What do you think?
Rate this book
The Self-Assembling Brain: How Neural Networks Grow Smarter
What neurobiology and artificial intelligence tell us about how the brain builds itself
How does a neural network become a brain? While neurobiologists investigate how nature accomplishes this feat, computer scientists interested in artificial intelligence strive to achieve this through technology. The Self-Assembling Brain tells the stories of both fields, exploring the historical and modern approaches taken by the scientists pursuing answers to the What information is necessary to make an intelligent neural network?
As Peter Robin Hiesinger argues, “the information problem” underlies both fields, motivating the questions driving forward the frontiers of research. How does genetic information unfold during the years-long process of human brain development―and is there a quicker path to creating human-level artificial intelligence? Is the biological brain just messy hardware, which scientists can improve upon by running learning algorithms on computers? Can AI bypass the evolutionary programming of “grown” networks? Through a series of fictional discussions between researchers across disciplines, complemented by in-depth seminars, Hiesinger explores these tightly linked questions, highlighting the challenges facing scientists, their different disciplinary perspectives and approaches, as well as the common ground shared by those interested in the development of biological brains and AI systems. In the end, Hiesinger contends that the information content of biological and artificial neural networks must unfold in an algorithmic process requiring time and energy. There is no genome and no blueprint that depicts the final product. The self-assembling brain knows no shortcuts.
Written for readers interested in advances in neuroscience and artificial intelligence, The Self-Assembling Brain looks at how neural networks grow smarter.
How does a neural network become a brain? While neurobiologists investigate how nature accomplishes this feat, computer scientists interested in artificial intelligence strive to achieve this through technology. The Self-Assembling Brain tells the stories of both fields, exploring the historical and modern approaches taken by the scientists pursuing answers to the What information is necessary to make an intelligent neural network?
As Peter Robin Hiesinger argues, “the information problem” underlies both fields, motivating the questions driving forward the frontiers of research. How does genetic information unfold during the years-long process of human brain development―and is there a quicker path to creating human-level artificial intelligence? Is the biological brain just messy hardware, which scientists can improve upon by running learning algorithms on computers? Can AI bypass the evolutionary programming of “grown” networks? Through a series of fictional discussions between researchers across disciplines, complemented by in-depth seminars, Hiesinger explores these tightly linked questions, highlighting the challenges facing scientists, their different disciplinary perspectives and approaches, as well as the common ground shared by those interested in the development of biological brains and AI systems. In the end, Hiesinger contends that the information content of biological and artificial neural networks must unfold in an algorithmic process requiring time and energy. There is no genome and no blueprint that depicts the final product. The self-assembling brain knows no shortcuts.
Written for readers interested in advances in neuroscience and artificial intelligence, The Self-Assembling Brain looks at how neural networks grow smarter.
384 pages, Hardcover
Published May 4, 2021
About the author
Peter Robin Hiesinger
2 books14 followersRobin Hiesinger is professor of neurobiology at Freie Universität Berlin, Germany. He teaches undergraduate and graduate students, leads a research laboratory (http://flygen.org) and a multilab research consortium on neural networks (http://robustcircuit.org). For more information, please visit http://selfassemblingbrain.com.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 20 of 20 reviews
October 19, 2021
Intelligence is evolutionary, a brain should be grown algorithmically and there might not be a workaround. This is the central idea of the book. The book is quite deep itself starting from the history of AI/AL, adding biology in significant details, and digging it very deep to the hardest philosophical questions. I kind of liked it a lot, but can't call this book easy to read and for just everybody.
June 10, 2021
This is a cheerful trip through the world of Artificial Intelligence neural net design (ANN) and biological neural net design within a broader context of neuroscience and Artificial Intelligence (AI) history. It is also a serious discussion of what the similarities and differences are. It is also a call for the two fields to understand each other better, for each has a lot to offer the other. The format is entertaining: Hiesinger has a series of seminars which move from level to level. For each there's an initial presentation of philosophical questions by 4 people in the two different fields, and then a very readable historical and detailed description of aspects of the fields. The descriptions are very interesting and clear. Some of the philosophical points the four discuss are confusing.
Hiesinger attempts to categorize what information is and presents a thesis that information is incomplete until the neural network is fully completed. I'm not sure if he may be defining information in different ways. It was nice to see that he includes algorithmic probability invented by Ray Solomonoff. Though he doesn't make use of that version of program hierarchy, (a probability distribution) when he talks about completing information, it suggests something like that to me: i.e. "a full explanation" --- that you won't be able know you've fully explained something and what that explanation is until "the end".
In general I'd like to see more focus on programming as it relates to both fields. and a greater awareness of non-binary aspects of AI history. But the book covers a large area, dealing in a thoughtful way with difficult questions, and it has chosen some valuable ideas and views. I learned a lot and enjoyed it, so, though the book can be a little confusing, I give it 5 stars.
Hiesinger attempts to categorize what information is and presents a thesis that information is incomplete until the neural network is fully completed. I'm not sure if he may be defining information in different ways. It was nice to see that he includes algorithmic probability invented by Ray Solomonoff. Though he doesn't make use of that version of program hierarchy, (a probability distribution) when he talks about completing information, it suggests something like that to me: i.e. "a full explanation" --- that you won't be able know you've fully explained something and what that explanation is until "the end".
In general I'd like to see more focus on programming as it relates to both fields. and a greater awareness of non-binary aspects of AI history. But the book covers a large area, dealing in a thoughtful way with difficult questions, and it has chosen some valuable ideas and views. I learned a lot and enjoyed it, so, though the book can be a little confusing, I give it 5 stars.
May 15, 2021
This is an excellent read with the author consistently pounding the “algorithmic description” as a fundamental concept to explain developmental neurobiology while staying away from all new cell and molecular developmental neurobiology. At first reading, I came up with at least a half dozen new experiments to exploit the concept and give it new teeth, wishing all the while that I were 20 years younger to pursue them. The essays about Artificial intelligence are close to what it seems today.
I reread the chapter on “Deeply engraved worship of tidy looking dichotomies.” The author’s comments were well constrained in providing information about Sperry’s dogmas. Bravo. Any well-read student of developmental neurobiology will get the message clearly. What I marvel about the book is the author’s intrusions into the philosophy at times -- absolute gems.
The last paragraph of the introduction is rather important. I do envision heated discussion amongst developmental neurobiologists.
Readers must refresh their memories by reading older literature to partake in the discussion. The book includes excellent discussion points for graduate student courses.
I have thoroughly enjoyed reading this book and look forward to the critique and evaluation of the concept. Excellent job.
Sansar Sharma
New York Medical College
Valhalla ,N.Y.
The Self-Assembling Brain: How Neural Networks Grow Smarter
I reread the chapter on “Deeply engraved worship of tidy looking dichotomies.” The author’s comments were well constrained in providing information about Sperry’s dogmas. Bravo. Any well-read student of developmental neurobiology will get the message clearly. What I marvel about the book is the author’s intrusions into the philosophy at times -- absolute gems.
The last paragraph of the introduction is rather important. I do envision heated discussion amongst developmental neurobiologists.
Readers must refresh their memories by reading older literature to partake in the discussion. The book includes excellent discussion points for graduate student courses.
I have thoroughly enjoyed reading this book and look forward to the critique and evaluation of the concept. Excellent job.
Sansar Sharma
New York Medical College
Valhalla ,N.Y.
The Self-Assembling Brain: How Neural Networks Grow Smarter
March 14, 2022
Fantastic read. As a scientist, I have always been fascinated by the human brain. Authors like Sacks and Damasio have been my go to for accessible yet deeply scientific insights into how we think and how our brains work. If you are a fan of this caliber, I highly recommend this work by Hiesinger. Both historical and contemporary, he blends thoughtful discussion across a diverse platform of speakers with his extensive knowledge of neuronal circuitry making this both a source of academic insight as well as a comfortable and easy to read glimpse into how our brains work. Can just as easily be picked up as a relaxing read or used in a classroom setting for those interested in either AI or neuroscience.
September 20, 2022
The fundamental thesis here is a sizzler: that the key feature of biological intelligence is the algorithmic unfolding of development over time. Our brains aren't loading data into a preexisting blank network; rather, they are creating the map as they navigate it. The implications of this for AI are profound. Attempts to skip development by extracting the end result and abstracting the brain's activities to a certain level of of detail (say, action potentials in neuronal firing) risks leaving the special sauce out of the recipe. How can you know that you captured the salient features of intelligence and didn't throw out the baby with the bathwater? Today's ANNs may have success on discrete problems like image recognition, but general intelligence (or human mind uploading) likely requires a total rethinking of network architecture.
Hiesinger does a great job explaining algorithmic growth in nervous systems. He blows up transhumanist dreams in pointing out that brains aren't storing bits of information and performing logical operations, but instead encode memories in specific algorithmic patterns that have to run in time to work. Try recalling your PIN out of order, or remembering the lyrics to a song backwards. It's possible that such an algorithm arises from relatively simple initial rules, but the unfolding pattern is so complex that it's mathematically undecidable: even if it's deterministic, you have to run the program to get the result. There are no shortcuts. Hiesinger is using this to describe how neural nets function, but he drops some tantalizing speculations about whether our entire universe is such an algorithmic process. You can't tell it's deterministic from inside the unfolding program, so maybe?
Hiesinger's text builds to the consequences unspooling from this insight in creative ways. He splits the book into alternating roundtables among four fictional archetypical figures (AI researcher, robotics engineer, developmental geneticist, and neuroscientist) and seminars explaining the conceptual issues. It keeps things readable, especially with numerous references to Douglas Adams spicing things up.
However, that charm can't quite conceal that the kernel to be extracted from this book is rather small. The same insight on the time-delimited nature of growth is repeated in many ways, but the idea is quickly grasped. While it's useful to explore specific applications, it grows a little repetitive and returns diminish. Really, this could have been an article, or at least a much shorter book. My rating reflects that rather than the quality of what's here. While I enjoyed reading it and found it insightful, ultimately this took up more time than necessary. Further, Hiesinger doesn't really get into the weeds on AI design or specifics of neural function, where some of the non-intuitive results might be found. Although this book has the virtue of being largely correct, many other neuro books out there are more respectful of your time. I'd seek those out first.
Hiesinger does a great job explaining algorithmic growth in nervous systems. He blows up transhumanist dreams in pointing out that brains aren't storing bits of information and performing logical operations, but instead encode memories in specific algorithmic patterns that have to run in time to work. Try recalling your PIN out of order, or remembering the lyrics to a song backwards. It's possible that such an algorithm arises from relatively simple initial rules, but the unfolding pattern is so complex that it's mathematically undecidable: even if it's deterministic, you have to run the program to get the result. There are no shortcuts. Hiesinger is using this to describe how neural nets function, but he drops some tantalizing speculations about whether our entire universe is such an algorithmic process. You can't tell it's deterministic from inside the unfolding program, so maybe?
Hiesinger's text builds to the consequences unspooling from this insight in creative ways. He splits the book into alternating roundtables among four fictional archetypical figures (AI researcher, robotics engineer, developmental geneticist, and neuroscientist) and seminars explaining the conceptual issues. It keeps things readable, especially with numerous references to Douglas Adams spicing things up.
However, that charm can't quite conceal that the kernel to be extracted from this book is rather small. The same insight on the time-delimited nature of growth is repeated in many ways, but the idea is quickly grasped. While it's useful to explore specific applications, it grows a little repetitive and returns diminish. Really, this could have been an article, or at least a much shorter book. My rating reflects that rather than the quality of what's here. While I enjoyed reading it and found it insightful, ultimately this took up more time than necessary. Further, Hiesinger doesn't really get into the weeds on AI design or specifics of neural function, where some of the non-intuitive results might be found. Although this book has the virtue of being largely correct, many other neuro books out there are more respectful of your time. I'd seek those out first.
November 14, 2021
Ponder Benzer's beautiful experiments, let evolutionary selection do the work. Accelerate evolution in the lab and dream about becoming an evolver of artificial intelligence.
'In a field that is party Artificial Life, and partly AI, some researchers therefore try to understand what it takes to evolve an AI. Arend Hintze and Chris Adami are amongst the pioneers who train ANNs not by feeding them large datasets, but by evolving genomes that encode neural networks. The neural network determines the behavior of an agent that can be selected based on its behavioral fitness. For example, the agent may be an artificial rat in a maze and selected based on its performance to find its way through the maze. Thousands of iterations of behavioral selection of random mutations will yield a genome that encodes a high performance behavior based on a specific, unpredictable distribution of synaptic weights in the ANN. In its simplest form, the genome contains one gene per synaptic weight. In effect, this 'direct encoding' yet again does away with any kind of development. However, the evolutionary training period bring a kind of algorithmic learning process in through the back door. Efforts that include indirect and developmental encoding are much rarer in this field, but the pioneers are on it. There are now simulations of genomes that indirectly change developing gene regulatory networks, which in turn change synaptic weights of recurrent neural networks that drive agent behavior. This kind of work is in its infancy, but first experiments with indirect and developmental encoding already revealed remarkable effects on robustness and adaptability.
Evolved systems allow understanding of aspects of representation in networks that are beyond most current approaches, but they have not become part of mainstream applications. The game changer could be a realization of what evolutionarily programmed developing AIs can achieve that current technology does not. '
'In a field that is party Artificial Life, and partly AI, some researchers therefore try to understand what it takes to evolve an AI. Arend Hintze and Chris Adami are amongst the pioneers who train ANNs not by feeding them large datasets, but by evolving genomes that encode neural networks. The neural network determines the behavior of an agent that can be selected based on its behavioral fitness. For example, the agent may be an artificial rat in a maze and selected based on its performance to find its way through the maze. Thousands of iterations of behavioral selection of random mutations will yield a genome that encodes a high performance behavior based on a specific, unpredictable distribution of synaptic weights in the ANN. In its simplest form, the genome contains one gene per synaptic weight. In effect, this 'direct encoding' yet again does away with any kind of development. However, the evolutionary training period bring a kind of algorithmic learning process in through the back door. Efforts that include indirect and developmental encoding are much rarer in this field, but the pioneers are on it. There are now simulations of genomes that indirectly change developing gene regulatory networks, which in turn change synaptic weights of recurrent neural networks that drive agent behavior. This kind of work is in its infancy, but first experiments with indirect and developmental encoding already revealed remarkable effects on robustness and adaptability.
Evolved systems allow understanding of aspects of representation in networks that are beyond most current approaches, but they have not become part of mainstream applications. The game changer could be a realization of what evolutionarily programmed developing AIs can achieve that current technology does not. '
August 29, 2021
I am one of the lucky people on earth who had the chance to get to know and work with Robin in person. As an accomplished neuroscientist, his approach to scientific questions/problems is strikingly different than those who find joy to follow the mainstream, textbook knowledge in their safe zones. He is always up for a challenge and this insightful and thought-provoking book is the reflection of his different mindset.
If you ever get your hands on this book, which I strongly recommend irrespective of your discipline, you will see that he successfully laid out the very principles of biological brain assembly and then show readers the fundamental differences between biological and artificial neural networks. This follows with how AI developers can use biological principles to create smarter artificial neural networks. I believe, in time, this book might turn out to be one of the reference books for AI developers but you don't really need to hold any title to enjoy this book.
If you ever get your hands on this book, which I strongly recommend irrespective of your discipline, you will see that he successfully laid out the very principles of biological brain assembly and then show readers the fundamental differences between biological and artificial neural networks. This follows with how AI developers can use biological principles to create smarter artificial neural networks. I believe, in time, this book might turn out to be one of the reference books for AI developers but you don't really need to hold any title to enjoy this book.
May 27, 2022
A perfect Interdisciplinary book.
July 14, 2022
The book shows the comparison between the development/function of our brain and AI.
The individual chapters are sorted in ascending order (complexity of development). There are different characters who represent the respective point of view and conduct dialogs among themselves.
Overall, I think it is more of a biological book than an interdisciplinary book. I would have liked to see more reference to computer science.
The individual chapters are sorted in ascending order (complexity of development). There are different characters who represent the respective point of view and conduct dialogs among themselves.
Overall, I think it is more of a biological book than an interdisciplinary book. I would have liked to see more reference to computer science.
August 8, 2021
Five starts. A thought provoking book that makes us question whether we are taking the right approach with AI?
The book is a real pleasure to read. It provides the unusual mix of strong scientific evidence and alternating fictional arguments between a geneticist, a neurobiologist, an engineer, and an AI researcher. The book compares the approaches of these different fields to figure out what makes neural networks intelligent. In the biology part, the book explores in depth HOW genes make brains. The author uses algorithmic information theory and cellular automata to argue that the developmental outcome of a single gene mutation is not predictable. Basically, only evolution can program the genes based on selection of outcomes. How evolution itself plays an irreplaceable role in neural network formation and consequently in brain related behaviours is currently missing from engineering approaches to AI.
For me, the most impactful point was learning that AI can't reconstruct overnight what nature has been constructing and moulding through years and years of evolution. And realizing that each individual is unique and carries a special set of ancestral wisdom that has been genetically refining as part of the evolutionary process. How does the brain fit in the genes? Are younger generations smarter than our ancestors? How can we account for evolution with AI? Acknowledging the significance of developmental biology and evolution when tackling learning with AI. Definitely, a perplexing, controversial and thrilling concept that marks the beginning of new approach.
The book is a real pleasure to read. It provides the unusual mix of strong scientific evidence and alternating fictional arguments between a geneticist, a neurobiologist, an engineer, and an AI researcher. The book compares the approaches of these different fields to figure out what makes neural networks intelligent. In the biology part, the book explores in depth HOW genes make brains. The author uses algorithmic information theory and cellular automata to argue that the developmental outcome of a single gene mutation is not predictable. Basically, only evolution can program the genes based on selection of outcomes. How evolution itself plays an irreplaceable role in neural network formation and consequently in brain related behaviours is currently missing from engineering approaches to AI.
For me, the most impactful point was learning that AI can't reconstruct overnight what nature has been constructing and moulding through years and years of evolution. And realizing that each individual is unique and carries a special set of ancestral wisdom that has been genetically refining as part of the evolutionary process. How does the brain fit in the genes? Are younger generations smarter than our ancestors? How can we account for evolution with AI? Acknowledging the significance of developmental biology and evolution when tackling learning with AI. Definitely, a perplexing, controversial and thrilling concept that marks the beginning of new approach.
August 17, 2023
The Self-Assembling Brain is a fascinating examination of the intersection between neurobiology and artificial intelligence. As the title suggests, the author Jonas Hielsinger posits that the brain - let's call it BNN or biological neural network for this review - is a self-assembling system with simple low-level rules resulting in incredibly complex high-level behaviors and cognition. Through dense yet lucid descriptions of cutting-edge research in neuroscience, the author makes the case that understanding how the brain wires itself may hold the key to advancing AI or artificial neural networks, henceforth called ANNs - again for this review.
The core argument underpinning the book is that neurobiology and AI are deeply intertwined fields with much to learn from each other. It emphasizes the need for collaboration between experts across disciplines to unlock the secrets of ANNs and BNNs. In some ways, the author's views are too biased toward the potential payoff of connections between the fields. The book could have benefited from more focus on the enormous divergence that has grown between these fields by now, but it still does not take away anything from the enormous value it provides, regardless.
The book truly shines while discussing the details of neuroscience. A particular highlight is the in-depth discussions of how simple local learning rules, evolved over millions of years, lead to the complex phenomena we associate with cognition and consciousness. Take language acquisition as an example – babies are not explicitly programmed with grammatical rules but rather absorb the statistical regularities in the speech patterns around them. The brain, a BNN, is wired to detect and internalize these regularities through brute repetition, unlike how we train ANNs these days.
The book illustrates this and other similar concepts through clever hypothetical dialogues between experts at the start of each chapter. In one exchange, an AI researcher presses a neuroscientist on how children acquire language without direct instruction. The neuroscientist explains how the rapid formation and pruning of neural connections allow the BNN to build statistical models reflecting the environment. While fictional, these dialogues neatly encapsulate the core themes around self-assembly and help make the later technical sections more intuitive.
An early section analyzes systems like our BNNs that are fundamentally unpredictable despite relying on simple deterministic rules. And, then, there is the reverse. Networks of neurons in lower-level areas operate largely randomly at an individual level yet produce reliable signals when aggregated. Out of disorder emerges order. The book covers the opposite phenomenons exceptionally to describe various aspects of both neural networks' complexity.
The book argues that grappling with these chaotic systems holds lessons for AI researchers seeking to build adaptable, resilient models. The brain achieves robustness despite – or perhaps because of – underlying chaos and randomness percolating through its networks.
While the author makes a strong case for collaboration between neuroscience and AI, the rapid progress of artificial intelligence over the past decade suggests the arrow of learning between the two fields has reversed in crucial ways. This reviewer feels that back when ANNs were in their infancy, AI researchers had much to gain from understanding the workings of organic BNNs. Insights into biological neural architecture and plasticity accelerated early ANN development. However, ANNs today operate unconstrained by the limitations of their organic counterparts - they do not have to be energy efficient or constructible from genetic code. They are not survival maximizers without a goal. The environments and design parameters for ANNs are now so distinct that neuroscience, for all its intricacies, likely has more to learn from AI than vice versa moving forward. While exceptions exist, the utility of modeling AI systems on detailed neurobiology has also diminished because of the incompleteness of our understanding of low-level brain function.
In summary, while conceptual inspiration clearly flowed from neuroscience to AI originally, ANNs have evolved so dramatically in recent years that they operate under very different principles and design constraints compared to BNNs. While fascinating, the complex mechanics of actual brain processes seem unlikely to offer meaningful shortcuts for today's leading AI techniques.
Such disagreements aside, here is a book where one learns in every para. The details are exhaustive but also fascinating when one begins to think how evolution has produced a gadget of such intricacy. The book not only succeeds at conveying the awe-inspiring complexity and magic of the BNNs but also throws light on how we will struggle to truly understand and master ANNs despite being their creators.
The core argument underpinning the book is that neurobiology and AI are deeply intertwined fields with much to learn from each other. It emphasizes the need for collaboration between experts across disciplines to unlock the secrets of ANNs and BNNs. In some ways, the author's views are too biased toward the potential payoff of connections between the fields. The book could have benefited from more focus on the enormous divergence that has grown between these fields by now, but it still does not take away anything from the enormous value it provides, regardless.
The book truly shines while discussing the details of neuroscience. A particular highlight is the in-depth discussions of how simple local learning rules, evolved over millions of years, lead to the complex phenomena we associate with cognition and consciousness. Take language acquisition as an example – babies are not explicitly programmed with grammatical rules but rather absorb the statistical regularities in the speech patterns around them. The brain, a BNN, is wired to detect and internalize these regularities through brute repetition, unlike how we train ANNs these days.
The book illustrates this and other similar concepts through clever hypothetical dialogues between experts at the start of each chapter. In one exchange, an AI researcher presses a neuroscientist on how children acquire language without direct instruction. The neuroscientist explains how the rapid formation and pruning of neural connections allow the BNN to build statistical models reflecting the environment. While fictional, these dialogues neatly encapsulate the core themes around self-assembly and help make the later technical sections more intuitive.
An early section analyzes systems like our BNNs that are fundamentally unpredictable despite relying on simple deterministic rules. And, then, there is the reverse. Networks of neurons in lower-level areas operate largely randomly at an individual level yet produce reliable signals when aggregated. Out of disorder emerges order. The book covers the opposite phenomenons exceptionally to describe various aspects of both neural networks' complexity.
The book argues that grappling with these chaotic systems holds lessons for AI researchers seeking to build adaptable, resilient models. The brain achieves robustness despite – or perhaps because of – underlying chaos and randomness percolating through its networks.
While the author makes a strong case for collaboration between neuroscience and AI, the rapid progress of artificial intelligence over the past decade suggests the arrow of learning between the two fields has reversed in crucial ways. This reviewer feels that back when ANNs were in their infancy, AI researchers had much to gain from understanding the workings of organic BNNs. Insights into biological neural architecture and plasticity accelerated early ANN development. However, ANNs today operate unconstrained by the limitations of their organic counterparts - they do not have to be energy efficient or constructible from genetic code. They are not survival maximizers without a goal. The environments and design parameters for ANNs are now so distinct that neuroscience, for all its intricacies, likely has more to learn from AI than vice versa moving forward. While exceptions exist, the utility of modeling AI systems on detailed neurobiology has also diminished because of the incompleteness of our understanding of low-level brain function.
In summary, while conceptual inspiration clearly flowed from neuroscience to AI originally, ANNs have evolved so dramatically in recent years that they operate under very different principles and design constraints compared to BNNs. While fascinating, the complex mechanics of actual brain processes seem unlikely to offer meaningful shortcuts for today's leading AI techniques.
Such disagreements aside, here is a book where one learns in every para. The details are exhaustive but also fascinating when one begins to think how evolution has produced a gadget of such intricacy. The book not only succeeds at conveying the awe-inspiring complexity and magic of the BNNs but also throws light on how we will struggle to truly understand and master ANNs despite being their creators.
August 13, 2023
Interesting, descriptive and profound. It touches on AI, neuroscience and biology, and the intersection of all those fields. Presents old and new ideas, that could well serve as the spark for new research directions in the area of AI and computer science.
The bibliography is a must next step after finishing it. The idea of presenting some of the concepts through a dialogue of characters makes this book easy to read and understand for people with few previous knowledge about the subject.
In my case, it opened a new door of interest, that I did not even know it existed.
The bibliography is a must next step after finishing it. The idea of presenting some of the concepts through a dialogue of characters makes this book easy to read and understand for people with few previous knowledge about the subject.
In my case, it opened a new door of interest, that I did not even know it existed.
March 6, 2024
3 stars- enjoyed, mostly.
I liked the discussions about intelligence, memory and evolution. However, way this was written was a little odd; it consisted of lots of conversations and lots of biographical information. The conversations were a little confusing, because the people talking never had the same thought processes that I did, and their conversations went in one direction, while my questions went in the opposite direction. Overall, an informative but oddly written book.
I liked the discussions about intelligence, memory and evolution. However, way this was written was a little odd; it consisted of lots of conversations and lots of biographical information. The conversations were a little confusing, because the people talking never had the same thought processes that I did, and their conversations went in one direction, while my questions went in the opposite direction. Overall, an informative but oddly written book.
September 5, 2022
Interesting overview of neurobio history. Don’t really agree with the conclusions. Focus of the book was a bit misplaced. Growth is not an essential component in understanding method of function. We know how the joints work without following their development.
March 17, 2024
Great insights from a long-time ignored area, developmental biology of the brain for neural networks and the future of machine learning.
It is one of the rare books that I came across has so many eye opening intriguing and interesting points.
It is one of the rare books that I came across has so many eye opening intriguing and interesting points.
December 10, 2021
The Hinton, Bengio and LaCunn ACM Turing Award, has made Deep Neural Networks, de jour. This book, is not useful to me personally, but supplies a fair deal of context, for the uninitiated.
February 1, 2022
I liked how complex and detailed it was
October 31, 2022
Very engaging read! Hiesinger tackles big questions regarding life, brains, and AI. Beyond its rich content, I especially enjoyed the way the book is written - different characters engaging in seminars - a clever way to structure the book, as it allows for a deep dissection of the topics form multiple perspectives simultaneously.
Great book - clear recommendation!
Great book - clear recommendation!
October 18, 2023
The same idea is repeated every single chapter. New examples are added, but essentially the message is the same every time. This one might not be worth the time.
June 25, 2021
This book provides a fascinating comparison of biological brains with artificial intelligence. The key argument is this: in biology, brains grow based on genes and are already smart before they start learning. By contrast, artificial neural networks are designed, not grown, and start out with random connections - they get smart through learning alone.
The author first tells the history of both fields, which is alone worth the entire book. But then it gets interesting: there are four scientists having arguments inbetween the chapters. There is the neuroscientist and the artificial intelligence scientist and even a robotics engineer - and they seem to agree on rather little. The discussion is very much up to date - all the newest neurobiology approaches and artificial intelligence developments are in there, but it doesn't get lost in the detail.
The book is accessible, but it is not a light read. I went back to the chapter on the historic roots several times to make sense of current ideas. I had just read Jeff Hawkins' 'A Thousand Brains' and wondered how there the idea still seems to be to design a neural network without genes, and without development. The Self-Assembling Brain provides a different idea that includes the intelligence of butterflies and humans alike, based on developmental growth of neural networks. A basic and important concept. And a book that provides a lot of information, so you can make up your own mind.
The author first tells the history of both fields, which is alone worth the entire book. But then it gets interesting: there are four scientists having arguments inbetween the chapters. There is the neuroscientist and the artificial intelligence scientist and even a robotics engineer - and they seem to agree on rather little. The discussion is very much up to date - all the newest neurobiology approaches and artificial intelligence developments are in there, but it doesn't get lost in the detail.
The book is accessible, but it is not a light read. I went back to the chapter on the historic roots several times to make sense of current ideas. I had just read Jeff Hawkins' 'A Thousand Brains' and wondered how there the idea still seems to be to design a neural network without genes, and without development. The Self-Assembling Brain provides a different idea that includes the intelligence of butterflies and humans alike, based on developmental growth of neural networks. A basic and important concept. And a book that provides a lot of information, so you can make up your own mind.
Displaying 1 - 20 of 20 reviews