bricks and mortar Amazon

We all know that traditional retail is dying because Amazon can deliver anything to your front door. Now, in a strange irony, Amazon is opening some physical stores.

Last year, Amazon opened its first physical bookstore in the University Village shopping mall in Seattle. The store features thousands of books, a tiny sampling of those on Amazon’s website, most of them with customer ratings of four stars and above.

The books sell for the same price in the store as they do on Amazon’s site. Because book prices regularly change on the site, visitors to the store scan books using a mobile app to find out how much they cost.

Although the store is called Amazon Books, it prominently features a growing array of Amazon-made devices, including the Kindle tablet, the Fire TV set-top device and Echo, its home speaker and virtual assistant.

mosquito bites

Here’s some practical advice from NPR on how to avoid mosquito bites. DEET works, and has been more or less proven safe. So, somewhat surprisingly, does eucalyptus oil.

I am fairly careful about mosquitoes, and yet I have been bitten by them in Florida, Thailand, Singapore, Indonesia and less exotic places on the U.S. eastern seaboard. So far I haven’t caught West Nile, St. Louis, dengue, chikungunya, Japanese encephalitis, and had never even heard of Zika until last month although it is supposedly common in places I have been. Then again, I could have an asymptomatic infection with one or more of these, who knows. If you talk to enough people from tropical countries, the majority have had dengue at some point, usually with a full recovery. But it can cause a particularly cruel and grizzly death in a few unlucky cases.

None of this is to make light of the horrible complications we are seeing with Zika. It’s just that like anything, your best bet is to take reasonable precautions and try to be as rational about the likelihood and consequences of unfamiliar risks as a human can.

containing technologies

This post is about CRISPR and gene drive, which are interesting in their own right. What I am going to quote is the author’s ideas on how to develop a promising but potentially dangerous technology responsibly:

For starters, public notification and broadly inclusive discussions should always precede and inform development of gene drive interventions in the lab. A clear description of the potential impact of an experiment – as my colleagues and I have provided for the technology as a whole – must be followed by transparency throughout the development process. This community-guided approach to research provides opportunities to identify and address potential problems and concerns during development. If a perceived problem cannot be adequately addressed, researchers should be prepared to terminate the project…

Another feature of a responsible approach would be a commitment by scientists to evaluate each proposed gene drive intervention – say, immunizing mice so that they cannot transmit Lyme disease to ticks – individually, rather than making a blanket decision on the technology as a whole. After all, the benefits and risks of each intervention would be entirely different.

A final safeguard against the irresponsible development of gene drive technology is to ensure that early interventions are developed exclusively by governments and nonprofit organizations. Given the potential of financial incentives to skew the design and results of safety tests, keeping the profit motive out of the development and decision-making processes will encourage balanced assessments.

ghost fleet

I’m a sucker for hypothetical future war books. I don’t know why I find them so fun. Obviously I would not find it so fun if this actually happened.

From Amazon:

What will the next global conflict look like? Find out in this ripping, near-futuristic thriller.

The United States, China, and Russia eye each other across a twenty-first century version of the Cold War, which suddenly heats up at sea, on land, in the air, in outer space, and in cyberspace. The fighting involves everything from stealthy robotic–drone strikes to old warships from the navy’s “ghost fleet.” Fighter pilots unleash a Pearl Harbor–style attack; American veterans become low-tech insurgents; teenage hackers battle in digital playgrounds; Silicon Valley billionaires mobilize for cyber-war; and a serial killer carries out her own vendetta. Ultimately, victory will depend on blending the lessons of the past with the weapons of the future.

Ghost Fleet is a page-turning speculative thriller in the spirit of The Hunt for Red October. The debut novel by two leading experts on the cutting edge of national security, it is unique in that every trend and technology featured in the novel — no matter how sci-fi it may seem — is real, or could be soon.

The gold standard, for me, will always be Clancy’s 1986 Red Storm Rising, which was about a hypothetical U.S.-Soviet Union War. He tried to pull an encore of sorts in 2001 with The Bear and the Dragon, but it just wasn’t that great. A similar hypothetical U.S.-China war novel is 1999’s Dragon Strike, by Humphrey Hawksley, which was a little better than the Clancy version even though Clancy invented the genre (and you wonder if Clancy read Dragon Strike before he published his novel, or maybe had already written the novel and was annoyed someone beat him to the punch with similar subject matter).

One more future war novel I found interesting and thought provoking was Deep Sound Channel by Joe Buff. In that one, yet another German-led axis of evil arises. The novel focuses on the hypothetical use of nuclear weapons in fairly limited and tactical ways in naval and submarine warfare.

Maybe I like these books for the chance to put my petty everyday concerns and irritations in perspective.

dots moving around on a map

This is just dots moving around on a map, but I find these dots very engaging in helping me understand urban planning concepts and results of a simulation.

I found this on R bloggers, which talks about how the simulation and map were created.

Data Scientist Todd Schneider has followed-up on his tour-de-force analysis of Taxi Rides in NYC with a similar analysis of the Citi Bike data. Check out the wonderful animation of bike rides on September 16 below. While the Citi Bike data doesn’t include actual trajectories (just the pick-up and drop-off locations), Todd has “interpolated” these points using Google Maps biking directions. Though these may not match actual routes (and gives extra weight to roads with bike lanes), it’s nonetheless an elegant visualization of bike commuter patterns in the city.

tree canopy volume

I had never thought about modeling tree canopy volume in 3D before. I’ve played around with simple algorithms to place trees on a map, assume a mature canopy area per tree, and estimate the total canopy area. This is useful because cities sometimes set targets and metrics in terms of number of trees, and sometimes in terms of tree canopy. The latter is better because it is more relatable to other goals a city might have related to the hydrologic cycle, carbon, heat, air quality, aesthetics and property values, biodiversity and habitat, and the financial cost to public offers of achieving these goals. Once you have an algorithm relating number of trees to canopy area, you can add more variables like type of tree, growth over time, and some assumed attrition rate or half life. Come to think of it, I have played around with leaf area index which is a quasi-3D concept. Anyway, without further ado here is the article that prompted my line of thought:

Local Impact of Tree Volume on Nocturnal Urban Heat Island: A Case Study in Amsterdam

The aim of this research is to quantify the local impacts of tree volumes on the nocturnal urban heat island intensity (UHI). Volume of each individual tree is estimated through a 3D tree model dataset derived from LIDAR data and modelled with geospatial technology. Air temperature is measured on 103 different locations of the city on a relatively warm summer night. We tested an empirical model, using multi-linear regression analysis, to explain the contribution of tree volume to UHI while also taking into account urbanization degree and sky view factor at each location. We also explored the scale effect by testing variant radii for the aggregated tree volume to uncover the highest impact on UHI. The results of this study indicate that, in our case study area, tree volume has the highest impact on UHI within 40 meters and that a one degree temperature reduction is predicted for an increase of 60,000 m3 tree canopy volume in this 40 meter buffer. In addition, we present how geospatial technology is used in automating data extraction procedures to enable scalability (data availability for large extents) for efficient analysis of the UHI relation with urban elements.

swing the election

Here’s an interesting interactive tool on FiveThirtyEight.com where you can play around with U.S. voter turnout and preferences among various demographic groups.

I ran a few scenarios:

  • The default scenario is that each demographic group (educated white, uneducated white, black, hispanic/latino, and Asian) votes for the same party in the same proportions as 2012, and turns out at the same rate, but the absolute size of each group is adjusted for changes between 2012 and 2016.
    • electoral votes 332-206 in favor of DEMOCRATS
  • Let’s go back to the default, and all the Asian people stay home.
    • 332-206 in favor of DEMOCRATS (just not enough people, and maybe already concentrated in democratic states)
  • Back to the default, and all the hispanic/latino people stay home.
    • 283-255 in favor of DEMOCRATS (perhaps hispanics/latinos are also concentrated in already democratic states?)
  • Back to the default, and black turnout falls from 66% to 29%
    • 286-252 in favor of REPUBLICANS (perhaps this flips some key midwest swing states like Pennsylvania, Ohio, Michigan, Wisconsin, etc.)
  • Back to the default, and uneducated whites swing strongly to the right, from 62% last time to 69% Republican (maybe a terrorist attack? a major incident with China or Russia? I don’t want to say false flag, this is not one of those conspiracy websites…)
    • 282-256 in favor of REPUBLICANS (probably those swing states again)
  • Stay with the previous scenario, but educated whites swing ever so slightly to the left, from 56% Republican last time to 54% Republican (what would cause this? I don’t know, some crazy right-wing candidate spouting racist nonsense maybe, I’m not naming names…)
    • 275-263 in favor of DEMOCRATS

So the bottom line is that the minority groups tend to vote Democrat.The uneducated whites tend to vote Republican. The educated whites are the swing voters who end up being the deciding factor. So it is hard to see how a Republican candidate who appeals strongly to uneducated whites but alienates educated whites could ever stand much of a chance.

on the issues

Ontheissues.org is a little bit junky but it has a lot of information on where the candidates stand, well, on the issues. It then graphs them on an interesting spectrum based on where they stand on government intervention in the social and economic spheres.

Social Questions:  Liberals and libertarians agree in choosing the less-government answers, while conservatives and populists agree in choosing the more-restrictive answers.

Economic Questions:  Conservatives and libertarians agree in choosing the less-government answers, while liberals and populists agree in choosing the more-restrictive answers.