AI must walk the ethical path

Anthropic, now more in the news for its chatbot, Claude, recently held a summit in which the leaders of the Christian faith and different philosophical traditions were invited for their advice in developing “Claude’s moral and spiritual responses to complex and unpredictable ethical queries” as its chief executive, Dario Amodei, felt the need for infusing in it a ‘moral character’. During deliberations it was even asked whether Claude could be called a ‘child of God’, with a ‘spiritual value’ beyond that of a simple machine.

A participating Catholic priest clarified that it is more about ‘building ethical thinking into the machine’. Many believe that as AI transitions from experimental software into a core pillar of daily life, tech leaders face unprecedented pressures to account for their societal impact. Of late, many prominent scientists, including computer scientists, tech entrepreneurs, investors and futurists have been using words like ’godlike’, ‘prophecy’ and ‘apocalypse’ – to describe the evolving power of AI. These words, usually reserved for the divine, signify a profound cultural and philosophical shift in thinking in the hitherto secular tech world.

Advertisement

Nick Bostrom, the founding director of the Future of Humanity Institute, Oxford, in his 2014 book ‘Super-intelligence: Paths, Dangers and Strategies’, envisioned an impending ‘intelligence explosion’, a scenario in which an upgradable AI enters a positive feedback loop, rapidly undergoing re c ursive self-improvement, leading to its jump in intelligence from human-level to an unimaginable super-intelligence. Computer scientist and futurist Ray Kurzweil, in his 2024 book, ‘The Singularity Is Nearer: When We Merge with AI’, presented a futuristic p ersp ective around ‘technological singularity’, a theorised tipping point where technology vastly exceeds human cognitive capacity, and the possibility of humans merging with machines in a gradual evolutionary transition through direct technological integration by 2045 is envisioned.

OpenAI CEO Sam Altman, called it a “magic intelligence in the sky’ which transports ‘humans to the post-biological era, vastly extending lifespans, eradicating diseases, and heralding an era of radical abundance.” However, there are many sceptics who argue that AI will not overtake human intelligence anytime soonif ever, and that the current AI models can be excellent at pattern recognition and data synthesis, but they fundamentally lack common sense, consciousness, and the flexible adaptability of the human brain”. Gary Marcus, a cognitive scientist and AI entrepreneur, a critic of the ‘scaling up’ hypotheses, asserted that the “idea of making the current AI models bigger by feeding them with more computation data will lead to artificial general intelligence (AGI), is simply misleading” as “human intelligence is deeply intertwined with emotions, social context, and subjective experiences”.

“The true AGI requires a physical and bodily interaction with the real world,” underscored Rodney Brooks, a roboticist and former director of the MIT AI Lab. There are also techno-optimists who talk about a ‘gentle singularity’, where humans and machines slowly and safely integrate into a cooperative ecosystem, dismissing the idea of ‘an imminent and malevolent robot takeover’. Many are also apprehensive that by “mystifying algorithms as omniscient or sentient deities, ignoring their algorithmic limitations, will only obscure human accountability for AI bias and ethical risks”. MIT physicist Max Tegmark and research engineers at the Distributed AI Research Institute (DAIR) dubbed such attempts as “dangerous, hubristic pseudo-religious pursuits that distract the world from immediate, material risks like corporate monopolies, data theft, and algorithmic bias”.

While institutional religions called for ‘drawing a firm line between machine calculations and actual divinity’, Pope Leo XIV, in his recent encyclical titled ‘Magnifica Humanitas’ or ‘Magnificent Humanity’ critiqued Silicon Valley’s trans-humanisation of AI, and ‘building God’ through AI. Amid concerns about humanity’s existential crisis vis-a-vis AI, eminent physicists like Stephen Hawking cautioned that “building a true AGI poses a severe challenge to human survival, and its super-intelligent version (ASI) may either eliminate or make humanity subservient to it”. Geoffrey Hinton, called the ‘Godfather of AI’, warned that a tool that has been embedded with “artificial intelligence may eventually eclipse humanity’s and disrupt traditional religious b eliefs and human exceptionalism”.

Peter Thiel, the tech entrepreneur, forewarned that “since both, the AGI and ASI, exceed the traditional boundaries of science and engineering, their uncontrolled rise could lead to severe societal collapse and biological threats”. Proponents of trans-humanism like Julian Huxley, Nick Bostrom, David Pearce, and James Hughes have advocated for the path of pursuing radical cognitive and moral improvements through a process of biological and digital symbiosis. Historian and philosopher Yuval Noah Harari in his essay ‘Homo Deus’, exhorted that the ‘human being has to evolve in parallel with machines’, while Elon Musk’s Neuralink project envisages a synergy between human intelligence and AI through implantable brain computer interfaces (BCI).

In 2024, the Global Future Councils (WEF), presented a roadmap for AI guardrails aligned to human values and ethical principles, and suggested continuous engagement among governments, businesses, civil society and individuals in transcribing abstract ethical principles into practical technical guidelines, tailored to specific cultural, legal and societal contexts, to make sure that AI systems remain auditable and transparent. Now, it is time for collective global action to ensure that technology serves humanity’s best interests, and shared values.

(The writer is former Director-General, Doordarshan & AIR, and former Press Secretary to the President)