Most Helpful Customer Reviews
|
|
16 of 16 people found the following review helpful:
4.0 out of 5 stars
Entertaining, basically right but bad critique, October 14, 2002
Damien's Spike is at very least an entertaining read -- somewhat like an amusement park thrill ride "we're going to be scanned into computers and live on as ever expanding immortal intelligences and/or exceeded by wild run away artificial intelligences if we don't get turned into nano gray-goo before then, Weeeeeooo..." and around and around the ride goes. Thus, it's easy not to take the subject matter very seriously...if I didn't see whispers of the spike in my work and field. The basic premise of the book, and other's like it such as "Age of Spiritual Machines" by Ray Kurzweil, is that technology doesn't just advance linearly, it feeds back upon itself. For example: faster computers let people simulate and explore new computer architectures and new materials which in turn help overcome bottleneck's to developing ever faster computers. Like interest payments on your money, technology lends itself to compound, not just linear growth. We'll, that's spike enough, but the further argument is that the day will come when a computer program will be able to figure out how to make itself function better/smarter and what new hardware changes it needs to run faster. Or once robots can design and build better robots, the compounding will itself compound and the rate of change goes hyper-exponential. I see this in my own work -- teraflop (T trillion floating point operations per second) machines will be on your lap before the decade is out. These will double for several cycles even if semiconductor technology hits a heat or quantum wall. But cheap 1, 2, 4 teraflop machines on every researcher's desk will cause material breakthroughs or trigger post-silicon computing and then 8, 16, 32, 64, 128 teraflop (T) machines are next...I put the date of 128 teraflops as 2022. Could go faster due to above advancement cycle, could slip 10 years if post-silicon is needed and harder than expected. 128 T is my (not Damien's) magic number because some claim ... that the human brain averages ~4 T, but since we don't understand the nuances of intelligence, I figure we can brute force by approximating at something like the rule of thumb where you can start relying on the law of large numbers in statistical sampling, or a factor of ~30. 4*30 T = 120 T. If robotic advances keep coming, especially the advent of strong, fast artificial muscle fiber ... we're looking at a very strange world in the 2020's. Take one example: When will the "war on terrorism" end? How about 2025 when we can mass produce robot spies and soldiers in weeks for a ~$2000 a pop while fanatics still take 12-25 years and a minimum of $25K even at 3rd world rates? Yes, they might be able to steal a few, just like they can steal a few airplanes, but techno societies will be able to produce millions. Downsides of the book: ================= Spends time showing past attempts at extrapolating curves that failed (speed of transport should essentially be infinite by now, power per person should be 1 sun apiece), but then dismisses the possibility of misreading the curves here. "This time for sure". Dwells on "minting" where especially nano-robots can make themselves and then turn around and make you anything you want. Everything will be free. Then mentions that things like brewing bear that is already essentially nano (yeast)-engineering of just this sort and last I checked, beer is not free. But...this time it's different. Little critique of scanning and uploading the brain other than having some moral/emotional qualms if the upload is destructive of the original body. Well, I've got some basic critiques. (1) You're brain isn't going to be very happy, or at least effective if you are simulated in a computer and not running a robot body well matched to our limitations: 2 legs, 2 arms, basic degrees of freedom in motion, because so much of our brain is built premised on the nature of our body, eyes, senses. (2) The ability to download other's knowledge is also fundamentally limited. You can train of a pattern recognition machines, but the structure tends to get fixed at which time it can't just incorporate scads of new knowledge without flushing the old. This is a fundamental limitation, not fast hardware. The "you" that is "you" will get washed out in the great accelerated interchange of knowledge and so where's the immortality other than in the general sense that we already have: "not us as individuals, but all of life". Your choice is to stay much unchanged and heavily bored by the eons or to loose/replace the you that is you. Ah well. End of review: In actuality, this book better serves as a sort of entertaining reference to the scientists and philosophers working in the field. So I recommend it -- 4 stars, good airplane read, good to have sitting nearby to remember who you want to look up on Google.
|
|
|
24 of 26 people found the following review helpful:
5.0 out of 5 stars
Future Shock!, April 30, 2001
The 'Spike', also known as the 'Singularity', is simply science and technological advances happening so rapidly that they appear as an almost vertical line when charted against the passage of time. Advances in disparate fields tend to feed on each other in a synergistic manner, making the graph of the Spike even steeper. We seem to be on the threshold of possibilities that will transform human life on the planet and beyond past most peoples' wildest imaginations. Damien Broderick presents a very balanced view, giving both the optimists and pesimists their viewpoints. According to Broderick, advanced artificial intelligences and nanotechnology may be two of the technologies that will predominate when the Spike arrives, but he says there probably will also be much we can't even conceive of now. Broderick writes that the Spike is not inevitable, as a disaster of one kind or another may overtake us, but most likely we will see one. If a Spike does take place it could transform everything about us, it would make for very interesting times indeed. Post Spike possibilities include immortal life for us, and a posthuman life throughout the cosmos, nano-manufacture of almost anything we want for free or nearly free, to the gray goo scenario in which nanobots are set free on the planet to reduce everything, including us, back to their component atoms. But the Luddites are wrong, we cannot stop or turn back, the promises of these technologies are just too great, and Broderick discusses this area superbly. Damien Broderick quotes several prominent researchers in various relevant fields of science and technology, their views make excellent reading, and several of them give guesses as to when a Spike may occur, but in the end we can only surmise the barest outline a Spike may take. As Broderick states in the book, "we can't yet imagine the shape of things to come". This is a book well worth reading, with extensive notes and suggested further reading at the back of the volume.
|
|
|
16 of 18 people found the following review helpful:
4.0 out of 5 stars
Fasten Your Seatbelts, June 4, 2001
In "The Spike," Damien Broderick has written an interesting, thought provoking, if sometimes frustrating book. The title of the book refers to what occurs on the right hand edge of any graph showing exponential growth. His thesis is that within the next 50 years, one or more technologies (nanotech, artificial intelligence, etc) will hit the spike and the world, and humanity itself, will be transformed beyond recognition.Because the book covers possible scenarios ranging from the probable to the almost absurd, it will be easy for many to dismiss the entire thing as a lot of nonsense. But I think that would be a mistake. There is much in this book that should be thought provoking, even if you reject some of the wilder ideas. It is easy to forget sometimes that the pace of change is speeding up dramatically. As recently as 10 years ago, the worldwide web (at least as we all know it today) did not exist - now try and imagine your life without it! I do regret that Mr. Broderick seems to have largely bought into the typical left-wing rubbish about how much worse off the poor (or maybe all of us) are today than they were in the past. Would that be the past of pre-civil rights sharecroppers in the South? Or maybe the past of the wave of immigrants of the late 19th/early 20th century described by Upton Sinclair and his contemporaries? And I have to confess that I am still struggling with the math that says a guaranteed stipend from the government of $25,000 for a family of four would "cost as much a small war." Last time I checked, 2.5 trillion dollars was a lot! Thankfully in other places in the book, Mr. Broderick seems to understand these complaints - for example when he asks the reader to imagine being in medieval Paris or Pharonic Egypt (as a member of the working classes). I also wish Mr. Broderick had devoted more pages to the possible backlash against technology we have witnessed in the past few years (e.g., the fight over genetically modified food). But overall those are fairly modest quibbles to what is otherwise a truly thought-provoking book. The human mind has trouble grappling with truly large numbers - be it the size of the universe or the exponential growth of computer processing capability. This book will get you thinking about the possibilities and consequences of those numbers.
|
|
|
Most Recent Customer Reviews
|