New Dark Age | James Bridle

Summary of: New Dark Age: Technology and the End of the Future
By: James Bridle


Venture into the depths of James Bridle’s ‘New Dark Age: Technology and the End of the Future’, a thought-provoking exploration of the entwined relationship between technology and modern society. This summary illuminates the fascinating origins of computation, the consequences of our digital habits, and the pitfalls of an ever-accelerating digital world. Delve into the realms of big data, climate change, artificial intelligence, surveillance, conspiracy theories, and the frightening phenomena arising from the marriage of capitalism and technology. Prepare yourself for a gripping journey that challenges your perception of the digital landscape and its implications on the environment and our future.

The Military and Weather Machines

The history of modern computation is tied to the military’s need to predict and control the weather. Mathematician Lewis Fry Richardson’s thought experiment of a “weather computer” during World War I was the precursor to the first computers developed during World War II. These early machines were used to simulate the effects of bombs and missiles under different weather conditions. However, the true functions of these computers were often concealed and their oversimplified view of reality and bad data caused serious errors, as seen with the SAGE network mistaking birds for a Soviet bomber fleet.

Climate Change and Technology

The impact of climate change on both ancient and modern technologies is explored in this summary. Climate change is a hyperobject, making it difficult for us to comprehend fully. The Syrian conflict is an example of how climate change led to a refugee crisis, and how agriculture and other ancient technologies are vulnerable to changing weather patterns. However, even new technologies like the internet are vulnerable as data centers require extensive physical infrastructure highly susceptible to extreme weather conditions. The world’s digital culture becomes faster, making it more energy-consuming to maintain data centers. More energy is used to store and transmit data, and climate change might hinder our ability to integrate all this information effectively.

The Fallacy of Big Data

Moore’s law predicts that computing power doubles every two years. However, this doesn’t necessarily mean that technology improves our lives. The belief that more computation and data leads to a better understanding of the world has led to a research system that values automated testing generating tons of data. But, this approach isn’t effective, giving rise to “Eroom’s law,” where drug research has decreased by half every nine years. The same fallacy encourages quantity over quality in science, leading to the replication crisis. Although scientific research gathers more data about the world, the current overflow of information negatively affects our ability to process it and slows scientific discovery.

The Unseen Digital Economy

Slough, a small town outside London, houses the physical base of our digital world, including LD4 – London Stock Exchange’s data servers. High-speed fibre optic cables transport financial data across the globe, enabling high-frequency trading. Algorithms and bots monitoring prices react almost instantly, making flash crashes common in the digital finance world. Meanwhile, Amazon uses fleets of robots to store, sort, and pick out products, with humans guided and monitored maximising efficiency. However, little attention is paid to the social security system to replace full-time employment. Technology concentrates power in the hands of the few rather than being a great equalizer.

Dangers of Machine Learning

Machine learning has its limitations and can be inherently biased due to inherent human biases. The book explores these pitfalls through intriguing anecdotes, highlighting the importance of ethical considerations when training machines to think. For instance, a computer designed to recognize camouflaged tanks failed miserably in the field despite being trained with multiple images. It turned out that it had only learned to differentiate between a sunny and a cloudy day instead of identifying tanks. The mysterious nature of a machine’s mind can justify the conclusions, including controversial and dangerous ones. The idea that algorithms and computation are unbiased is misplaced since the machines tend to be trained with data, and the only data we have is of our past, which is full of violence, injustice and racism. The story of Nikon Coolpix S630, which repeatedly displayed the error message “Did someone summary part?” when Asian-Americans tried to take family photos, further highlights the limitations of machine learning and the ethical considerations required in training them.

Want to read the full book summary?

Leave a Reply

Your email address will not be published. Required fields are marked *

Fill out this field
Fill out this field
Please enter a valid email address.
You need to agree with the terms to proceed