By Dhurjati Mukherjee
For the last few years, analysts were afraid about the induction of Artificial Intelligence (AI) as it was perceived to be a threat to employment, particularly in a labour-intensive country such as India. Various propositions were advanced about its positive effects in the technology sphere and some even observed that it may offset the loss to jobs, directly or indirectly. But in recent days, AI has captured the headlines as it is stated to pose an existential threat to humanity and human life.
Last week at the UN Security Council, AI and international security was discussed after industry leaders, at a meeting convened by France, calling for a coordinated slowdown of the AI systems where representatives from OpenAI, Anthropic and Chinese firms, DeepSeek and others participated in the deliberations.
The head of AI company Anthropic Dario Amodei told the 15-member council that if managed poorly, AI could be a risk to humanity as a whole. Even Altman of OpenAI echoed the same viewpoint.
The discussions were undertaken at an opportune time in view of the cry the world over of slowing down development of AI capabilities because of an imminent risk of recursive self-improvement (RSI) where it escapes control and does very bad things.
Anthropic came out with a 154-page report apart from one by OpenAI and Amodei about the impending threat to our lives. The proposals affirmed the need for widely mandated safety bars for continued development of AI to be enforced by a network of third-party auditors, government agencies and international bodies.
If one delves deep into the matter, the warnings that have hit the headlines are not new. AI pioneers Geoffrey Hinton and Yoshua Bengio, whose work helped make modern AI possible, had put the risk around 10 and 20 per cent. But now the warnings are coming from people inside the frontier labs, working directly on the systems being designed.
It is now time to make AI democratic as the most important decisions cannot be made by labs at San Francisco, as rightly pointed out by Altman at the Security Council. “They must be shaped through democratic processes and by governments accountable to the people they serve,” he stated.
The need for international cooperation was highlighted at the Council meeting as Altman observed “no company, no nation can handle this alone”.
Thus, safety in the everyday uses of AI in defence, medicine and hospitals, education and law enforcement agencies are vital and need to be regulated in view of the dystopian threat.
A slowdown as has generally been agreed upon would, however, leave the central problem unresolved unless special efforts are made by the companies jointly with the governments.
Meanwhile, safety testing has started and Britain’s AI Security Institute now has a team working on this. However, very little is known about the safety techniques being experimented with and how artificial intelligence will actually benefit the human race.
Safety standards must be undertaken by most governments like the EU AI Act which requires the most powerful models to be assessed for systemic risks and gives power to regulators to demand safeguards. But this alone may not be sufficient in the long run as an international panel of experts need to be set up to analyse and formulate the risks.
In keeping with the ongoing debate over the induction of AI, though the warnings of extreme technical risks cannot be undermined as history shows that humans are not the strongest or smartest species, one has to adopt a wait-and-watch policy.
Strangely the rise of AI has to be taken in our stride keeping the basic core of human existence. When AI handles the mundane machinery of survival, what do we do with ourselves?
The transition to higher levels of technology has to be in tune with our life and living. We cannot be naïve about the social and economic displacement coming our way for which there is a need to take proper steps after critically analysing the whole issue in consultation with experts, technology leaders and scientists from across the globe.
