Michael Nielsen – Why aliens will have a different tech stack than us
From Dwarkesh Patel
Michael Nielsen•Pioneer in Quantum Computing, Research Fellow, Astera Institute
Executive Summary
The history of scientific discovery is far more complex and non-linear than simplified narratives suggest; for example, the Michelson-Morley experiment did not directly lead to Einstein's theory of special relativity, nor did it immediately falsify the existence of an 'ether'.
Scientific progress is not necessarily subject to diminishing returns or a 'low-hanging fruit' problem.
New fields, like computer science or deep learning, can emerge from unexpected domains, creating entirely new frontiers with abundant opportunities for rapid breakthroughs.
The trajectory of scientific and technological development is highly path-dependent.
Different starting conditions or cultural biases could lead civilizations to explore vastly different parts of the 'technology tree,' implying that our current scientific understanding is just one of many possibilities.
In the age of AI, there is a critical distinction between the deep, integrated understanding gained from high-stakes, demanding intellectual work (like writing a textbook) and the potentially superficial knowledge acquired through easy interactions with tools like LLMs.
12 quotes
Concerns Raised
Simplified narratives of scientific history misrepresent how discovery actually works.
The ease of using AI tools like LLMs may encourage superficial learning and hinder the deep thinking required for breakthroughs.
The increasing capital intensity of frontier research (e.g., deep learning) could become a bottleneck to progress.
Opportunities Identified
New scientific fields can emerge from unexpected, esoteric questions, creating vast new frontiers for discovery.
The path-dependent nature of technology implies that vast, unexplored domains of knowledge exist.
Understanding the true, messy history of science can provide better models for fostering future innovation.