The Dressler Blog

I have opinions. Lots of opinions.

Back

Digital in four dimensions Digital environments don’t naturally have any dimension. Our sense of dimensional digital environments, like the graphical user interface, are fictions that exist to make human beings more comfortable. But, the simulation of dimensionality in digital environments is becoming crucial. Persistent worlds in virtual reality require the power to simulate three dimensions, plus time. MMORPG’s also require four dimensional simulated environments. As often happens with new technology, these comparatively frivolous applications precede more serious ones. The need to create and control fleets of self-driving vehicles requires these simulated environments. The logic that will govern these systems needs to be extensively tested in a non-real world environment before it can be installed in vehicles. It doesn’t matter how many car accidents there are every day in this country, if one self-driving car kills a pedestrian or rear-ends someone’s grandmother, it will be front page news. Improbable, a British technology company with strong VC-backing has begun to apply their Spatial OS platform to these and other problems (see article below). Because the future of digital will be in four dimensions. Application to Marketing: Many people in marketing like to believe what they do is “storytelling.” For a long time, that has been true. Most marketing took the form of little stories we told to consumers to illuminate something about a brand or product. I have made lots of little stories in my career. But this is a mindset with an expiration date. Four dimensional digital environments do not lend themselves to tidy story-telling. Consumers will not be passive consumers of these environments. They will be explorers. In a way, the stories will be created by the consumers with themselves as the protagonists. Instead of telling stories, marketers need to focus on creating environments rich in narrative possibilities. The worlds we create must tell the consumer something about our brand and our product, but we will not have complete control. Next Steps: I’ve said it before. Get yourself an Oculus headset and start playing around. Read More Not Ideal. “Impact” venture capital and the idealistic startups they fund have become big business. Idealistic startups are companies that are interested in making money by making the world a better place. But I believe that a startup can only afford to be idealistic in one aspect of its business. If a startup tries to be idealistic in multiple aspects of business, the startup will fail because the idealism overwhelms the focus on profitability. For example: suppose you decided to found a company dedicated to teaching more disadvantaged youth to code. That could work. Suppose you decided to found a company dedicated to teaching more disadvantaged youth to code and grow their own vegan food, using only open source technologies and employing exclusively transgendered individuals. I suspect such a company might fail. Not because the goals are unworthy, but because they are too numerous. This is problematic because idealists don’t confine their idealism to a single issue. So they over-reach in their idealism and tackle too much. Don’t get me wrong, idealism has a place in the startup world. Ideally, any company’s idealism should be central to how it plans on making money. (Like Warby Parker attacking the Luxottica monopoly to reduce the price of glasses.) Once a company is successfully making money from their core idealism, they can begin to tackle idealistic issues that are related to their core idealism. So an urban farming startup could (after becoming established) begin to tackle composting. But it would be a mistake for them to tackle adult illiteracy in Appalachia. Not because that issue isn’t important, but it’s just too far afield from their core idealistic competency. These examples are obvious. However, in the real world idealistic over-reach can be harder to discern. Founders may be unaware that their tech team has an ideological aversion to certain technologies that would reduce their total cost of ownership. This happens more often than you might imagine. It’s incumbent on the founder to be up front about their main goals and to restrict non-core idealism until the product is well-developed and established in the market. Application to Marketing: Issue-based marketing has gotten popular in the last five years. It seems that every brand is trying to “start a movement” with Millennials by talking about obesity or bullying or Syrian refugees. This is all well and good. But almost all of these efforts are outside of the brand’s core idealistic competency. They exist for the sake of marketing. McDonald’s makes cheap, filling food. You may think that that cheap and filling food is bad for people, but there is an idealism in that core mission. If McDonald’s wants to start a movement, they would be better off focusing on the persistent and lingering issue of hunger in our society, rather than bullying. A good rule of thumb is that if your issue-based marketing campaign requires substantial marketing support to publicize, then it is not core to your business. Next Steps: Are you an idealist? That’s great. Idealism makes the world a better place. But focus your idealism to make it a reality. Read More Machine learns French The existence of multiple human languages is a drain on innovation and commerce. I might be an entrepreneurial genius, but if I only speak Catalan my sphere of possible influence is severely limited. As a company, Google depends on the internet functioning as a free-flowing marketplace of ideas accessible to all. The more people use the internet, the more money Google makes. So naturally, Google has dedicated some of its extensive machine learning expertise to the problem of translation. Translations are tricky because they are rarely simple one-to-one replacements. Schadenfreude and ennui are two examples of words within closely related languages that don’t really have exact English translations. (English speakers handle such situations by stealing the words from other languages.) Once you get further afield, into Cantonese or Tagalog, translations become trickier and trickier. Idiomatic expressions make direct translations even more difficult. My Dutch Father-in-law says that the expression – “It’s like an angel peeing on my tongue” – is a Dutch way of describing something delicious. (I suspect he’s lying.) Machine learning with its cumulative understanding based on trial and trial and trial, is able to navigate these complexities in a way that simple word substitution systems cannot. Application to Marketing: Mistranslation can be disastrous for any company trying to sell its products in a foreign country. A single misused article will mark the company as both foreign and profoundly disrespectful of the native culture. Historically, Google has struggled in China partly because its search algorithms failed to grasp the subtleties of local idioms. A truly reliable translation service would allow marketers to become global without having marketers and support staff with native fluency in every language. Next Steps: Let Google solve this for you. They have lots of money. Read More Containers, not boxes. In the bad old days, when you needed to add functionality to a system, you installed a separate server to do that one thing. A server might handle authentication or security or email. It was just a box that got added to your server rack. The box couldn’t be turned on and off as needed. If you didn’t need that functionality, well, too bad. It just kept wasting space and capacity and energy. Then along came Virtual Machines (VM’s). This allowed administrators to create imaginary boxes within servers that could run a single function. The advantage of these systems is that multiple functions could be housed within a single, physical server. The disadvantage was that these Virtual Machines were a step removed from the physical server. Because they needed to run an operating system on top of an operating system, they didn’t have the performance of the old dumb boxes. Containers are an improvement on Virtual Machines. They are also self-contained systems that can coexist on a single server, but they have bare-metal performance since they don’t have a virtualization layer. So they have the performance of dumb boxes and the utilization advantages of VM’s. Application to Marketing: Yeah, I’m sure you don’t care. Most marketers are never going to have to think about this. Server capacity has gotten so cheap that marketers (and most other people in business) have stopped thinking about their server usage at all. But you should care a little bit. Because this is less a capacity and usage issue and more an issue of performance. If your current digital initiatives are using Virtual Machines or (heaven forbid) dumb boxes, you might be throttling performance down the road. Next Steps: When it comes to core infrastructure it pays to build for the next generation of marketing technology. Which means build in more speed, capacity, and performance than you think you need. Read More

Sign up to receive weekly Uneven Distribution emails about technology, design, marketing, and user experience.