Early Foundations and Milestones In his seminal 1950 paper Computing Machinery and Intelligence, renowned mathematician Alan Turing first posed the fundamental question, “Can machines think?” and then introduced the Imitation Game (now known as the Turing Test) as a way to gauge machine intelligence. However, the term "Artificial Intelligence" was coined in 1956 by John McCarthy during the Dartmouth Summer Research Project on Artificial Intelligence, a seminal event which set the stage for decades of exploration. Early research in the 1960s and 1970s focused on symbolic AI and logic-based programs (the era of “Good Old-Fashioned AI” (GOFAI)) that could prove mathematical theorems and solve puzzles. These periods of over-optimism were followed by “AI winters” when funding and interest waned, however, foundational work continued. By the 1980s, expert systems, i.e., rule-based programs encoding human expert knowledge, became popular. Yet, these systems were hard to maintain and required manual knowledge engineering. Emergence of Machine Learning Machine Learning (ML) enabled algorithms to learn autonomously from data without explicit programming. This shift in the 1990s was due to significant improvements in computing power, data storage, and connectivity. ML techniques like neural networks, decision trees, and support vector machines began outperforming rule-based systems in tasks like image classification and language translation. World Chess Champion Garry Kasparov’s 3½ - 2½ defeat to IBM Deep Blue in a six-game rematch in 1997 demonstrated the ability of machines to outperform humans in domains considered to require strategic reasoning. This inspired early exploration in financial applications as well. As a financial sector application, HNC Software’s Falcon system was screening two-thirds of all credit card transactions worldwide by the 1990s. ML application grew in the 2000s, and in finance, early ML models were deployed for specific, well-defined tasks: for instance, using neural networks, Banks also adopted ML for credit scoring beyond traditional logistic regression, using larger datasets to enhance prediction accuracy. The Deep Learning Revolution and Generative AI The 2010s saw further breakthroughs with the rise of deep learning, a subset of ML that involved multi-layered neural networks. A major milestone during this period was the release of the 2017 paper “Attention is All You Need” by researchers at Google, which introduced the Transformer architecture that laid the foundation for large language models (LLMs). The power of deep learning’s ability to carry out complex pattern recognition was validated by landmark achievements such as computers surpassing human accuracy in image recognition in 2012 and when Google DeepMind’s “AlphaGo” defeated Go champion Lee Sedol in 2016. Soon after, voice assistants became commonplace, and self-driving cars took to the roads. AI was no longer confined to labs; it began to surface in everyday products and services. In late 2022, Generative AI tools brought the power of advanced AI directly to the public. ChatGPT reached 100 million users in just two months after launch1, highlighting the unprecedented pace of adoption. Techniques such as retrieval-augmented generation (RAG), mixture-of-experts (MoE) architectures are further enhancing capabilities. From generating images to creating complex reports using a suite of agents, AI has moved beyond just being a niche technology to gradually reshaping the way we work. Unprecedented Progress As per the AI Index report 2025 by Stanford University, AI systems now outperform humans in nearly all tested domains. Complex reasoning is the last major frontier, but even here, the gap is narrowing quickly. Open-source AI models are rapidly catching up to closed models, narrowing the gap from 8% to just 1.7%. Smaller models are also showing significant gains in efficiency and capability. The year 2024 marked a shift in national strategy with record public investments: India ($1.25 billion), France ($117 billion), Canada ($2.40 billion), China ($47.50 billion) and Saudi Arabia.