This article presents an extended dialogue between Angshuman Guha, a researcher with hands-on experience in neural networks dating back to 1993, and Google’s large language model (LLM), Gemini. It explores a set of closely connected questions at the intersection of AI and cognitive science. Are modern LLMs genuine reasoning systems, or are they sophisticated Stochastic Parrots that recombine language without understanding? Do the so-called “emergent abilities” of large models reflect a real shift in machine capability, or are they artifacts of how we measure performance? And perhaps most fundamentally: can a system trained entirely on human-generated text be said to produce anything like “original thought”?
Most intelligent tasks we perform in our lives, we learn those skills through examples rather than being told step-by-step how to do them. For example, no one told us how to recognize the numbers, but showed us many examples, and our minds figured out some subconscious rules to distinguish a “1” from a “2” and all the other digits. Scientists quickly realized that if machines must do complex tasks, they cannot be taught algorithmically, with step-by-step instructions, as we ourselves may not know these logical steps, but rather by showing many examples of the correct behavior. Although that was the holy grail of AI, machine learning was a hard task.
DeepSeek’s rapid ascent in artificial intelligence (AI) has disrupted the sector, challenging industry norms and intensifying global competition. This analysis by the Editorial Board highlights the company’s breakthroughs, notably its cost-effective, high-performance AI model "R1" and open-source large language models (LLMs) "DeepSeek-V3" and "DeepSeek-R1".
In our relentless quest to understand the Universe, we often overlook the most intricate and profound object we have: the human brain. Weighing just 1.5 kilograms, this soft, wet tissue is a marvel of complexity, capable of solving problems, creating art, and exploring the depths of existence. Throughout history, we have sought to unravel the mysteries of thought and reasoning, constantly drawing parallels between our minds and the advanced technologies of our time. From clockwork mechanisms to modern computers, these analogies reflect our evolving understanding of intelligence and cognition. Written by one of the leading experts on AI, this article takes a journey through the lens of chess—a game long considered a benchmark for intelligence—tracing the development of Artificial Intelligence over the past 75 years. As we examine how machines have learned to play this strategic game, we will uncover the challenges and breakthroughs that have shaped the AI field, revealing the deep connection between our understanding of the mind and the machines we create.
