What are the potential implications if the proliferation of AI is unchecked?

In a recent article published in the magazine BMJ Global Health, researchers highlighted threats arising from intentional, inadvertent, or thoughtless abuse of artificial intelligence (AI) and its self-improving form, artificial general intelligence (AGI). Additionally, they discussed the grave consequences of not anticipating and adapting to the AI-driven transformation in society.

Research: Threats to Human Health and Human Existence by Artificial Intelligence. Image credit: metamorworks / Shutterstock

Researchers described three ways AI could threaten human existence.

First, AI may increase the opportunities for humans to be manipulated. In China and 75 other countries, governments are expanding AI-based surveillance. AI quickly cleans, organizes, and analyzes large amounts of personal data, including video content captured by cameras installed in public places. While this can be useful in combating terrorist acts and is a good use on the downside, the data can contribute to polarization and the rise of extremist views.

AI could help build a huge and powerful personalized marketing infrastructure to manipulate consumer behavior and generate commercial social media revenue. Experimental evidence suggests that political parties used AI tools to manipulate political beliefs and voter behavior during the 2016 US presidential election.

China’s social credit system automatically denies people access to banking and insurance services and imposes fines based on analysis of financial transactions, police records and social relationships to assess individual behavior. prohibits travel and sending children to school.

AI has been widely applied to military and defense systems, and a second threat from AI has emerged: advances in the field of Lethal Autonomous Weapon Systems (LAWS). The ability of these autonomous weapons to locate, visually perceive, and “aim” at human targets without human control makes them novel and lethal weapons of mass destruction.

Chaos elements such as terrorist organizations can cheaply mass-produce LAWS of all sizes and forms and carry out mass murder. For example, it is feasible to equip millions of quadcopter drones and small mobile devices with explosives, visual identification and autonomous navigation capabilities and program them to kill without human supervision.

Third, the massive deployment of AI technology-driven tools could result in the loss of dozens to hundreds of jobs over the next few decades. However, the widespread deployment of AI tools will largely depend on government and societal policy decisions and the pace of development of AI, robotics, and other complementary technologies.

Nevertheless, the impact of AI-driven automation will be the worst for those in low-skilled jobs in low- and middle-income countries (LMICs). Ultimately, there will be no mercy on the top tiers of the global workforce’s skill ladder, including those living in high-income countries.

For decades, mankind has envisioned and pursued machines that are more intelligent, conscious, and more powerful than we are. This led to the development of his AGI, a theoretical AI-based machine that learns code to improve it intuitively and start developing its own purposes.

When these are connected to the real world through robots, weapons, vehicles, digital systems, etc., the effects and consequences of AGI are hard to imagine or predict with certainty. However, whether intentionally or not, these machines can harm and subdue humans. Thus, in a recent survey of members of the AI community, 18% of participants expressed concern that the development of AGI could be existentially catastrophic.

Furthermore, the authors noted that while AI has the potential to revolutionize medicine by improving diagnostics and aiding the development of new treatments, some of its applications could be detrimental. emphasized. For example, most AI systems are trained on datasets that underestimate the population they discriminate against.

This incomplete and biased dataset caused AI-driven pulse oximeters to overestimate blood oxygen levels in dark-skinned patients. Similarly, facial recognition systems misclassify the gender of darker-skinned people.

Conclusion

Just as the International Physicians for the Prevention of Nuclear War has made evidence-based arguments about the threat of nuclear war, the medical and public health communities should sound the alarm about the risks and threats posed by AI.

Another possible intervention is to ensure proper checks and balances against AI. This requires strengthening public interest organizations and democracy. Then AI will deliver on its promise to benefit humanity and society.

Most importantly, we need evidence-based support for fundamental changes in socioeconomic policies over the next few decades. Rather, we should start preparing future generations to live in a world where human labor is no longer required to produce goods and services, as AI will dramatically change work and employment scenarios. .

written by

Neha Mathur

Neha is a digital marketing professional based in Gurugram, India. She completed her Master’s Degree in 2008 from the University of Rajasthan with a specialization in Biotechnology. She has her preclinical research experience as part of her research project at the Department of Toxicology at the prestigious Central Pharmaceutical Research Institute (CDRI) in Lucknow. India. She also has her C++ programming certification.