Eye tracking

This company coming seemingly out of nowhere with a stand-alone eye-tracking device in a partnership with Steel Series, an eye-tracker-enabled gaming laptop in partnership with MSI, and a growing number software partnerships with companies like Ubisoft and Avalanche might be a new force in gaming, but it’s been studying eyeballs — and tracking them — for a long, long time. When you play Assassin’s Creed: RogueThe Hunter, The Hunter: Primal and many more games to come (they showed me some, but I can’t talk about them), you will experience what data researchers and ability engineers have known for over a decade: Your eyes know what you’re thinking before your hands do.
My time at Tobii is full of interesting experiences, but one thing sticks out — and it happens multiple times — the people here are almost supernaturally aware of what I’m thinking. In multiple interviews, Tobii employees will comment on how I am interacting with them, how I maintain eye contact while they are speaking, to encourage them to continue; how I nod at them to indicate that I am listening; and smile to influence them to expand on what they might be saying. Subtle interview tricks I learned so long ago I’m no longer aware of doing them. But the people I talk to at Tobii are aware. They’re aware and responding. Because they’re watching my eyes. And they’re programming computers to do the exact same thing.

Profile of Swedish company Tobii and their eye-tracking technology.

This bit is interesting, on how eye tracking can remove one step of abstraction from interaction with software interfaces using the typical modern computer hardware.

Bouvin demonstrates by picking up a pencil from the table in front of us. First he looks at the pencil, then he moves his hand to pick it up. In the computing world, we’ve become used to this type of interaction, but everyone who is currently alive who knows how to use a computer has had to train their mind to add a step between what they see and interacting with it.
We’ve had to learn the motor control of seeing, and then moving a cursor with a mouse or a trackpad, and then interacting.
Tobii eye tracking will remove that inter-evolutionary step, making it possible for us to interact with computers in the same way we interact with the world. Look at a thing. Interact with the thing. No cursor required.
“What you get with eye tracking is you can substitute pretty much all of that positioning, all that directional input, because your eyes are there before you touch,” says Bouvin. “You’re looking there. The movement and directional thing becomes unnecessary. Same with a mouse. You’re already there with your eyes. Moving the mouse cursor there is a step you can more or less do without. The only thing that’s left is the actual action.”


An engineering theory of the Volkswagen scandal

Paul Kedrosky in The New Yorker:

In a powerful book about the disintegration, immediately after launch, of the Challenger space shuttle, which killed seven astronauts in January of 1986, the sociologist Diane Vaughan described a phenomenon inside engineering organizations that she called the “normalization of deviance.” In such cultures, she argued, there can be a tendency to slowly and progressively create rationales that justify ever-riskier behaviors. Starting in 1983, the Challenger shuttle had been through nine successful launches, in progressively lower ambient temperatures, across the years. Each time the launch team got away with a lower-temperature launch, Vaughan argued, engineers noted the deviance, then decided it wasn’t sufficiently different from what they had done before to constitute a problem. They effectively declared the mildly abnormal normal, making deviant behavior acceptable, right up until the moment when, after the shuttle launched on a particularly cold Florida morning in 1986, its O-rings failed catastrophically and the ship broke apart.
If the same pattern proves to have played out at Volkswagen, then the scandal may well have begun with a few lines of engine-tuning software. Perhaps it started with tweaks that optimized some aspect of diesel performance and then evolved over time: detect this, change that, optimize something else. At every step, the software changes might have seemed to be a slight “improvement” on what came before, but at no one step would it necessarily have felt like a vast, emissions-fixing conspiracy by Volkswagen engineers, or been identified by Volkswagen executives. Instead, it would have slowly and insidiously led to the development of the defeat device and its inclusion in cars that were sold to consumers.
If this was, in fact, the case, then Horn was basically right that engineers were responsible. The scandal wouldn’t have been caused by a few rogue engineers, though, so much as by the nature of engineering organizations themselves. Faced with an expensively engineered diesel engine that couldn’t meet strict emissions standards, Volkswagen engineers “tuned” their engine software. And they kept on tuning it, normalizing deviance along the way, until they were far from where they started, to the point of gaming the emissions tests by detecting test conditions and re-calibrating the engine accordingly on the fly.

My uncle who works in aerospace/aeronautics has long told me that most of the public has no idea how far the Space Shuttle program were pushing the boundaries of what was possible.

Encouraging to see Kedrosky get a byline in The New Yorker. Om Malik has a piece in The New Yorker on Jack Dorsey's return to Twitter. I love The New Yorker, but their coverage of the tech industry has occasionally struck me as a bit biased (part of the larger East Coast/West Coast culture war, with Silicon Valley in the role of the nouveau riche). I'm encouraged to see their stable of tech voices broadening.

Warriors, come out to play

Coming off legacy stuff is complicated for customers. For something like an Oracle database, that software has all the financial information and is all but impossible to leave behind. When Oracle introduced its cloud and fast data analysis products, they were aimed primarily at those virtually locked-in legacy customers.
All that now looks shaky. At its annual customer conference, Amazon on Wednesday introduced new features and services aimed at offering legacy customers on all sorts of computing systems not just easy ways to get off the old technology, but also better and faster ways their old data can work on A.W.S.
Among the most notable, there was a 47-pound data storage device that A.W.S. would ship to a customer, and for $200 would suck down 50 terabytes of data, incidentally converting it from an older system to a more modern one. There was a service called Database Migration, which takes data in proprietary systems and converts their schema to open-source products.

Fun to see my old colleagues and friends at Amazon Web Services continue to move aggressively. Clever ideas here to minimize switching costs for customers on legacy systems.

First AWS was for individual developers, then for startups, then for some large scale technology companies. Now it's going after any company that needs computing. Up (down?) the ladder they go.

This is how technological progress works

Generation n-2: This is our save icon. It's for saving a file to a floppy disc.

Generation n-1: I don't use floppy discs anymore, I only save files to my hard drive, but I remember floppy discs, so the iconography still works for me.

Generation n: Why is that the save icon? 

Generation n-1: Because people used to save files to floppy discs, then later to hard drives, but the icon stuck around. That's a 3 1/2 inch floppy disc, which actually isn't really floppy. There was a 5 1/4" disc which was floppy, but no one uses that as the save icon.

Generation n: Why not update the icon? How about something like one of these instead?

Generation n+1: What does it mean to save a file? Why do we even need that function. I just update my files in Google Docs or Quip, they are always up-to-date.

Generation n+2: What are files?

Our precious soil

Top cities became hotbeds of innovative activity against which other places could not easily compete. The people clustered together boosted each others’ employment opportunities and potential income. From Bangalore to Austin, Milan to Paris, land became a scarce and precious resource as a result; the economic potential of a hectare of a rural Kentucky county is dramatically lower than that of a hectare in Silicon Valley’s Santa Clara county. And there is only so much of Santa Clara to go around.
Yet more Santa Clara could be built, were it not for the second and more distressing factor behind land’s return: the growing constraint imposed by land-use regulation. The Santa Clara town of Mountain View, for instance, is home to some of the world’s leading technology firms. Yet nearly half of the city’s homes are single-family buildings; the population density is just over 2,300 per square kilometre, three times lower than in none-too-densely populated San Francisco.
The spread of land-use regulation is not hard to understand. The clustering that adds to local economic vibrancy has costs, too, as the unregulated urban booms of the 19th century made clear. Crowded slums were fertile soil for crime and epidemics; filthy air and water afflicted rich and poor alike. Officials began imposing new rules on those building in cities and, later, on those extending them: limiting heights and building designs; imposing maximum densities and minimum parking requirements; setting aside “green belts” on which development was prohibited. Such regulations have steadily expanded in scope and spread to cities around the world.
As metropolitan economies recovered from their mid-20th-century slump populations began growing again. The numbers of people living in the central parts of London and New York have never been higher. And as demand for quality housing increased the unintended consequences of the thicket of building regulation that had grown up in most cities became apparent.

Great piece in the Economist about how land has become a constrained commodity, a brake on our economic growth. A lot of wealth has been made off of rent capture, and most of it accrues to the already wealthy while the middle class are saddled with mortgage debt and the poor, who rent, are just priced out of desirable regions.

This is a nightmarish scenario for the economy. It's bad any time you have any constraint on growth, but when that scarce commodity is land, it seems particularly difficult to remove because it has to be done through a political system that is in thrall to the moneyed, connected, real-estate-owning class.

The ugliest effect of the return of land, though, may be the brake it applies to the economy as a whole. One of the main ways economies increase worker productivity, and thus grow richer, is through the reallocation of people and resources away from low-productivity segments to more efficient ones. In business this means that bad firms go bust and good ones grow to great size. Something similar should hold for cities. Where workers can be put to use at high levels of productivity labour scarcity will lead to fast growing pay packets. Those pay packets will attract workers from other cities. When they migrate and find new, high-paying work, the whole economy benefits.
Mediterranean Avenue to Boardwalk
But that process is now breaking down in many economies. For workers to move to the high wages on offer in San Francisco, they must win an auction for a home that provides access to the local labour market. The bidding in that auction pushes up housing costs until there are just enough workers interested in moving in to fill the available housing space. Salaries that should be sending come-hither signals are ending up with rentiers instead, and the unfairness can trigger protest, as it has in San Francisco. Many workers will take lower-paying jobs elsewhere because the income left over after paying for cheaper housing is more attractive. Labour ends up allocating itself toward low-productivity markets, and the whole economy suffers.
Chang-Tai Hsieh, of the University of Chicago Booth School of Business, and Enrico Moretti, of the University of California, Berkeley, have made a tentative stab at calculating the size of such effects. But for the tight limits on construction in California’s Bay Area, they reckon, employment there would be about five times larger than it is. In work that has yet to be published they tot up similar distortions across the whole economy from 1964 on and find that American GDP in 2009 was as much as 13.5% lower than it otherwise could have been. At current levels of output that is a cost of more than $2 trillion a year, or nearly $10,000 per person.

First and foremost, let's acknowledge that this is all solvable if we just relax land use regulations, build more housing, and increase the population density of our urban centers. Supply and demand still works in this universe. For a variety of reasons, some clear to me, like NIMBYism, some not clear to me, that seems intractable. Does the new David Simon show explain how hopeless it all is? I need to watch it so I find some entertainment value in my despair.

If that's a road to nowhere, can the germ of a solution come from the private sector? Tech companies tend to be creative about trying to solve problems that constrain their growth because they arise from a culture of ignoring accepted impediments. For now, though, they haven't made a ton of progress on this issue. At most they've turned to providing shuttle services with wi-fi that widen the geographic footprint in which their employees can live and still get to work.

But that doesn't work if real estate is expensive everywhere within that radius. What's next? Perhaps a deeper investment in conferencing or virtual reality technology to amplify the sensation of proximity and intimacy of even the furthest flung workers? It's long been a promise of technology, but it has yet to be realized in full.

What about turning office space into living space at night when it lies idle? It's sounds ridiculous, but a company did that with the slack time of cars and seems to raise a billion dollars of capital every other week. I know, it sounds terrible, living at the office, but I'm wary of making paternalistic prescriptions about how a person should spend their free time. Some of my closest friends in life are people I spent a lot of time with at the office at Amazon during my years there.

Maybe tech companies will start their own housing developments? The economic productivity of the average productive tech worker likely still exceeds their compensation, and tech companies, led by Google in particular, have been aggressive in pushing into that gap with increased salary, benefits, stock options, etc. Housing is just another form of investment, and they need not provide it for free, it could just be subsidized. Of course, that means they'd need to acquire land, and that, paired with a thicket of land-use regulations, still restrict the human density achievable with even the most aggressive development.

Finally, solving the problem for just tech workers doesn't feel like a path towards solving it for the rest of the population. Frankly, no one feels much sympathy for tech workers these days (with the exception of some Amazonians who are working long hours, though I'm still suspicious; a lot of it feels secretly like Schadenfreude in disguise). Bay Area complaints about high real estate costs are going to fall on deaf ears, even if it's symptomatic of a dangerous trend for our economy, as noted by the Economist article.

Group all the things I miss the most about NYC, and the root of all of them was the unmatched human density. It's not just the visceral sensation of the people around you (which some dislike), but the diversity of businesses and communities that can sprout and thrive when the potential customer base is so large and tightly packed. The variety of entertainment, like theater and museums, the variety of local cuisine, the ability to find someone to share almost any interest, from curling to revivalist arthouse cinema to hip hop dancing.

I have no answers, only a longing for the Bay Area to experience the liveliness of density in the physical world to match that of the the networks and communities they've built in the virtual space, where real estate is cheap and plentiful.

History lessons from Singer sewing machines

So, this crowdsourced sewing machine could be sold and distributed widely. But why did Singer's prove to be the one with staying power? It was not due to Isaac Singer himself, who Liebhold describes as more of a “scalawag” than a businessman. Rather, it was the smart businessmen who took charge of the company, particularly lawyer Edward Clark, who co-founded I.M. Singer & Co. He created the company’s early advertising campaigns and devised the “hire-purchase plan” for customers who could not afford the machine’s high price—the first installment-payment plan in the United States.
The company expanded the practice of door-to-door sales, in part because the hire-purchase plan required canvassers to collect weekly payments, but which also allowed salesmen to bring the product into prospective customers’ homes, and show them how such a novel machine could simplify their lives. The company opened up flashy showrooms where it could demonstrate how the machines work (a scale model of an original Singer showroom will be included in the exhibit), and took machine demonstrations to county and state fairs.
Singer Co. also became active in buying up used sewing machines and tamping down secondary markets of used sewing machines. Like the latest iPhone today, Singer would roll out a new sewing machine model and encourage consumers to replace their old one.

A fascinating history of how Singer sewing machines emerged from a pack of companies starting from roughly the same place. In fact, all the companies had joined together to create the first patent pool so that they could shift from litigating the bejesus out of each other to growing the market.

So many of the lessons of Singer's success still echo in today's tech world. How it's not enough to build a better mousetrap, or even the first mousetrap. You need to figure out distribution and how to overcome customer hangups and purchase hurdles. We tend to have a cultural bias against sales and marketing in the tech world, and perhaps in business in general. We revere product design as a craft. In our new networked age, however, the cognitive challenge of getting users to hear about and try out your product or service is quite underrated.

Uber for disembodied companionship

Kashmir Hill worked as an invisible boyfriend/girlfriend for a month. We are already living with one foot in the near future.

With each job, I would see the person’s first name, last initial and hometown; “how we met;” and my own assigned name, age, and which of six personality types they’d given their Invisible. Now I’m adventurous and fun. Now I’m cheerful and outgoing.
There were 3 major rules:
1. I was always supposed to be upbeat in my messages.
2. I’m not supposed to break character.
3. No sexting. (Photos are blocked on the service.)
I’d get the story of how we met and the last 10 messages we’d exchanged. This setup is designed to create the illusion of continuity; ideally, an Invisible Boyfriend would seem like a steady, stable presence in a user’s life, instead of what it really is: a rotating cast of men and women. And it is both: a woman who works for the service previously told me she prefers playing the role of boyfriend because she knows what a woman wants to hear.

Hill probably got paid more for this article than she did for her work as invisible companion.

It’s hard to put a price on love. But Crowdsource did. It’s worth a whopping five cents. That’s how much I got paid to write each of these texts.
If I spent an hour answering texts, and took the full five minutes to write each one, I’d be making 60 cents an hour, far below the minimum wage. This is legal because all the workers on the platform are classified as independent contractors rather than employees. “Contributors have a tremendous amount of control over their decisions—for example, when to perform a task, when to complete it, and even if they want to complete it at all,” said Jeffrey H. Newhouse, an employment lawyer at Hirschler Fleischer, by email. “That means the contributor isn’t an employee and, as a result, employee protections like the minimum wage don’t apply.”

Not surprising considering the required skill set is the ability to write with decent grammar, that's reasonably commodified.

As with Uber, the laborers already fear displacement by technology like self-driving cars.

I assumed that, when artificial intelligence is good enough, Invisible would just cut the crowdsourced humans out of the equation and use chat bots, which you don’t have to pay per message, instead.
No, he said. “Having humans in the flow is the key to the service,” said Tabor. “There are things that only humans can respond to and understand, like inside jokes.”

Loneliness continues to be one of the great problems being tackled by technologists. See also: Her, The Diamond Age.


Daniel Miessler thinks we're underestimating the Internet of Things.

IoT isn’t about smart gadgets or connecting more things to the Internet. It’s about continuous two-way interaction between everything in the world. It’s about changing how humans and other objects interact with the world around them.
It will turn people and objects from static to dynamic, and make them machine-readable and fully interactive entities. Algorithms will continuously optimize the interactions between everyone and everything in the world, and make it so that the environment around humans constantly adjusts based on presence, preference, and desire.
The Internet of Things is not an Internet use case. Quite the opposite, the IoT represents the ultimate platform for human interaction with the physical world, and it will turn the Internet into a mere medium.

Great succinct read on the key technical components that will make up his vision for IoT.

The Internet of Things is a terrible name, which doesn't help matters. Miessler suggests four alternatives though they don't catch me on first read. Something less tech jargony, not terrifying (the phonetics of daemon aren't great though it's a cool word), shorter (a la the singularity). it sounds silly but a name is all that gives a futuristic scenario like this a personality right now.