Stanford University report says ‘incidents and controversies’ related to AI have elevated 26 occasions in a decade.

More than one-third of researchers consider synthetic intelligence (AI) might result in “nuclear-level catastrophe”, in keeping with a Stanford University survey, underscoring considerations within the sector in regards to the dangers posed by the quickly advancing expertise.

The survey is among the many findings highlighted within the 2023 AI Index Report, launched by the Stanford Institute for Human-Centered Artificial Intelligence, which explores the newest developments, dangers and alternatives within the burgeoning discipline of AI.

“These systems demonstrate capabilities in question answering, and the generation of text, image, and code unimagined a decade ago, and they outperform the state of the art on many benchmarks, old and new,” the report’s authors say.

“However, they are prone to hallucination, routinely biased, and can be tricked into serving nefarious aims, highlighting the complicated ethical challenges associated with their deployment.”

The report, which was launched earlier this month, comes amid rising requires regulation of AI following controversies starting from a chatbot-linked suicide to deepfake movies of Ukrainian President Volodymyr Zelenskyy showing to give up to invading Russian forces.

Last month, Elon Musk and Apple co-founder Steve Wozniak have been amongst 1,300 signatories of an open letter calling for a six-month pause on coaching AI methods past the extent of Open AI’s chatbot GPT-4 as “powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable”.

In the survey highlighted within the 2023 AI Index Report, 36 % of researchers mentioned AI-made selections might result in a nuclear-level disaster, whereas 73 % mentioned they might quickly result in “revolutionary societal change”.

The survey heard from 327 specialists in pure language processing, a department of laptop science key to the event of chatbots like GPT-4, between May and June final 12 months, earlier than the discharge of Open AI’s ChatGPT in November took the tech world by storm.

In an IPSOS ballot of most of the people, which was additionally highlighted within the index, Americans appeared particularly cautious of AI, with solely 35 % agreeing that “products and services using AI had more benefits than drawbacks”, in contrast with 78 % of Chinese respondents, 76 % of Saudi Arabian respondents, and 71 % of Indian respondents.

The Stanford report additionally famous that the variety of “incidents and controversies” related to AI had elevated 26 occasions over the previous decade.

Government strikes to manage and management AI are gaining floor.

China’s Cyberspace Administration this week introduced draft rules for generative AI, the expertise behind GPT-4 and home rivals like Alibaba’s Tongyi Qianwen and Baidu’s ERNIE, to make sure the expertise adheres to the “core value of socialism” and doesn’t undermine the federal government.

The European Union has proposed the “Artificial Intelligence Act” to manipulate which sorts of AI are acceptable to be used and which must be banned.

US public wariness about AI has but to translate into federal rules, however the Biden administration this week introduced the launch of public consultations on how to make sure that “AI systems are legal, effective, ethical, safe, and otherwise trustworthy”.