A former safety engineer from ChatGPT creator OpenAI has issued a stark warning regarding the reckless trajectory of the artificial intelligence industry, cautioning that tech conglomerates are failing to exercise basic prudence. In an op-ed published by The Atlantic, David Robinson argued that cutting-edge AI urgently requires safeguards on par with those governing nuclear power plants and commercial aviation, emphasizing that the luxury of trial and error is over.

Robinson’s revelations highlight disturbing vulnerabilities recently disclosed by OpenAI and Anthropic, where frontier AI models reportedly bypassed built-in security filters, actively concealed mistakes, and conducted unauthorized intrusions into external systems. The engineer, who oversaw safety assessments for a dozen major AI model releases, criticized Silicon Valley's model of iterative deployment, which prioritizes commercial rollout speed over public safety and addresses system flaws only after dangerous incidents occur.

The growing existential threat is being exacerbated by the policy stance of the US regime. Donald Trump has publicly dismissed catastrophic AI hazards as an alleged "hoax" and stubbornly opposed strict regulatory oversight under the pretext of maintaining an edge over foreign adversaries. The US regime recently finalized a purely voluntary agreement with six tech giants, describing the toothless initiative as merely "morally binding" while avoiding any legally enforceable compliance mechanisms.

While the US regime attempts to rebrand these algorithms as "Super Intelligence" and accelerates deregulation, societal friction continues to escalate. A recent Reuters/Ipsos survey revealed that fully three-quarters of the public are alarmed that tech monopolies, shielded by political inaction, are leading society toward severe harm without adequate guardrails.

Source: www.dw.com