Labeling yourself an “AI Optimist” or “AI Pessimist” means you’ve already chosen a side. It’s too early for that.

Interestingly but unsurprising, there are some who feel the need to adopt personal labels in this current revolution of AI. The most effortless and irresistible of these monikers are “optimist” and “pessimist,” which often find their way into personal opinions when trying to help others make the mental shortcut to understanding one’s overall position. We know what an optimist is; a person with a disposition or tendency to look on the more favorable side of conditions, usually expecting the most favorable outcome. The inverse is true of pessimists. By definition, these approaches favor a particular set of outcomes over others, which by most measures, qualifies as bias—the exact thing we’re trying to avoid when setting AI precedents. And therein lies the hidden danger.

Taking the position of an “ist” (optimist or pessimist) has its allure. It makes the work of contemplating massive, complex topics less burdensome by focusing on a subset of potential outcomes rather than contemplating the entire vector of possibilities. This approach may be defensible when events are better understood or more fully in focus—for example, climate change or the economy. Someone may have already considered all possibilities and arrived at their final position. This approach becomes more risky in situations like the advent of new technology, where outcomes and consequences are nascent and highly fluid.

Optimists and pessimists often focus on narrow applications of AI. The optimists often talk about the capitalistic version of AI that is very visible. That version of AI includes workforce productivity enhancements, personalized educational tools, and medical research accelerators. These applications may initially render specific tasks and professions obsolete but ultimately create a larger, more opportunistic economy. For whom we don’t know. Pessimists tend to be more focused on the invisible AI. The AI that determines which immigrants are allowed access into a country, which applicants are offered jobs, which young minds are granted college admittance, and who qualifies for welfare benefits. This invisible AI is hard to be optimistic about and understandably encourages pessimism.

Neither of these perspectives is necessarily wrong, though each is asymmetric. Asymmetrical thinking can be risky, often correlating with asymmetric influence, resources, or capabilities. Therefore, a disproportionally positive or negative outlook on AI could result in disproportionally unfair outcomes, as decisioning power is centered around a small, potentially overly-influential minority, unrepresentative of the larger population. The safeguard is inclusivity to capture diverse, non-stratified thinking to create the most informed and robust opinion possible.

We live in an Oppenheimer-like era, where we must consider all possibilities and make informed decisions by gathering myriad perspectives. We have not yet determined which perspectives to include, and those that sharply contrast need not be considered rivalrous. In short, it’s too early to take sides, which is what declaring yourself an “ist” effectively proclaims. If you need to brand yourself with a label, brand yourself as “AI Neutral” and remain open to all thinking and probability of outcomes. We simply don’t yet know what outcomes will result from AI, but we do know that premature opinions often lead to ill-informed decisions.