The Dressler Blog

When you build digital stuff all day, you develop opinions. Lots of opinions.

Back

Digital Trends

Approaching the Quantum Era Technology blogs are abuzz with rumors that Google’s quantum computing lab in Santa Barbara is nearing some kind of breakthrough. While other quantum computers have been built, notably by D-Wave and IBM, the systems have a small number of qubits (processing units) and have trouble maintaining coherence, the quantum state necessary to perform quantum processing. Google is rumored to have built qubits that can maintain coherence for longer and scaled up to 100 qubits, a point at which we can power real world quantum applications. Quantum computers are supposed to be exponentially better than conventional processors at certain types of functions, namely image and pattern recognition and machine learning. Application to Marketing: I believe quantum computers are going to cause a revolution in business, marketing, and technology. The reason is simple. Systems tend to be self-organizing from first principles. In the case of technology, the first principle is the binary processing chip. Everything is either false or true, zero or one. While computer science has grown infinitely complex, it continues to adhere to an approach that is intolerant of ambiguity. Spend enough time around programmers and you begin to appreciate how their work has encouraged them to demand certainty and definition. They don’t so much dislike ambiguity, as find ambiguity inapplicable to their work and experience. Something either is or is not. Maybe doesn’t happen. Today’s computer scientists imagine quantum computing primarily as an adjunct to existing conventional processors. They focus only on the weaknesses of their current systems (for image recognition and machine learning.) Their approach suggests using a conventional computing system to handle the data before and after the “quantum interval.” They believe that quantum computing will be a discrete event in the largely unchanged processing of data. Frankly, I doubt this will last. Certainly, the first generation of quantum computer scientists will apply quantum processing discretely, but quantum states do not lend themselves to limited application (because observation leads to decoherence and much of the information is lost.) Inevitably, a new generation of programmers – quantum natives – will become comfortable with extended quantum processing. Why translate back and forth between fundamentally incompatible approaches if you don’t have to? Quantum computing requires more than new chips, it requires a revolution in thinking. Technology will never be the same. Next steps: Read what you can about quantum computing. The article below from MIT Technology Review will get you up to speed on Google’s efforts. Read More Systems Evolve Once upon a time, information technology was like physics. The belief was that everything could be known and that, once known, everything could be controlled. Now, many people believe that information technology is more like biology. The systems have grown so complex that no single individual is capable of grasping them in their entirety. Everyone relies on subject matter experts to explain and maintain aspects of their code base. In technology circles, the question is whether you would rather have a bank’s systems or Google’s. A bank powers their technology using a massive number of Sun servers and Oracle arrays. If anything goes down, the entire system goes down, but that also means that it is possible to locate the problem immediately. Companies like Google, or Netflix, or Facebook work under the assumption that any single aspect of their system may fail at any time. The point is not to establish total control, but to create a resilient, fault-tolerant system that can evolve new capabilities and eliminate bad code. Hardware is not evaluated and maintained individually, but as part of a larger collective. If the group of servers continues to function at an acceptable level, it doesn’t matter if individual servers fail. Application to Marketing: While technology companies have learned to prioritize flexible and resilient systems, marketers continue to assume they can exercise an impractical level of control over technology. Basic services like load balancing through a cloud service provider or working with companies like Cloudflare that offer resiliency and protection from denial of service attacks are treated like unnecessary overhead. And yet, when technology fails or sites go down, marketers quickly realize how dependent on their digital infrastructure. With an infinite time horizon, all systems fail. Preparation for that day only looks like overhead until that day comes. Next Steps: What is your time to recovery if some aspect of your digital infrastructure were to crash right now? If you don’t know the answer already, chances are the answer won’t please you very much. Read More The Ad Blocking Reality This is a rabbit that we cannot fit back into the hat. Ad blocking is here to stay. As marketers, the fault is entirely ours. (Actually, it may be more my fault than yours since I worked on some of the first banner ads back in the day.) Ad blocking exists because the marketing and publishing industries got lazy. We tied the future of the industry to a resource with infinite supply – digital media. And then we neglected the hard work of attribution. Without establishing definitively which ads were seen and which were effective, we de-valued digital media to the point where desperate, cash-strapped publishers loaded up pages with more ads than any consumer could possibly wish to see. Ad blocking is already approaching 50% of the most desirable demographics. Soon the only people who won’t be blocking ads will be your grandmother and people who actually work in advertising. Application to Marketing: Everyone has a solution for ad-blocking. Some rail against the consumers who are “stealing” content. Others insist that we just need to limit inventory. But limited inventory is a non-starter for anyone with even a cursory knowledge of game theory. The only way to have a market-based limit on inventory is to nail down attribution. I think the blockchain could offer a solution that would satisfy marketers, publishers and privacy advocates, but I’m open to anything. If we don’t solve this, things are going to get much, much uglier very soon. Next Steps: If I were a client, I would stipulate that my ads be the only ads that appear on a page. That, at least, is a start. Read More Google Doomsday Scenario Enough, already. This summer has seen a spate of articles to the effect that “all is not right at Google!!!” The latest of these articles focus on their decision to cancel their modular smartphone, known as Project Ara. I admit this makes me sad. I was excited about these phones. But Google remains one of the only technology companies on earth that is actually profitable. Let me amend that, Google isn’t just profitable. Google is insanely profitable. When Larry and Sergey took over, I was nervous that their utopian streak would overwhelm the discipline put in place by former CEO Eric Schmidt. Instead, they went in the opposite direction, trimming unprofitable projects and streamlining divisions. They have demonstrated a deep level of corporate discipline and an awareness of future challenges. The Alphabet reorg is an attempt to get in front of future anti-trust problems. Many other corporations should aspire to Google’s vision and planning. Application to Marketing: The lesson (as always) is to take what you read in the press with a grain of salt. No journalist writes an article titled “Google continues to be an awesome company” because no one would bother reading that article. It’s much more fun to predict catastrophe. If you’re right, you’re a genius. If you’re wrong, no one remembers anyway. Google will continue to be the dominant force in marketing for the next ten years (at least.) Next Steps: Move along, nothing to see here. Read More

Give us your email to sign up for our weekly Dressler Digital Trends. Stop trying to keep up and start getting ahead.