The Center for AI Safety, a non-profit organization, has recently published a statement regarding the potential risks associated with artificial intelligence (AI). Signed by executives from OpenAI and Google DeepMind, as well as university professors in machine learning, computer science, and philosophy, the statement indicates that mitigating the risk of AI-induced extinction should be a global priority, akin to other societal-scale risks such as pandemics and nuclear war.
Despite these concerning warnings, some believe that AI presents a unique opportunity to optimize food and agriculture production. Puneet Mishra, a researcher at Wageningen University & Research (WUR), explains that AI technology has only recently begun to show its true potential. Automated machine learning founded on AI is now capable of making decisions and issuing commands for future operations.
Traditionally, data availability has been the biggest bottleneck for AI adoption, as data is often stored in different silos and formats. However, the situation is changing as more effort is put into combining or generating data specifically for AI purposes. As a result, the use of AI in food and agriculture is becoming more feasible.
AI technology is expected to reduce energy usage and labor costs in agricultural settings, such as in WUR’s Autonomous Greenhouses project which uses AI to train robots operating within the autonomous farms. While the robots cannot be trained for every situation, researchers simulate greenhouse environments and plant structures to allow the robot to identify a greater variety of crops and situations.
AI can also optimize food manufacturing, as illustrated in WUR’s research into chocolate production. Sensors are used to generate data which is then combined with AI to achieve quality and consistency targets.
To keep potential AI threats at bay, regulation is necessary. This will ensure that AI does not generate non-sensical information disguised as facts, and that data used to train AI is transparent and unbiased. This will give human beings full control of a super helpful tool that can currently be perceived as “out of control.”