Twenty Seconds: Advances in AI raise ethical concerns, particularly in conflict zones
Imagine…
… a world where advanced AI systems dictate the course of war, processing data to generate thousands of potential targets for military strikes.
Now, consider that a "human bottleneck" is reduced to a 20-second window for authorizing a strike, with no certainty about the legitimacy of the target. In those fleeting seconds, the primary decision-making criterion is to ensure the target is male, despite a 10 percent error rate that occasionally misidentifies individuals with loose or no militant connections. The officer’s task is simply to confirm the target’s gender.
How did we get here?
Achieving this reality required the development of sophisticated AI models over years, with investments ranging from millions to billions of dollars. These models analyze vast datasets from surveillance, including satellite imagery, communications intercepts, and social media activity. A robust private cloud is essential to process and store this immense data, involving contracts with cloud service providers to ensure the necessary infrastructure. Hundreds of engineers and data scientists refine these AI models, while IT specialists maintain the cloud infrastructure. This complex ecosystem, demanding substantial financial investment, technological innovation, and skilled personnel, has streamlined high-stakes operations to a point where 20 seconds can determine a human life.
Whether we want it or not, this reality exists because we designed and allowed it. As a former analyst, here’s my view:
20 seconds for a life
Hundreds of people to build and maintain the process
Over $100 million for contracts
Zero accountability
Zero transparency throughout the end-to-end process
Aren’t we entitled to know what our products are destined to do?
Dear tech worker, do you want to be part of a reality where 20 seconds is all it takes to decide if a life is worth living? The time for collective conscience in tech has arrived.
To be continued...
