Discussion Topic Frequency
In almost every dinner conversation, the topic of “p (doom)” tends to arise. This statistic, also known as the probability of doom, has gained popularity among Silicon Valley researchers, as reported by The New York Times.
When engaging a tech enthusiast in a discussion about AI, be prepared for inquiries regarding your stance on the likelihood of AI causing humanity’s demise. This rather morbid subject has become a common point of conversation lately, often accompanied by a cup of coffee.
The concept of p (doom) serves as a convenient way for individuals within the AI community to gauge each other’s perspectives on the technology. Those with a 50% p (doom) rating might be labeled as “doomers,” akin to the former interim CEO of OpenAI, Emmet Shear, while individuals with lower percentages, such as 8% or 5%, are viewed as optimists. Regardless of one’s position, it acts as a conversational icebreaker, facilitating discussions on a somewhat grim topic.
Aaron Levie, CEO of Box, mentioned to The New York Times that this topic arises frequently during mealtime conversations.
Rise of Technical Terminology
The emergence of terms like p (doom) is not surprising in the tech realm. With the proliferation of relational AI and advanced speech models like OpenAI’s ChatGPT, concerns surrounding technology, ranging from everyday worries to doomsday scenarios, have become more prevalent. Prominent figures in the tech industry, including OpenAI CEO Sam Altman, have actively contributed to fueling these apprehensions.
The exact origin of the term “p (doom)” remains unclear. According to The New York Times, it likely originated on the forum LessWrong over a decade ago. Programmer Tim Tyler initially used it to denote the likelihood of catastrophic outcomes without specifying a precise timeframe or definition of “doom,” as he shared with the publication.
Eliezer Yudkowsky, a prominent figure on the website, is known for his radical views on AI and has even suggested extreme measures like bombing data centers to avert an apocalypse. His interpretation of p (doom) transcends mere numerical predictions, as conveyed to The New York Times.
Varied Perspectives
Some viewpoints on AI’s future echo the sentiments of the technology’s prominent figures, often dubbed as the godfathers, who have expressed remorseful warnings about its potential dangers. Geoffrey Hinton, a notable figure among them, recently departed from Google, expressing regret over his contributions and issuing stark cautions about AI’s societal risks.
Nevertheless, there are individuals in the industry who maintain a pragmatic outlook. Aaron Levie, for instance, indicated to The New York Times that his p (doom) rating is “as low as it could be,” expressing concerns about regulatory overreach hindering technological progress rather than an imminent AI doomsday.
The focus on sensationalized AI doomsday scenarios in popular culture could overshadow the more mundane but pressing issues faced by AI today, such as widespread copyright infringements. The prevalence of discussions around p (doom) reflects a sense of existential awareness among those involved in developing AI, prompting introspection on one’s own apprehensions about the future.
