Recent calls to halt the development of artificial intelligence (AI) have been made by numerous medical professionals and health policy specialists from all over the world owing to the possibility that it could constitute a “existential threat” to humanity.
A group of five specialists determined that there are three ways that AI might wipe out humanity.
According to medical authorities, the AI Singularity will cause widespread destruction.
This declaration comes shortly after a letter calling for the suspension of AI research for similar reasons was signed by over a thousand scientists, including Yuval Noah Harari, Elon Musk, John Hopfield, and Rachel Branson from the Bulletin of Atomic Scientists.
It is important to note that the catastrophic risks that the health policy experts are referring to are not present with current text-based AI services like OpenAI’s ChatGPT.
The first danger is that AI could contribute to the expansion of authoritarian strategies like surveillance and disinformation. They make the point that authoritarian or totalitarian regimes may find it easier to rise to power and hold onto it as a result of AI’s ability to quickly clean, categorize, and analyze large volumes of personal data, including photos taken by the cameras that are becoming more and more commonplace.
The experts also expressed their concern that AI could hasten mass murder by expanding the use of Lethal Autonomous Weapon Systems (LAWS).
The medical experts also raised concern over the possibility of severe economic collapse and human suffering as a result of millions of people losing their employment due to AI-driven automation. The authors stated that estimates of the number of jobs that AI technology will eliminate over the next 10 years range from tens to hundreds of millions.
The team, led by a doctor affiliated with the United Nations University’s International Institute for Global Health, emphasized that its strongest cautions apply to a highly developed but yet hypothetical type of artificial intelligence called self-improving general-purpose AI (AGI).
AGI or AI Singularity
AGI would be able to execute a variety of activities that are currently only capable of being completed by humans, unlike present AI, which is unable to fully learn from and adapt its own code. Such an AGI “could theoretically learn how to surpass any constraints in its code and start developing its own objectives,” the medical professionals claim in their comments.
“There are scenarios where AGI could present a threat to humans, and possibly an existential threat,” the experts said in their statement, “by intentionally or unintentionally causing harm directly or indirectly, by attacking or subjugating humans or by disrupting the systems or using up resources we depend on.”
Although such a threat may be some decades away, the commentary by health policy experts, which was published in the British Medical Association journal BMJ Global Health, emphasized that there are several opportunities for the misuse of AI technology at this stage.
Governments will abuse AI, weaponize it
The authors discussed potential challenges to “democracy, liberty, and privacy” posed by governments and other large organizations using AI to automate complicated activities involving mass surveillance and digital disinformation campaigns.
They highlighted China’s Social Credit System as an illustration of a governmental instrument used to ‘manage and oppress’ human people when discussing mass monitoring.
The authors cautioned that in the case of digital disinformation, AI-driven information systems may exacerbate social division and conflict by causing a general breakdown of trust and undermining democracy. Deepfakes are an increasingly sophisticated method of distorting or misrepresenting reality.
AI destroying the concept of jobs
Last but not least, the researchers talked about the risk of poverty and misery brought on by the “widespread deployment of AI technology,” highlighting the potential for “tens to hundreds of millions” of people to lose their jobs.
“While there would be many benefits from ending work that is repetitive, dangerous and unpleasant,” these medical professionals wrote, “we already know that unemployment is strongly associated with adverse health outcomes and behavior.”
Also read : AGI: What is it? Perhaps sooner than you think, AI will become self-aware
The commentary’s writers cited a survey of AI society members in which 18% of participants held the view that the emergence of advanced AGI would be disastrous for humanity. Additionally, between 2040 and 2065, according to predictions made by half of the participants polled, AGI would be developed.