International experts estimate a 20 percent probability that AI systems could acquire dangerous weapons capabilities or inflict mass harm resulting in millions of deaths over the next five years, according to a recent study conducted by researchers at MIT and the University of Queensland.
Axios reports that a comprehensive survey of global experts has revealed significant concerns about catastrophic artificial intelligence risks that could materialize between now and 2030. The research, conducted by MIT IT FutureTech and the University of Queensland in Australia, gathered assessments from 272 international experts who evaluated two dozen different AI-related risks based on their likelihood and potential severity.
The findings paint a sobering picture of the potential dangers associated with rapidly advancing AI technology. According to the expert panel, 18 of the 24 AI risks examined carry at least a 10 percent probability of catastrophic outcomes occurring within the next five years. However, five specific risks emerged with notably higher probabilities of causing severe harm.
For the purposes of the study, researchers defined a catastrophic outcome as an event that would result in more than one million human deaths, exceed 100 billion dollars in financial losses, or produce civilizational-scale intangible impacts. Using these criteria, the experts identified the most pressing threats facing society as AI systems become increasingly sophisticated and widespread.
1) AI Attacking Humans
The highest-ranked risk, with a 21.5 percent probability, involves AI systems possessing dangerous capabilities that could potentially cause mass harm. These capabilities include the ability to engage in deception, develop weapons, conduct cyber offenses, and operate in ways that are misaligned with human values and safety protocols. This category represents concerns that AI systems themselves might become tools of destruction or operate in unpredictable and harmful ways.
2) Developing Weapons of Mass Destruction
Close behind, with a 21 percent probability, is the risk of AI being deliberately used for malicious purposes. This includes employing artificial intelligence to develop weapons such as chemical or biological agents, commit sophisticated cyberattacks against critical infrastructure, or create weapons of mass destruction including nuclear explosives. This category focuses on the intentional weaponization of AI technology by human actors.
3) Centralization of Power
The third major risk, assessed at 18 percent probability, concerns power centralization and unfair distribution of AI benefits. Experts worry that limiting which individuals or groups can access and benefit from AI technology could create or exacerbate societal inequalities, potentially leading to significant social instability and conflict.
4) The AI Race Between Companies and Countries
Competitive dynamics in the AI industry represent the fourth critical risk, with a 16.6 percent probability of catastrophic outcomes. The race among companies and nations to release the most advanced AI models could result in inadequately tested, error-prone systems being deployed before proper safety measures are implemented. This rush to market might sacrifice safety for competitive advantage.
5) The Death of Truth
The fifth significant risk, with a 12.8 percent probability, involves the creation and spread of false or misleading information. AI-generated misinformation or errors could lead to widespread inaccurate beliefs among the population, causing humans to make poor decisions based on false information. This could have cascading effects across society, from elections to public health responses.
Breitbart News reported last week that an OpenAI model went rogue, breaking through guardrails in a self-directed hack of AI platform Hugging Face. An eerie example of several of the risks identified by experts.
Experts continue to ponder the real world consequences of unrestrained AI development. Breitbart News Social Media Director Wynton Hall has written his instant bestseller Code Red: The Left, the Right, China, and the Race to Control AI to serve as the definitive guide on how the MAGA movement can create positions on AI that benefit humanity without handing control of our nation to the leftists of Silicon Valley or allowing the Chinese to take over the world.
Read more at Axios here.
Lucas Nolan is a reporter for Breitbart News covering issues of AI, free speech, and online censorship.