[VIC – 89] Fixed vs variable cost. Mesh networks. The uncanny valley. So many clocks.

Business & Money

I work at a software company that provides a marketing intelligence platform (marketing intelligence = business intelligence for marketing). One of the things that make our offering different than technologies of yesterday is that we provide it as a SAAS platform, or software-as-a-service where customers can pay a flat monthly fee for the right to use our software from any web browser or mobile device.
Business intelligence in the old world was provided as most software applications were provided in the past. There was a huge up front cost for software installation & integration, then a small annual maintenance fee. In other words, customers recognize a big fixed cost up front, with negligible variable costs over time. In the SAAS model, there is no massive up front cost, but instead, a variable cost over time (based on usage) that gets recognized as an operational expense every month.
Most software is going the way of the SAAS model these days, but the fixed cost vs variable cost debate is alive and well. You can find a great example of this if you compare Snap(chat) and Facebook. These companies have taken a drastically different approach to their cost structure which could have massive implications over time.
Facebook has built 8 or 9 massive data centers (don’t quote me on that number) all over the world in which they’ve invested billions of dollars. They recognized these data centers as a massive fixed cost up front, with the variable costs over time (electricity, wages for employees working in the data center, etc) being relatively marginal compared to the initial investment. This means that over time, as Facebook has adds more users (closing in on 2 billion) and more advertisers to its platform, their margins and profitability have grown significantly. They are basically spreading that fixed cost over more and more users and customers as they scale, which is a boon for the business model.
Conversely, Snap has decided not to invest in their own data centers, but instead rely on the infrastructure of other cloud providers (Google Cloud) for computing and storage. Thus, they’ve opted for a variable cost or operational expense model that continues to grow linearly as they bring on more customers and advertisers. Of course, they can negotiate better rates as they spend more money, but this model does not scale nearly as well as the one Facebook chose.
Not to say that Snap’s can’t succeed with its model, but it will definitely be much harder, especially given that they’ve chosen to focus on a much smaller niche market.

Human Progress

As the nation continues to be bombarded with increasingly powerful storms (cough cough.. climate change.. cough cough), one of the most important things to consider is cell phone coverage. As critical infrastructure and power grids fail, people often rely on wireless networks to call for help and stay tuned for the latest weather updates. But said infrastructure includes cell towers, which of course rely on electrical grids to stay on line and route calls and data. So what happens when the power goes out?
One startup I’ve kept an eye on over the years, goTenna, is trying to solve this problem. They’re building mesh networks which allow for communications to be routed via other mobile devices. So in an area without power, and thus no cell coverage, data can hop from device to device until it gets from sender to recipient. Thus, more wireless devices in an area equates to a stronger mesh network, while completely circumventing the traditional cell network.
This technology has the ability to not only improve communication during natural disasters, but also improve communication in any densely populated area or areas where coverage is hard to reach. Think about concerts, for example, where too many devices are trying to ping the same tower. Or in NYC when you’re traveling on the subway in between stations. And the best part? Since the number of devices is exploding with wearables, IOT, and the like, you are inherently increasing the strength of the network without building any additional towers or wireless infrastructure. The devices themselves are the network.
Here’s a quick video of the founder giving a presentation at the New York Times as part of their “Cities For Tomorrow” conference series earlier this year.


There’s this concept in the tech community known as the “uncanny valley.” It is the concept that is most often used to describe the development of humanoid robots (and artificial intelligence by extension). Basically, as we develop robots and intelligence technologies, for a long time these are very clearly non-human (take a Roomba vacuum or a Firby for example). Even when they start to stand upright and have a silicone covering that looks a lot like human skin, they lack the basic skills of non-verbal communication, the fluidity of motion, the ability to discern things like sarcasm or deceit, and a whole host of other basic underpinnings of being a human being, even if the outward appearance seems to suggest otherwise. Not until we reach near perfection on all of these “soft” components of humanity, would we be able to escape the “uncanny valley” where the robots are not too far from human, but still very creepy and easily identifiable replicas of ourselves.
But I believe this concept can just as easily be applied to areas outside of technology. Specifically, this week I’ve been thinking about it in terms of human psychology. There’s a long list of human deplorables where disassociation is fairly straight forward. Hitler, Charles Manson, Ariel Castro, Osama Bin Laden, you’d be hard pressed to find many people that would willingly associate themselves with these people. These are the robots that are very clearly non-human.
But then you have this huge gray area, this “uncanny valley” where individuals seem fairly human in some regards, but are merely creepy replicas. And these people are coming out of the wood work with increasing frequency. A number of people have dropped (or been forcefully removed) from my life because it became pretty clear to me that they lacked basic elements of what I consider to be core to humanity. Things like recognizing the fact that certain minority groups have been discriminated against and fundamentally disadvantaged in this country for a long time. And that these problems are still problems today. Things like embracing the idea that people that love each other, regardless of creed or color, should be encouraged and supported if they decide to spend their lives together in marriage.
In general, I’d say the world is stuck in an “uncanny valley” today. There are faint signs of equality and progress if you look closely, but it’s a mere replica of where I hope things end up at some point in the future.

My Latest Discovery

I never realized just how many clocks are in our lives. In my typical day, there’s the microwave, the oven, my alarm clock, both iPhones, my lap top, Apple watch, and a million other clocks in public spaces. I’ve discovered that looking at the clocks less brings more peace to my life. And I don’t really lose any sense of time. Given the routines of everyday life, and our natural instincts of time-based on the sun and what’s happening around you, you’d be surprised how accurate our sense of time is without ever looking at clocks.

I’ll be writing more about this one in the future…