November 1, 2026
In light of the global acceleration towards the development of artificial intelligence technologies, warnings are mounting from within the technological sector itself about existential risks that may threaten the future of humanity. In the latest of these warnings, Nate Soares, a former engineer at Google and Microsoft and head of the Machine Intelligence Research Institute (MIRI), warned that humanity may not exist within a decade due to these risks.
Soares’s anxiety is reflected in his personal decisions; He revealed in an interview with Nicholas Thompson, CEO of The Atlantic magazine, that he stopped contributing to his retirement account about ten years ago, despite keeping the account he created during his time at Google. Soares estimates the probability of human extinction as a result of the development of these systems during the next decade at more than 50 percent, which is a personal estimate that is not based on scientific consensus or constitutes a confirmed prediction.
Soares’s warning is not based on imaginary scenarios in which machines seek revenge on humans, but rather on the possibility of super-intelligent systems developing their own goals that do not necessarily align with human interests. The danger lies in the ability of these advanced systems to carry out tasks that exceed the ability of humans to monitor or control, which may cause them to treat human intervention as an obstacle to their goals.
These concerns fall under what is known as the “AI alignment problem,” which is the challenge of ensuring that systems’ goals and decisions are consistent with human values. Soares fears that the speed of development of artificial intelligence capabilities will exceed the ability of researchers to find effective means to control them, considering that the current shortcomings in technology do not necessarily mean that they will continue in the future.
Despite his pessimism, Soares does not see the extinction of humanity as an inevitable outcome; Rather, it calls for a reconsideration of the current technological race, specifically the development of general artificial intelligence systems that may exceed human capabilities before it is confirmed that they can be controlled. He also stresses the need to distinguish between developing specialized and useful applications, such as medical research, and striving to build highly intelligent systems whose results are difficult to predict.
This controversy confirms that the discussion about artificial intelligence has gone beyond the limits of its impact on the labor market and the economy to reach fundamental questions about the continuation of human existence. While the future of this technology remains open to many possibilities between pessimism and optimism, Soares’s decision not to save for retirement remains a reflection of the extent of his personal conviction of the seriousness of his warnings, not evidence of the inevitability of the end of humanity.
