In a recent profile published in New York Magazine, OpenAI CEO Sam Altman’s vision for the future of artificial general intelligence (AGI) has raised eyebrows and sparked controversy. Altman’s use of the term “median human” to describe the capabilities of AGI has led to concerns that he envisions a future where AI replaces what he perceives as “normal people.”
The “median human” conundrum
Altman’s assertion that AGI should possess the intelligence equivalent to a “median human that you could hire as a co-worker” has ignited a debate about the potential impact on the workforce. This vision implies that AI could perform various tasks currently carried out by human workers, from coding to practicing medicine. While Altman emphasizes the potential benefits of AGI, it’s hard to overlook the implications that this could lead to job displacement for many.
In a 2022 interview on the Lex Fridman podcast, Altman expanded on this concept, suggesting that AGI could perform tasks typically reserved for humans working remotely behind a computer. This notion leaves some wondering if those leading “median” lives might soon find their job prospects diminished.
Widespread use of “Median” terminology
It’s worth noting that Altman is not alone in using the term “median” in discussions about AGI. Many AI blogs and publications have employed similar terminology, and CNBC even published an article titled “How to talk about AI like an insider” featuring this concept. The idea behind this terminology is to describe AGI’s capabilities as falling within the range of what an average or median human can achieve.
However, this comparison has not been without its critics. Brent Mittelstadt, director of research at the Oxford Internet Institute, finds using the term “median” concerning and vague. He points out that there’s no concrete, measurable comparison of human intelligence within AI research, making it challenging to define a universal standard for what constitutes a “median” human.
The challenge of equating performance with intelligence
One of the key issues critics raise is the conflation of performance with intelligence. While AI models can achieve human-level performance in certain tasks, equating this with human intelligence raises ethical and philosophical questions. Henry Shevlin, an AI ethicist and professor at the University of Cambridge, notes that discussions around intelligence tend to be more sensitive due to the complex nature of human cognition.
Critics argue that attributing agency, comprehension, cognition, or reasoning to mechanistic AI models oversimplifies the complexity of human thought and consciousness. It reduces humans to mere statistical averages, a perspective many find dehumanizing.
Altman’s broader vision
Sam Altman is a figure of considerable influence in the AI industry, and his vision extends beyond just replacing “normal people” with AGI. He has frequently discussed using AI to address global challenges, such as climate change, Universal Basic Income, and clean energy. Altman has stated that OpenAI’s mission is to ensure that AGI benefits all of humanity if it ever becomes a reality.
However, as is often the case with ambitious visions, there is a diversity of opinions on what constitutes “benefiting all of humanity.” Altman’s use of the “median human” concept highlights the challenge of reconciling his vision with the ethical and practical implications of AGI development.
The controversy surrounding Sam Altman’s vision for AGI and his use of the term “median human” underscores the complex ethical and practical considerations inherent in AI development. While Altman’s goal of using AI to solve global problems is commendable, the potential consequences of replacing human workers with AGI deserve careful consideration.
As the AI field continues to advance, engaging in a meaningful dialogue about the implications of AI technologies on society, the workforce, and the essence of human intelligence is crucial. The debate surrounding the terminology used in the AI community reflects the need for a nuanced and ethical approach to AI development—ensuring that the benefits of AGI are truly inclusive and equitable for all of humanity.
