Effective Altruism (EA), a movement committed to using evidence-based decision making to maximize positive impact, has garnered widespread attention for its pragmatic approach to doing good. However, unlike other philosophical or ethical movements, it has gotten a lot of negative press lately.
From the downfall of SBF and now former member of OpenAI’s board Helen Toner, this ethical philosophy is causing a lot of people problems.
As EA’s reach spreads and tries to affect changes they see as good (by their definitions, which are of course assumed to be objective) its influence has made its way to the upper echelons of AI as it did with cryptocurrency.
We won’t go into the nitty gritty of that situation as that story is still quite fresh and has been a very prominent feature in news cycles. This article aims to explore how EA as a movement has been poking a stick in the spokes of its own wheels.
As EA has grown in the 2010s it has released quite a lot of studies and recommendations but today, we’ll only look into one and see how many issues arise with EA’s approach as pointed out by a study on looking at Effective Altruism through the lens of American Pragmatism . So, let’s go straight to the source and look into William MacAskill’s use of Quality-Adjusted Life Years (QALYs) formula and is expressed as:
QALY= Time x Health Related Quality of life
The proposition unfolds as follows: if one wants to become a doctor, the optimal course is to practice in a developing country where the demand for medical professionals is assumed to be higher. In such settings, the potential for a significantly higher marginal value emerges, translating into a substantially amplified impact – approximately 100 times more impactful, gauged in terms of Quality-Adjusted Life Years (QALYs).
Conversely, opting to be a doctor in a developed country presents a contrasting scenario. Here, the abundance of medical professionals surpasses the needs of the populace, resulting in a diminished impact measured in QALYs. The central tenet revolves around the concept of replaceability. MacAskill contends that a surplus of well-intentioned individuals, including doctors, is readily available to fill the role if one were to abstain from the profession.
Now from a counterfactual perspective, what if, prior to committing to a medical career, one meticulously assessed the anticipated positive contributions compared to alternative occupations? This entails a quantitative analysis where one QALY symbolizes a year lived in perfect health. Delving into the role of a doctor in a developing country, the potential to save approximately 300 lives over a 40-year career span is postulated, factoring in an assumed additional lifespan in perfect health of 36.5 years per life on average.
This analytical and strategic approach mirrors the essence of Effective Altruism, wherein decisions are informed by a meticulous evaluation of potential impact and a commitment to optimizing positive outcomes on a global scale.
However, this approach oversimplifies the intricacies of complex issues. Measuring subjective well-being and long-term impact proves challenging, highlighting the limitations of relying solely on quantitative metrics to gauge success. Stripping the qualia out of everything unknowingly transforming all that information he has gathered and analyzed into bias. Thus, leading to a whole lot of bad decision-making not just at an individual level but on organizations as we all saw what transpired the past week when EA force fed its methods into the corporate governance of OpenAI.