The nuclear race and Cold War that followed World War II were brutal and took a significant toll on the world, leading to a major technological competition between the U.S. and Russia. On one hand, this era gave birth to cultural myths like '007'. On the other, it spurred the creation of technologies like the internet and the Global Positioning System (GPS), which today help Domino's deliver pizza in less than 30 minutes and Zepto deliver my vegetables each morning.
Similarly, a monumental project, akin to a moonshot, is underway in AI, amidst a geopolitical race to AGI (Artificial General Intelligence) between the U.S. and China. Project Stargate is an ambitious initiative with a $100 billion investment in computing that marks a significant collaboration between two of the most influential entities in artificial intelligence: Microsoft and OpenAI. This partnership aims to push the boundaries of AI research and development with a clear and audacious goal: to achieve artificial general intelligence (AGI) by the year 2028, or perhaps earlier. Ilya Sutskever, co-founder of OpenAI, recently left the organisation and raised $1 billion for his months-old startup, interestingly named Safe Superintelligence (SSI), which aims to build safe AGI.
Apple Intelligence In iPhone & Mainstreaming AI
As demonstrated by the rise and fall of NVIDIA shares, this journey is fraught with the hype of massive expectations. But the AI ship is clearly advancing, propelled by three major waves. The first wave was when we were able to feed the entire internet's knowledge into training GPT models, which produced ChatGPT by OpenAI, enabling highly intelligent chatbots accessible to everyone. Impressive results from o1-preview (the latest breakthrough model from OpenAI) promise to bring reasoning and mathematical capabilities to everyone on a massive scale.
Apple's recent launch of the iPhone 16, equipped with hardware capable of running AI (Apple Intelligence) directly on your phone, signals the mainstreaming of this technology wave. Within a year, hundreds of millions of AI users will leverage this intelligence in their daily interactions - from using Siri to fixing their pictures, correcting their grammar in emails, and understanding their surroundings. The age of the "Intelligence Economy" is kickstarting.
AI Agents In Enterprises
The second wave involves building AI agents that are beginning to automate processes within enterprises, starting with call centre agents and AI tools that enhance coding capabilities. These AI agents will work collaboratively with humans to automate or displace close to 85 million jobs by 2025 (according to the World Economic Forum). They are not at an expert level right now, but they are improving daily, working round the clock with the ability to process massive token context windows - far larger than many novels - at once. Already, some of the best programmers are leveraging coding AI agents from companies like Cursor and powerful tools like GitHub Copilot. The year 2025 is expected to be a breakthrough in the enterprise integration of AI agents at scale. We will see AI customer agents, creative agents, data agents, and security agents transforming initially horizontal enterprise workflows, followed by vertical ones.
Race to AGI
The third wave is when these AI agents will not only change the way we work but also transform how we live and function in the physical world. Physical-digital AI agents, building on the previous two waves, which are constantly learning in simulations and applying that learning in real-world settings, represent this development. The understanding of concepts like gravity through robotics, as suggested by researchers like Pulkit Agrawal at MIT and Deepak Pathak at CMU, could play a significant role in the path toward AGI. That's why embodied cognition and physical intelligence at scale will be critical on the way to AGI.
To explore this acceleration to AGI safely, we might need contained environments like AI cities, where we can run experiments at the scale of automating 40-50% of our workflows and assess whether these are sustainable and safe for society. These AI cities will have AI computing and high-bandwidth connectivity infrastructure integrated with physical infrastructure to enable robots and intelligent machines to understand our world and function seamlessly and safely. They will also contain sandboxes built to test AI's safe applications in healthcare, agriculture, logistics, manufacturing, and mining, allowing for sustainable adoption. By building STPIs (Software Technology Parks), India was able to gain the maximum from the IT wave. By creating AI cities to establish a safe environment for building, deploying, and testing AI at scale, India can capture a lion's share of the new 97 million jobs and $15.7 trillion of new economic value that AI is expected to create by 2030 (according to PwC). Unleashing intelligence at scale without reimagining our society in the form of AI cities could have dramatic and unforeseen consequences, both social and economic.
Doing & Building
The neuroplasticity of the human brain is extraordinary, and given a fair chance, it will rapidly adapt to the massive new tools that we are building with AI. However, our current education system, with its flawed focus on memory-driven exams easily passable by AI, is creating a massive trap and drag on our demographic dividend.
China already has 392 robots per 10,000 employees, with a target to reach 500 by 2025. What we really need to fear is not the flooding of cheap humanoids from China, but rather the massive numbers of "human bots" - humans raised and trained to master outdated exams in pursuit of paper degrees - who may be unemployable in the future. Doing is the new learning, and we should be aiming for the art of co-building things with AI. We must teach our demographic dividend of 900 million to "build" the Viksit Bharat future in collaboration with AI in these AI cities. It might be worthwhile to encourage our 300 million young children to explore their inner selves and learn to build their dreams with their unique purpose and passion, given that "Androids are yet to dream of electric sheep".
[Umakant Soni - Umakant Soni is chairman of AIfoundry and co-founder of ARTPARK (AI & Robotics Technology Park) and a passionate advocate for inclusive development]