Take a photo of a barcode or cover
A review by tobin_elliott
Nexus: A Brief History of Information Networks from the Stone Age to AI by Yuval Noah Harari
dark
informative
medium-paced
5.0
I'm very obviously getting old. Why?
Because I'm finding more horror in the non-fiction I read, than in the horror novels I read.
A couple of months ago, I almost set down Annie Jacobsen's excellent NUCLEAR WAR: A SCENARIO, because it was scaring the hell out of me and filling me with dread. Which is what a good horror story should do.
And now we come to NEXUS and once again, I was filled with dread, I was scared as hell, and I almost stopped reading several times. And each time I overcame that feeling and read on, the dread settled in worse.
Because I recognize the world Harari's describing. The past horrors. What's going on right now, right in front of us on our televisions and computers and smartphones and tablets. And then there's the projections of where AI could lead us.
A little while ago, I read Kurzwiel's THE SINGULARITY IS NEARER and, while I really enjoyed it, I did find the author's starry-eyed projections of our bright AI future a little too clean and sanitized when compared to what we're faced with on a daily basis in our real world. Because, despite that starry-eyed, hopeful future, we all know that any technology that's created for good is typically perverted almost immediately. 3D printing is amazing...until we're printing guns that can avoid metal detectors. Deep fake is really cool...until we're putting female celebrity faces on nudes.
Harari is a lot more clear-eyed and realistic about both how AI can be perverted, as well as how we constantly push forward, whether we understand the ramifications of what we're putting out into the world or not.
I remember being horrified to find out that, prior to the first atomic bomb being detonated, there were concerns that it could start a chain reaction that would ultimately ignite our atmosphere. It would have killed everyone. Ultimately, the tests went ahead because the scientists determined the chances of that to be..."unlikely."
We rolled the dice on the idea that catastrophe was unlikely. Not impossible. Not even improbable.
Unlikely.
And now, we have a system that teaches itself, and has been found to head off on unexpected tangents with a fair amount of regularity.
Once again, we're rolling the dice on "unlikely."
And that's terrifying.
This book is a must read. But it's not a hopeful one.
Because I'm finding more horror in the non-fiction I read, than in the horror novels I read.
A couple of months ago, I almost set down Annie Jacobsen's excellent NUCLEAR WAR: A SCENARIO, because it was scaring the hell out of me and filling me with dread. Which is what a good horror story should do.
And now we come to NEXUS and once again, I was filled with dread, I was scared as hell, and I almost stopped reading several times. And each time I overcame that feeling and read on, the dread settled in worse.
Because I recognize the world Harari's describing. The past horrors. What's going on right now, right in front of us on our televisions and computers and smartphones and tablets. And then there's the projections of where AI could lead us.
A little while ago, I read Kurzwiel's THE SINGULARITY IS NEARER and, while I really enjoyed it, I did find the author's starry-eyed projections of our bright AI future a little too clean and sanitized when compared to what we're faced with on a daily basis in our real world. Because, despite that starry-eyed, hopeful future, we all know that any technology that's created for good is typically perverted almost immediately. 3D printing is amazing...until we're printing guns that can avoid metal detectors. Deep fake is really cool...until we're putting female celebrity faces on nudes.
Harari is a lot more clear-eyed and realistic about both how AI can be perverted, as well as how we constantly push forward, whether we understand the ramifications of what we're putting out into the world or not.
I remember being horrified to find out that, prior to the first atomic bomb being detonated, there were concerns that it could start a chain reaction that would ultimately ignite our atmosphere. It would have killed everyone. Ultimately, the tests went ahead because the scientists determined the chances of that to be..."unlikely."
We rolled the dice on the idea that catastrophe was unlikely. Not impossible. Not even improbable.
Unlikely.
And now, we have a system that teaches itself, and has been found to head off on unexpected tangents with a fair amount of regularity.
Once again, we're rolling the dice on "unlikely."
And that's terrifying.
This book is a must read. But it's not a hopeful one.