Click here to view the book on Amazon:

Key Concept

The key concept is that artificial intelligence is going to become more naturally integrated to our biological systems. When this happens, we will expand & augment our brain capacity.

This sets the stage for us to increase our brains computational power millions of times over.

Who this book is for

This book doesn't solely focus on AI. It actually talks a lot about the meta trends of technology and society as a whole and how things have been improving.

If you have ever read the book Factfulness by Hans Rosling, this book is similar in that it talks about how things are improving and how we are on the cusp of a technological revolution.

Buy this book if...

• If you are interested in some abstract and crazy concepts about the future of Technology
• If you enjoy the longer perspective on things, e.g seeing how things have improved over the last 100 years
• You want to hear some theories about humans could live for 1,000 years
• You don't mind reading conjecture and some guesses based on macro trends
• You don't mind a book that jumps around a bit into different topics and areas. This book covers 3D printing, building, nanobots.

Don't buy this book if...

• You are looking for a step by step path of how this is going to happen (this book is more about the macro trends)
• Conjecture based books annoy you
• You want a book that is dedicated to just AI and the technical aspects of it. The majority of this book is about historical trends and how we are on the cusp of a technological revolution.

On the whole, I'd say this is more aimed at executives and high level thinkers, rather than those down in the weeds.

Key Takeaways:

Predictions:

• Turing Test passed in 2029
• Singularity Achieved in 2045 (Defined as when we merge with AI)

The rise in computational power of the brain:

When we start to augment digital AI into our brain, we can potentially unlock computational processing of several billion cycles per second.

This will have large scale ramifications for society and will allow abstract cognition that is difficult to comprehend today.

He predicts that this will happen throughout the 2030s.

The AI Effect

He talks about an interesting concept, essentially a task is considered hard or unsolvable until it is solved by AI. Once AI solves it, it is considered trivial.

As a result our true progress has been more significant than it seems in hindsight.

The Connectionist Approach of AI

The connectionist approach to problem solving is where you have a network of nodes that are interconnected. This is how the brain works, and it is how AI is starting to work.

Importantly, it allows you to solve problems without needing to understand the underlying mechanics of the problem.

Unreliable Training Data Can Lead to Good Performance

He talks about how its not intuitive that bad training data can still lead to good performance in AI. It's because the errors can cancel each other out.

GDP and Technology

He makes an interesting argument around how technology changes are exponential. E.G a phone today costs $1,000 but has more computational power than a $1,000,000 computer from the 1970s. Yet when you buy this phone, it only represents $1,000 of GDP.

Wikipedia is missing from GDP

He talks about how Wikipedia is a free resource, yet it has more information than any other resource in history. Yet it doesn't show up in GDP. Back in the day this would show up through the purchasing of Encyclopedias.

As a result, the argument is that society is actually much richer and more productive than it appears by traditional metrics.

Conclusion

This book is 292 pages and for that I learned an incredible amount. The thing I liked here is that I learned about concepts beyond just AI, for example the point around how GDP doesn't quite capture the true wealth of society.

5 stars

*****