The Contagion of Fear

Late one night during the author's first year at university, a group of computer science students hatched a devious plan to spread fear among their humanities peers. They convinced themselves that a computer virus was spreading through the general computer lab and warned the unsuspecting students to eject their floppy disks to prevent further infection. The resulting panic and chaos was akin to shouting "fire" in a crowded theater, leaving the pranksters with a deep regret for their actions.

The aftermath of their ill-advised prank was catastrophic for the unsuspecting students who had already powered off their computers to avoid the virus. Many lost their work, leading to a wave of fury and trouble for the pranksters. The facilities director, a stern figure, confronted the students, forcing them to write letters of apology. The incident, though deeply regretted, served as a valuable lesson for the author and his peers.

Fast forward to the present day, where former Anthropic employee Jacob Coxon boldly claims that the probability of AI killing all humans is 10% within the next decade. This chilling assertion has spread like wildfire, leaving the public fearful and questioning the implications of artificial intelligence. Coxon's claim, supported by Evan Hubinger, has become a mainstream concern that has permeated discussions on AI and its potential consequences.

The author argues that the fear surrounding AI is being fueled by experts who may not fully understand the inner workings of the technology they are discussing. In their rush to spread fear, these experts inadvertently sow doubt and uncertainty, often relying on speculation rather than concrete evidence. The author warns that this contagion of fear, like the one experienced during their college prank, can have a profound impact on society and can overshadow the actual risks and possibilities of AI development.

Drawing on their own expertise in computer science, the author emphasizes the importance of distinguishing between fear and reality. They argue that while the potential risks of AI are significant, the likelihood of AI causing the extinction of humanity is overstated at this point in time. The author reminds us that building computers is not just about intelligence, but also about understanding and engaging with the physical world.

The technology we develop must be grounded in reality and be subject to human reasoning and oversight.

In conclusion, the author cautions against allowing fear to dictate our actions and decisions regarding AI. By fostering a culture of responsible and informed dialogue, we can address the concerns surrounding AI without succumbing to the contagion of fear that threatens to overshadow the potential benefits of this transformative technology.