295 - Overload

A core reason why "branding" and "franchises" work is because they establish themselves as the "safe option", the option that maintains relative certainty even in otherwise uncertain and/or unfamiliar environments. This means that when someone is overloaded, they are more likely to select such a safe default option, as it minimizes their cognitive expense.

Reductions to uncertainty hold substantial value, as the existence of such franchises shows. This is also why you'll find fast food restaurant chains that charge substantially more than reliably superior local ones in many countries, because travelers often choose those perceived safe defaults, avoiding that cognitive expense that uncertainty brings.

The same is true universally when it comes to cognitive expenses related to uncertainty. When I have to evaluate options relating to traveling across countries, dealing with Visas, Immigration systems, and various other sources of wholly pointless paperwork, these are often major sources of uncertainty. Unlike food options, where food poisoning may be your biggest typical risk to consider, such sources of major uncertainty also entail major (often exponential) negative economic and practical impacts.

If you're randomly rejected entry to a country at the border you may lose thousands of dollars in booking accommodations, and any contingent business opportunities, and incur all of the major costs of last-minute travel and booking, as well as the massive spike in stress, all of which there may be no real remediation for. This is a low probability but massive impact event, which can never be overlooked when considering the uncertainty that such systems leave everyone exposed to through inconsistencies in application. Of course, this uncertainty costs countries a ton in lost revenue, which most remain oblivious to.

Critically, inherently probabilistic systems don't reduce your uncertainty, they merely bias it in approximate directions, but across only a subset of the applicable dimensions, parroting some cognitive biases absent the grounding of contextual understanding. You can reduce statistical "Noise" with simple deterministic algorithms and processes, or with the new types and generations of dominantly deterministic but dynamically growing and adapting graph-based systems that my team works with.

Today people are overwhelmed by cognitive biases like "Choice Overload" at every turn, many times every day, and that is why brands and marketing more generally remain increasingly effective. However, remove or massively reduce that cognitive expense and the entire advertising industry may well implode up its own @ss. Simple manipulative tricks found in marketing and dark patterns are transient methods of exploitation, not features to be expected in society 10 years from now.

In the next several years the "safest option" may reliably become the one that a human reaches with the help of viable technology. The Dilbert comic below illustrates OpenAI.