apple divorce, eight months since Google 1.0 died, and six months since I tried divorcing Google.

Divorcing Google, but planning to go out with something else. That hasn't worked out so well; Apple in particular is exploring new domains of pain.

Meanwhile, the less-facebooky parts of Google+  are improving, even as Twitter enlists with the Sith (don't they know there can only be two?).

So I'm thinking about trying to reconcile with Google - assuming she's still into geeks. How could Google win us/me back?

Google could sell privacy.

Let me explain. In the modern world two populations have privacy. One is poor, lives on cash and checks, and doesn't have a cell phone. The other is Romney-class wealthy. The rest of us are the Transparent Society. We can't buy privacy.

That is, we can't buy it now, but Google could sell it.

Google could sell a yearly G+ privacy subscription for something like $200 year per person or $400 year per family (wild ass guesstimates). For that amount we'd have full control over what we share, and we'd opt out of all advertising and marketing. We'd still be able to opt in to ads if we wanted, and of course there'd be no shield from subpoena. We'd be able to turn on the parts of G+ we want, and disable those we don't want. We might even have the optional use of disposable avatars or identities.

It sounds like a lot of infrastructure to build for a few users, but Google needs to sell into the German and EU market. Their privacy laws are much stricter than America's privacy "suggestions". Google would also like to provide services for the under 13 group, and even in the US that requires enhanced privacy protection. So they have to build this infrastructure anyway.

At a stroke, this would rebuild Google's geek appeal. Most would decide not to pay the price, but there would be no grounds for objections -- because Google's contract with its users would be transparent.

Some of us would pay.

You can do it Google. Save yourself and we'll be happy again.

Is labor lumpish in whitewater times?

Krugman is famously dismissive about claims of structural aspects to underemployment (though years ago he wasn't as sure). DeLong, I think, is less sure.

Krugman points to the uniformity of underemployment. If there were structural causes, wouldn't we see areas of relative strength? It seems a bit much to claim that multiple broad-coverage structural shocks would produce such a homogeneous picture.

Fortunately, I fly under the radar (esp. under Paul's), so I am free to wonder about labor in the post-AI era complicated by the the rise of China and India and the enabling effect of IT on financial fraud. Stories like this catch my attention ...

Fix Law Schools - Atlantic Vincent Rougeau  Mobile

... the jobs and high pay that used to greet new attorneys at large firms are gone, wiped away by innovations such as software that takes seconds to do the document discovery that once occupied junior attorneys for scores of (billable) hours while they learned their profession..

Enhanced search and discovery is only one small piece of the post-AI world, but there's a case to be made that it wiped out large portions of a profession. Brynjolfsson and McAfee expand that case in Race Against the Machine [1], though almost all of their fixes [1] increase economic output rather than addressing the core issue of mass disability. The exception, perhaps deliberately numbered 13 of 19, is easy to miss ...

13. Make it comparatively more attractive to hire a person than to buy more technology through incentives, rather than regulation. This can be done by, among other things, decreasing employer payroll taxes and providing subsidies or tax breaks for employing people who have been out of work for a long time. Taxes on congestion and pollution can more than make up for the reduced labor taxes.

Of course by "pollution ... tax" they mean "Carbon Tax" [1]. The fix here is the same fix that has been applied to provide employment for persons with cognitive disabilities such as low IQ and/or autism. In the modern world disability is a relative term that applies to a larger population.

If our whitewater times continue, we will either go there or go nowhere.

[1] They're popular at the "Singularity University" and their fixes are published in "World Future Society". Outcasts they are. Their fan base probably explains why the can't use the "Carbon" word, WFS/SU people have a weird problem with letter C. 

See also:

• Slavery, technology, and the future of the weak 2/2012. What do you do with a labor surplus?
• DeLong, Mankiw and the Problem of the Weak 10/2006. When Mankiw was almost readable.
• Mass disability and the middle class 9/2011

Health care: We don't want more stuff, we want more years.

Stanford's Chad Jones and Robert Hall tell us health care spending really is different ...

Why Americans want to spend more on health care (Louis Johnston, MinnPost, 7/6/12)

... Income elasticity measures how much more of a good or service a person will buy if their income goes up by 1 percent. For most goods and services this number is less than 1; that is, if income rises then people will buy more of most goods but they will increase their purchases by less than 1 percent. 

Years of life are different. If you have a medical procedure that extends your life, then the first, second, third and however many extra years you receive are all equally valuable. So if your income rises by 1 percent, you will increase your spending on medical care by at least 1 percent, and possibly more.

Jones, along with Robert E. Hall (also of Stanford) embedded this idea in an economic model and found that it does a good job predicting the path of health care expenditures from 1950 to 2000. Further, they show that if this is true, then the share of GDP we devote to health care could easily rise to 30 percent or more over the next 50 years as people choose to spend more on health care to obtain more years of life.

Thinking about the rise in medical spending this way puts health care policy in a different light. People want to live longer, better lives, and they are willing to pay for it. They don’t want more stuff, they want more life...

Life extending [1] health care is an inexhaustible good. That's what simplistic happiness studies, like a pseudo-science [2] article claiming that $75,000 is "enough", usually miss. They implicitly assume, or indirectly measure, good health [3].

Years ago, when health care spending was a mere 12% of GDP (we're about 15% now), my partner, Dr. John H, saw no reason why it wouldn't, and shouldn't rise to a then unthinkable 15% or more. His point was that people like being healthy, and to the extent that health care works, they will want more of it.

Health care that is perceived to be effective is the ultimate growth industry.

That's why this is where we'll end up. We could do much worse.

[1] A shorthand for extending life that we care about, particularly life-years of loved ones. More years of dementia don't count, though significant disability has less impact that many imagine. I assume there's some amount of quality lifespan that would, depending on one's memory, have an income elasticity of less than one. Science fiction writers often put that at somewhere between 300 and 30,000 years.
[2] I read the published study; "Participants answered our questions as part of a larger online survey, in return for points that could be redeemed for prizes." Can you image a less representative population? Needless to say they didn't define what household income meant, yet they turned this into a NYT article.
[3] The Jimmy Johns' insultingly stupid parable of the mexican banker is a particularly egregious example. 

Where would I hide a military AI project?

If I were a somewhat different person, and life played out quite differently, I can imagine being a senior NSA bureaucrat.

I'd read about Google's cat-recognition engine, and the likelihood of a mouse-level AI within the decade, and I'd be thinking that the NSA needs to get there first.

Not because there's an obvious military application, but because there could be a weapon in there somewhere and because someone in China is thinking the same thing.

So I'd add a few hundred million a year to my off-budget budget; just  pocket change really. Then I'd build a data center for my testing and I'd get to a mouse level AI in six years. If I needed to I could pry the secret sauce out of Google's hands; I'm sure there a ways to do that but it's probably not necessary. Google publishes much of its AI research.

I'd build it just to see what it was like, and so I could assess the military potential.

Problem is, modern AI experiments take a lot of power and produce a lot of heat. I wonder how I'd disguise it...

How China might play the Koreas

Even by China's standards, North Korea is a hideous place.

Why does China keep it going then? Is it merely for cheap labor? A way to keep America busy? Fear that a free North Korea would inspire the Chinese people?

Or do China's rulers have another game in mind ...

Top South Korean Aide Steps Down Over Pact With Japan

... It quickly became apparent, however, that the government had underestimated South Koreans’ misgivings about cooperating militarily with Japan. Mr. Lee’s political opponents quickly seized on that disquiet to begin an election-year offensive, accusing Mr. Lee of kowtowing to Washington and, with various civic groups, likening the conservative governing camp to the past Korean “traitors” who secretly cooperated with Japan’s annexation of the Korean Peninsula in 1910...

...  he could not withstand the furor over reports that he played an important role in negotiating the pact with Japan, a country most South Koreans still view with animosity because of its often-brutal colonial rule in the early 20th century and its territorial claim to a set of islets administered by South Korea.

After the Lee government announced the deal last Thursday, accusations flew that the government was “pro-Japanese,” a far worse charge in South Korea than being “pro-North Korean.” ...

Hmm. Sounds like a country that could be turned.

Perhaps the prize in this game is separating South Korea from the West, either as a democracy or as an authoritarian oligarchy. Preferably the latter.

The prize for neutrality, or for alliance, would be unification on South Korean terms.

I wonder if NK's military leaders think this too.  I doubt they like the idea. They should be working hard to keep South Korea allied with the US.

The Standard Model - summarized

In a most excellent overview of the Higgs(es?) news, The Economist manages the best concise summary of the Standard Model that I've read anywhere (emphases mine) ...

The Higgs boson: Gotcha! | The Economist:

... the Standard Model, the best explanation to date for how the universe works—except in the domain of gravity, which is governed by the general theory of relativity. The model comprises 17 particles. Of these, 12 are fermions such as quarks (which coalesce into neutrons and protons in atomic nuclei) and electrons (which whizz around those nuclei). They make up matter. A further four particles, known as gauge bosons, transmit forces and so allow fermions to interact: photons convey electromagnetism, which holds electrons in orbit around atoms; gluons link quarks into protons and neutrons via the strong nuclear force; W and Z bosons carry the weak nuclear force, which is responsible for certain types of radioactive decay. And then there is the Higgs.

The Higgs, though a boson (meaning it has a particular sort of value of a quantum-mechanical property known as spin), is not a gauge boson. Physicists need it not to transmit a force but to give mass to other particles. Two of the 16 others, the photon and the gluon, are massless. But without the Higgs, or something like it, there is no explanation of where the mass of the other particles comes from.

For fermions this is no big deal. The Standard Model’s rules would let mass be ascribed to them without further explanation. But the same trick does not work with bosons. In the absence of a Higgs, the rules of the Standard Model demand that bosons be massless. The W and Z are not. They are very heavy indeed, weighing almost as much as 100 protons. This makes the Higgs the keystone of the Standard Model...

I've read elsewhere that in the absence of the Higgs particles would zip around at the speed of light. Evidently, not so! The problem is rather with the W and Z bosons. That's quite different, but there's something about this summary that feels more authoritative.

I've pasted that text into Notational Velocity/SimpleNote so I have it in my extended memory.

There's more in the article ...

...  the model requires its 20 or so constants to be exactly what they are to an uncomfortable 32 decimal places. Insert different values and the upshot is nonsensical predictions, like phenomena occurring with a likelihood of more than 100%.

... One way to look beyond the Standard Model is to question the Higgs’s status as an elementary particle. According to an idea called technicolour, if it were instead made up of all-new kinds of quark held together by a new interaction, akin to but distinct from the strong force, the need for fine-tuning disappears.

Alternatively, the Higgs can maintain its elementary status, but gain siblings. This is a consequence of an idea called supersymmetry, or susy for short. Just as all the known particles of matter have antimatter versions in the Standard Model, in the world of susy every known boson, including the Higgs, has one or more fermion partners, and every known fermion has one or more associated bosons....

Google's Project Glass - it's not for the young

I've changed my mind about Project Glass. I thought it was proof that Brin's vast wealth had driven him mad, and that Google was doing a high speed version of Microsoft's trajectory.

Now I realize that there is a market.

No, not the models who must, by now, be demanding triple rates to appear in Google's career-ending ads.

No, not even Google's geeks, who must be frantically looking for new employment.

No, the market is old people. Geezers. People like me; or maybe me + 5-10 years.

We don't mind that Google Glass looks stupid -- we're ugly and we know it.

We don't mind that Google Glass makes us look like Borg -- we're already good with artificial hips, knees, lenses, bones, ears and more. Nature is overrated and wears out too soon.

We don't mind wearing glasses, we need them anyway.

We don't mind having something identifying people for us,  recording where we've been and what we've done, selling us things we don't need, and warning us of suspicious strangers and oncoming traffic. We are either going to die or get demented, and the way medicine is going the latter is more likely. We need a bionic brain; an ever present AI keeping us roughly on track and advertising cut-rate colonoscopy.

Google Glass is going to be very big. It just won't be very sexy.

Computing 2012: The End of all Empires

I grew up in a bipolar world.

Yes, the USSR vs. USA, but also the bipolar world of Microsoft and Apple. One was ruthless and ruled by corporate power, the other was a stylish tyranny.

Times changed. The USSR fell apart leaving a Russian mafia state ruled by a mobster, and the USA fell into a spiral of fear, wealth concentration, political corruption, and institutional failure. China grew wealthy, but turned into a fascist state run by oligarchs and mobsters. The EU has Greece and Italy and the second Great Depression. India, Brazil, everyone has problems, nobody is a secure Power. Now we live in a multipolar world.

Weirdly, the same thing has happened to the world of computing (now including phones). Microsoft's slow collapse is this week's Vanity Fair special. Google joined the Sith and all it got was dorkware, a human-free social network, and a profit-free phone. Post-IPO Facebook is rich and frail looking. Dell, HP, Motorola, RIM and Nokia are history.

Ahh, but what about Apple? Isn't Apple going from power to power -- even in the old Mac/Windows wars?

That's how it looks - to the press. Today. But I'm just coming off an epic 1 week fiasco involving OS X Lion and iCloud. It ended with me deciding to keep my primary machine on Snow Leopard and reverting my iPhone to iTunes sync after years of MobileMe sync. I'll try again when Mountain Lion is out.

Yes, few people will run into the problems I have had (arising at least in part from an obscure geeky bug with OS X/Unix vs Windows "line termination"). Many people, however, will run into some problems. My experience shows that many months after Apple's grandiose iCloud launch and insane MobileMe/iCloud migration, they still don't have troubleshooting tools and procedures or, amazingly, any way to delete your iCloud.com data. It's as though they thought they'd get everything right the first time -- perhaps because everyone associated with MobileMe was purged.

That's a hell of a miss for a corporation with billions in the bank and a fifteen year history of bungling online services.

Then there's the Apple ID/FairPlay/iCloud problems. My friends are struggling with these. Other friends can't figure out how to manage Ringtones on iTunes.

Perhaps most worrisome of all, Apple is providing mega-compensation packages to its corporate executives because, apparently, they must be retained. An unavoidable step with inevitable consequences. Bad consequences.

Apple doesn't look strong to me. It looks vulnerable.

Google, Apple, Microsoft, Facebook. None of them are serving me well. None of them are looking all that strong.

All the Empires are falling. My personal balancing act is becoming more complex all the time.

Traveling in Europe with an AT&T unlocked iPhone 3GS

Peter, a friend of mine, recently bought a used A&T 3GS from another friend. Since the phone's contract had passed AT&T approved a request for iTunes carrier-lock removal. So it was no longer a carrier-locked phone.

Peter took the phone to Europe and did very well with it there. I've not read many success stories so I'll pass on what he emailed...

Traveling to Europe the past several years I've taken along a very basic 2007 era unlocked Nokia GSM cell phone. It was always very easy to find small mobile phone stores. I would just walk in and buy a SIM card for that country for about €10 or so. It usually came with some calling time credit (easy to buy more credit at phone stores and some local shops), and the young guys staffing the stores were always helpful with setup assistance.

Buying an in-country SIM card allowed cheap calls and texts within in that country, and more expensive calls out of the country. Keeping in touch with friends and family back home can also be done very cheaply, if calls are made from the US to the European network mobile phone by Google phone from a computer. These calls from the US to a European mobile phone will cost about 10-15 cents per minute (calls from a computer in Europe using Google phone to a US land line cost less than 10 cents per minute).

[Peter then acquired the iPhone ...]

Setup of this clean iPhone with factory settings was easier than expected. I didn't have a SIM card in the US, but the setup was very simple using my home wifi network and following on-screen prompts. During the wifi setup I easily register the phone under my name using my Apple ID number, and synced my Gmail contact data (since I had previously added phone numbers to my Gmail contacts, I now had both email and phone contact data on the iPhone). I then hooked the phone up to my MacBook (OSX 10.5.8) and synced with iTunes. I enabled iCloud for contacts, email and Photo Stream and then bought a few apps. Very easy and intuitive.

In Rome a few days later near my apartment I found a small neighborhood mobile store for a major Italian carrier (TIM). As usual the techy guy working there spoke English. The process was simple - I showed him the iPhone (though he was a little surprised that I had an un-locked US iPhone), and asked to buy a SIM card (€20 with €15 voice and SMS credit plus free unlimited data for 1 month). He then had to register it with my passport info and do a little setup. The whole process took about 30 minutes, and with a little lag in activation time I was up and running with iPhone voice and data service in Rome within 1 hour.

Note: iPhone 3Gs uses a regular sized SIM card, iPhone 4 and 4s use micro SIMs, which may be harder to find in Europe (though it is possible to "cutdown" a regular SIM to micro size). Also you can buy a new basic unlocked GSM phone in Europe for about €30.

During my 2 week stay in Rome, the iPhone was very helpful - calling local friends, restaurant reservations, email, camera, using GPS and maps for locating sites and restaurants, and finding walking routes between sites... Especially helpful purchased apps included: Italian dictionary and verbs, Rome2Go, Rome Travel Guide - Lonely Planet (linked to GPS maps this was incredible useful), Contact Sync for Google Gmail, TripColor and Kayak Pro.

Moral of this story: When traveling in Europe, an un-locked iPhone is a fantastic asset. Next best would be any unlocked GSM smart phone or basic phone.

I'm amazed that a 20 euro card came with unlimited data!

Air Conditioning and obesity

I ate lunch at my favorite diner. Great food, but too much.

Sometimes I put enough aside; but not today. I ate the whole thing.

After lunch I left the cool hole-in-the-wall and walked into the great heat wave of 2012 (which will seem benign in 2022). I walked slowly back to the office, trying not to touch myself.

I blame it on the air conditioning. If China Restaurant hadn't been pleasantly cool, I know I'd have stopped sooner. 

Of course there would be other consequences of a world without air conditioning. I'm relatively slender, and I feel the heat. If I were heavier, the heat would be even more uncomfortable. Another incentive to weigh less.

So is modern air conditioning a factor in our losing battle with fat?

The question has been asked ... 

International Journal of Obesity - Putative contributors to the secular increase in obesity: exploring the roads less traveled June 2006

...The thermoneutral zone (TNZ) is the range of ambient temperature in which energy expenditure is not required for homeothermy. Exposure to ambient temperatures above or below the TNZ increases energy expenditure, which all other things being equal, decreases energy stores (i.e., fat). This effect was shown in short-term controlled human experiments 41, 42 and the decreases in adiposity were evident in controlled animal experiments; these effects are widely exploited in livestock husbandry, where selecting the environment to maximize weight gain is critical.43 Animal44 and human45 studies show that excursions above the TNZ markedly reduce food intake. Herman45 cited a consumer survey suggesting that after an air-conditioning breakdown, restaurant sales drop dramatically...

Suggestive data, but the relationship to "TNZ" sounds dubious. The authors should have described the findings as interesting correlations with some evidence of causation, then mentioned TNZ as one mechanism among many.

I couldn't find any more recent references.

Bicycle light donation - a selfish way to give

Today's commute was as good as it gets. Warm, sunny, not too hot yet. Bit of a breeze. Streets quiet. Cars considerate and happy. A good ride for daydreams of doing good stuff.

Until one of my daydreams runs into something practical. Something like a program to donate towards $10 easy to install semi-sealed high reliability blinking red bike lights.

I could bring $200 to an interested bike shop, put up a small poster, and see how it goes.

It's the kind of low energy donation I might actually do, and, best of all, it's anonymous. So no spam!

Even better, it's selfish. I worry a bit about getting run over, but I worry a lot more about running someone else over. More blinkies, fewer nightmares. A program like this ought to appeal to people who never bike, particularly elderly drivers with good reason to fear low visibility bicycles. Unlike helmets, which some cyclists dislike (not me), just about everyone likes blinkies [1].

My next thought was that someone has to have set something up like this. Of course they have, including in Minneapolis four days ago ...

• I’m raising money so more people on bikes can Get Lit and Get Lit: This Chicago program focuses on front white lights.
• Be Bright at Night: A Cleveland campaign that gives lights to people who bike without at night. It's supported by the Sylvia Bingham fund, in memory a young woman killed by a bad right turn. (Lights wouldn't have helped her, to prevent those turns we need smart-car tech [2]).
• Portland (OR) police distribute bicycle lights
• Minneapolis police install bike lights (four friggin' days ago) - part of the Bike Cops for Kids program, "Children's Hospitals and Clinics of Minnesota and other generous persons have donated 1,500 helmets, 150 bikes and 6,00 bike lights."

These are great programs, but I didn't see one that involved distribution at local bicycle shops like Cycles for Change, Express Bike Shop, or my local favorites. So there's room for growth.

Anyone know of other programs like this?

[1] Except for the Apocalypse flavor of libertarian. In my searches I found someone complaining about the corrupting influence of bicycle light donations.
[2] It's in the lab. Machine vision/radar to identify pedestrians, animals, and bicycles and alert drivers.

After the decision - healthcare 2012

America's ObamaCare monster lurches forward.

It is hideous. Bits and pieces have fallen off, more will fall. It lost an elbow with the unexpected medicaid ruling.

Still, it moves. Lurches become steps. Trillions of dollars will build momentum. Plans made are being executed.

There will be ways to incent states to extend medicaid coverage. I am sure of that.

Some predict renewed GOP vigor and an enraged Tea Party.

I don't think so. I think that the American infotainment industry is now going to start talking about what the mandate really means. That mountain will become a molehill.

More importantly, the healthcare industry is going to pivot to making this concrete. Corporations hate uncertainty, and they hate reversing course. Romney will listen to them.

I would not be surprised if the ObamaCare hate dies as quickly as the 'defense of marriage' passion.

The monster is ours. We can rebuild him ...

Browsing the blog backlist - we need a new app

I follow about 150-250 active blogs including the "Core" blogs I read religiously.

It's a good reading list, far superior to the days when I read The Economist (RIP 2006, but a bit better lately) and the NYT. Still, there's something wrong.

The wrong bit is that there's rich material buried in the backlog of many of those blogs, and in blogs that no longer publish but are still online. Obviously some of that material is better than what I'm reading now.

We need a tool to surface that material and make it available. A blog backlist browser tool.

I'd like, for example, to give Reeder.app a set of 20 or so blogs and have a convenient way to read posts from years back. I've tried doing this in the "new" Google Reader, and it really doesn't work. 

Has anyone heard of anything like this?

Google's A.I. recognizes cats. Laugh while you can.

Google's brain module was trained on YouTube stills. From vast amounts of data, one image spontaneously emerged ...

Using large-scale brain simulations for machine learning and A.I. | Official Google Blog 

".. we developed a distributed computing infrastructure for training large-scale neural networks. Then, we took an artificial neural network and spread the computation across 16,000 of our CPU cores (in our data centers), and trained models with more than 1 billion connections.  

...  to our amusement, one of our artificial neurons learned to respond strongly to pictures of... cats ... it “discovered” what a cat looked like by itself from only unlabeled YouTube stills. That’s what we mean by self-taught learning... 

... Using this large-scale neural network, we also significantly improved the state of the art on a standard image classification test—in fact, we saw a 70 percent relative improvement in accuracy. We achieved that by taking advantage of the vast amounts of unlabeled data available on the web, and using it to augment a much more limited set of labeled data. This is something we’re really focused on—how to develop machine learning systems that scale well, so that we can take advantage of vast sets of unlabeled training data.... 

... working on scaling our systems to train even larger models. To give you a sense of what we mean by “larger”—while there’s no accepted way to compare artificial neural networks to biological brains, as a very rough comparison an adult human brain has around 100 trillion connections.... 

..  working with other groups within Google on applying this artificial neural network approach to other areas such as speech recognition and natural language modeling."

Hah, hah, a cat. That's so funny. Unless you're a mouse of course.

The mouse cortex has 14 million neurons and a maximum of 45K connections per neuron, so ballpark estimate, perhaps 300 billion connections (real estimates are probably known from the mouse connectome project but I couldn't find them). So in this first pass Google has less than 1% of a mouse connectome.

Assuming they double the connectome every two years, they should hit mouse scale in nine years, or around 2021. There's a good chance you and will still be around then.

I've long felt that once we had a "mouse-equivalent" connectome we could probably stop worrying about global warming, social security, meteor impacts, cheap bioweapons, and the Yellowstone super volcano.

Really, we're just mice writ large. That cat is looking hungry.

Incidentally, Google didn't use the politically incorrect two letter acronym in the blog post, but they put it, with periods (?), in the post title.

Best cities of North America - Chicago, New York, Montreal ... and Minneapolis

A report on bike share safety in NYC makes a number of safety recommendations and includes a graph of North America's best cities:


We see famous city names like Toronto, Chicago, San Francisco, Vancouver.

And then, way out on the right site of the best of the best ... Minneapolis (and, damn you, Portland to the right of us).

We rock.

Incidentally, the curve shows that as more people cycle the risk of death per cyclist falls (safety in numbers). Vancouver is everyone's target; Minneapolis and Portland need to study Vancouver's example.

My hunch for Minneapolis (and especially Saint Paul) is that the best way to reduce bicycle fatalities here is to enforce our neglected crosswalk law. Since that law primarily protect pedestrians that sounds a bit odd, but I think of this as falling under the 'broken windows' theory of bicycle and pedestrian safety. MSP drivers are gross violators of the crosswalk law and there's zero police enforcement. It's trivial to setup law enforcement sting operations, and it would make all drivers more conscious of their urban surroundings.