AI is here but future is still human

Prime Minister Narendra Modi’s call for a global governance framework for AI at the Paris Summit, the United States breaking ranks arguing that it will result in over-regulation, and the recent launch of the Chinese Large Language Model ‘Deep Seek’ have together reignited the man-versus-machine debate. In 1951, Alan Turing prophesied: “Once the machine thinking method had started… At some stage… we should have expected the machines to take control.” Now is a good time to ask what AI is trying to accomplish. It is about making machines intelligent, but what does that entail? AI, at its core, remains a sophisticated tool for pattern recognition and data processing, not an autonomous form of intelligence. Humans are intelligent to the extent that our actions can be expected to achieve our goals. All other aspects of intelligence, such as perception, thinking, learning, inventing, listening, and so on, can be understood in terms of their contributions to our ability to act effectively. Since the inception of AI, intelligence in machines has been defined in the same way: machines are intelligent to the extent that their actions can be expected to achieve their objectives. However, because machines unlike humans lack independent objectives, we assign them objectives to achieve.

Predicting that machine intelligence will surpass human intelligence at some point is counter-intuitive. Machines do not possess intelligence. If we consider intelligence as computation, AI has undoubtedly made significant advances over the last seven decades. We have also created learnable machines that improve their ability to achieve objectives through training. Over the last decade, deep learning systems have learned to recognise both human speech and images, as well as to translate between different languages. This progress has been made using simple, narrow, application-specific algorithms. The goal of AI has always been general-purpose AI: machines that can quickly learn to perform well across the entire range of tasks that humans can do. However, we are a long way from achieving general-purpose AI. It is important to understand what AI can and cannot do. AI, as it exists today, excels in automation, data analysis, and predictive modelling. Machine learning algorithms process vast amounts of data to identify patterns, automate repetitive tasks, and improve efficiency.

However, AI’s strengths are also its limitations. Unlike human intelligence, AI lacks adaptability, creativity, and an understanding of abstract concepts. Intrinsic to AI is that it is bound by data and operates within the confines of its training data. Every AI model, no matter how advanced, relies on human-generated inputs. It cannot independently acquire knowledge beyond what it has been exposed to, making it incapable of genuine innovation, it can only extrapolate from existing data. AI lacks general intelligence because consciousness is absent. AI, no matter how advanced, is ultimately a computational system running algorithms – its outputs are derived from statistical correlations, not genuine understanding, or self-awareness. Even if AI were to pass the Turing Test perfectly, it would only be simulating intelligence, not experiencing it. Sentience requires consciousness, emotions, and self-awareness – qualities that emerge from biological, not computational processes.

Overstating the powers

AI does not know that human beings exist at all or that they have minds. From the algorithm’s point of view, each person is simply a click history. There is empirical evidence on the limits of AI and leading AI researchers acknowledge this. Geoffrey Hinton, a pioneer of deep learning, has pointed out that AI models struggle with reasoning and understanding causality. Yann Le Cun, another AI visionary, emphasises that machines do not possess common sense or the ability to independently navigate complex, real-world problems. Stuart Russel suggests that the world will likely never see a general-purpose AI. We must learn from history to understand why AI will follow the pattern of past technologies. The belief that AI will replace human intelligence mirrors historical trends of technological hype. The industrial revolution, automation, and the advent of computers initially caused concerns about widespread job loss, but they ultimately enhanced productivity and led to new fields of employment. Early robotics was expected to eliminate entire job sectors, but instead, led to job transformation rather than elimination. History demonstrates that new technologies often create more opportunities than they eliminate. AI will likely follow the same trajectory – reshaping, not replacing, human roles.

So, how must we envision AI and the future world? India must push ahead with developing Foundational AI models and emerge as a leader in AI. The future lies in human-AI collaboration, where AI handles repetitive, data-intensive tasks, allowing humans to focus on strategic thinking, creativity, and ethical considerations. A simple axiom works well to drive home the point: ‘Technology does not solve problems, human beings do’.

Going back to read Adam Smith, who’s widely reviled as ‘The Apostle of Greed’, students are pleasantly surprised at what Smith says at the beginning of his first book: ’It is so obvious to everyone that each of us cares deeply about other people that it hardly merits saying it, but I’m going to say it anyway.’ If AI systems are going to be making decisions on behalf of the human race, what does that mean?; that’s something scientists and philosophers have grappled with in the past few decades. I take an optimistic view of AI and the future: there are areas that will not be automated, either because we don’t want to or because humans are just intrinsically better. Keynes called it ‘perfecting the art of life’. AI or no AI, humanity will be faced with man’s eternal quest: how to live agreeably, wisely, and well. Those who cultivate the art of life better will be much more successful in this future world.

(The writer is Director, School of Social Sciences, Ramaiah University
of Applied Sciences)