20130617

DVD by mail isn't dead yet

Discovery of the day: Netflix has made its original series "House of Cards" available on DVD. For a while, Netflix was in the paradoxical situation of offering almost all of the world's video content on disc except for its self-produced shows. NFLX has not been shy about its strategy to shift attention away from DVDs toward streaming, and I was wondering if this emphasis would include omission of its original series from the legacy platform. It looks like streaming will be the preferred service for premieres, with DVD release approximately 5 months after initial availability. This is similar to the standard DVD "window" for other shows and movies, so Netflix is acting like the other content owners. As a DVD subscriber, I'm glad to see that Netflix is not using their content as an exclusive wedge to further disadvantage the older service.

The logical question, though, is whether I will be able to wait 5 months to see season 4 of Arrested Development or will break down and buy a month of streaming just for one show.

20130614

Convenience (not savings) is the trojan horse for "smart" appliances

There is an alternate reality somewhere in which customers pay the constantly-changing spot price for electricity, smart appliances adjust load to maximize savings under varying realtime conditions, and legions of economists dance the happiest dance that economists are permitted. In our reality, regulators and customers are profoundly ambivalent about even simple pricing experiments like time of use metering, and the use case for consumer load shifting is mostly limited to electric vehicle charging. While stubbornly flat electricity rate plans will disadvantage the demand for smart appliances for the foreseeable future, at least one barrier fell recently: standardization.


Marvell's SEP 2.0-compliant thermostat

With the release of Smart Energy Profile 2.0, there is now a common language for manufacturers to provide functionality for startup, shutdown, monitoring, standby, and other functions. While this is valuable for utility planners, it's harder to make the case that consumers should care. The more affluent consumers who will be able to afford the first generation of smart appliances aren't going to be swayed by a $10 bill credit. They will, however, enjoy the ability to power up their vacation home from their smartphone as they drive so that it is warm and lit-up upon arrival.

An open standard will enable a rich and interoperable ecosystem of third party applications to talk to any device from any manufacturer. Bluetooth's "remote control profile" (AVRCP) enables users to pair a Bose speaker with an Apple iPod. Similarly, the stalled and heavily-siloed home automation controller market could find new opportunities in coordinating a a Nest thermostat and Phillips Hue lights from one place. Getting manufacturers to implement the standard may take some time, but convenience and lifestyle applications will provide more near-term possibilities than the utility's ability to change your air conditioner setpoints on a hot day.

Is this a way out of the chicken-and-egg logjam of device availability and rate plan introduction? I don't see any others on the horizon.

20130313

Electrification, and new vehicle ownership models

If battery technology continues its slow development, will a different ownership model be the accelerator that electric vehicles need to break through? Our approach to personal transport is dominated by two models, with BMW pointing the way to an exciting hybrid approach:

Own 1900-present
  • Lowest total cost of ownership for heavy users
  • Inflexible - one vehicle for all needs
  • Infrequent large inconvenience - user handles maintenance, inspections
Sharing 2000-present
  • Lowest total cost of ownership for infrequent users
  • Flexible - choose vehicle to suit needs
  • Frequent small inconvenience - car may not be available, walk to pickup location
Hybrid ???
  • Ownership advantages for most frequent use case (commuting)
  • Sharing advantages for infrequent use cases (road trip, furniture hauling)
BMW is offering an internal combustion vehicle to electric i3 owners for those occasional trips that will exceed its 100 mile range. It fits somewhere between traditional ownership model and the pure "mobility as a service" model of car sharing, combining the best aspects of both.

EV advocates have been claiming since the 90s that a 120-mile range is enough to handle the typical American commute and errand schedule between recharges. This analysis neglects the fact that we don't buy our cars to handle the 90% most frequent use cases. If that were true, we'd see entire fleets of single-seat cars with no trunk space. At our most rational, we buy our cars to handle 95-99% of the functionality we need so that even single people can drive friends around, take the occasional road trip, and bring home an umlauted cabinet from IKEA. Our desire for functionality can even be aspirational, with few people ever realistically intending to drive their SUV off-road or push their Mustang's cornering and acceleration limits. It's not enough to just handle the commute, and the technical responses (better batteries, on-board charging, battery swapping) have yet to demonstrate practicality. BMW has figured out a clever service-based model that addresses the other 5-10% of the user's mobility needs.

You don't want a drill, you want a hole. You don't want a car, you want mobility.

This hybrid model is an effective one that could carry us for the next 10-15 years. With the advent of driverless cars, I would expect that the pure service-based model will become more popular. Not piloting one's own car reduces personal investment in that particular metal box. If a shared car can drive itself from the remote garaging location, it removes one of the disadvantages of a sharing model. Driverless cars may make personal car ownership seem pointless. Given the surprising development speed of robot cars and the frustrating state of battery technology, does anyone want to take any bets on which one hits 10% penetration first?

20130127

My Camera's Last Remaining Advantage is a $1.99 Wrist Strap

Two years ago, I predicted that smartphones would replace dedicated cameras for the majority of consumers. I also had the smug notion that caring about aperture and exposure values would forever leave me with a "proper" camera. All it took was a trip to a damp Northern European island to show how wrong I was.

Lighting conditions in Scotland are challenging at best. You are usually dealing with a cloudy sky that gives you the choice of overexposed light grey overhead or an underexposed shadow of your subject matter. High Dynamic Range photography takes multiple exposures and composites them together with a bilateral filter to make the local gradients appear natural, enabling all the detail at each exposure to stand out. Though I had a choice of cameras for most subject matters, time after time I kept reaching for my phone's HDR camera over my high-end point-and-shoot.

HDR, 5MP phonecam Single-exposure, 10MP PowerShot
Sky has blue, foreground a bit neon but properly exposed Sky overexposed, foreground good

The trend will likely continue. The next versions of my camera (the Canon S95-110) have an in-camera HDR feature, but the camera's weak processor necessitates less-advanced image matching that creates ghosting and other artifacts. The matching and quality of the on-board image processing from my year-old smartphone is strong enough to handle moving objects in the foreground and whatever else I throw at it. The processing gap will only grow over time as camera manufacturers have a hard time justifying 8-core processors and other beasts. The science of optics has been stable for the last hundred years or so, but we're just beginning to experiment with superresolution, adaptive lighting control, and light field rendering. The action is in processing, not optics.

At this point, the most notable advantage that my point-and-shoot has over my phone is a wrist strap; I feel more comfortable dangling my well-secured camera over a canyon edge than my smooth, featureless slab of a phone. Build a retractable iPod Touch-style "loop" into my next phone and I may not bother bringing a dedicated camera along on my next vacation.