Vehicles: Experiments in Synthetic Psychology (MIT Press)

---

These vehicles gradually start to act more and more like sentient beings. After a couple of iterations, theystart to exhibit things like "memory", "trains Amazing little book that left me torn, just as the book is split into two distinct parts. After a couple of iterations, theystart to exhibit things like "memory", "trains of thought", "goal-directed behaviour.

The main take-away for me is the "law of uphill analysis, downhill synthesis". From the outside, complex system may look waaay more complex than they actually are. If you take an inside-out approach, like we do here by inventing the "vehicles" step-by-step, the actual internals of complex systems might in reality be less complicated than outside behaviour would indicate.

The second half of the book translates some of these findings to actual neuroscience, but I found this chapter difficult to follow without any background in the field.

See a Problem?

The content has a lot of analogies to the fields of Artificial Intelligence and Machine Learning. The internals of the later vehicles represent much of what we know as "Artificial Neural Networks". Definitely one I will be re-reading soon! Jan 01, zynphull rated it really liked it Recommends it for: How can complex behavior, such as that observed in vertebrates and, of course, humans, arise from "simple", inanimate objects such as electrons, atoms, molecules and cells? How is it that evolution gave birth to conscience? Braitenberg's approach to thi How can complex behavior, such as that observed in vertebrates and, of course, humans, arise from "simple", inanimate objects such as electrons, atoms, molecules and cells?

Braitenberg's approach to this issue consists in imagining what he calls 'vehicles', initially simple objects consisting of sensors for external input such as light, chemicals, etc. Unambitiously, he begins to show how some relatively simple changes to the internal wiring of these vehicles lead to apparently unpredictable, complex, and even 'human-like' behavior. For example, a sensor located at the vehicle's right-hand side that is strongly activated by light, if connected to its left-side motor and vice-versa makes its left side 'run faster', thus turning the vehicle toward the light.

It is substantially more difficult, he says, to understand what goes on behind the apparent behavior of an animal by watching it behave through induction than, as is the entire book's proposal, to deduce it from its internal, simpler workings, and play with its structure, watching how changes within it affect its outer behavior. Eventually, Valentino goes as far as introducing Darwinian evolution to its vehicles, which inevitably brings to mind today's genetic algorithms. The second part of the book is dedicated to a biological analysis of these vehicles, effectively showing whence Valentino a neuroscientist himself got his ideas from.

Animal's complex neurological structure, he shows, arises from learning, from sensory input, and, eventually, from internal "thoughts" or neuronal activations.

This second half is a tad biology-heavy, but reasonably readable for a layman like myself. Conscience he suggests, implicitly , then, arises not from God, not from magic, but from biological necessity - from neuronal cross-wiring. How could it be any different? Aug 23, Jewel rated it really liked it.

Recommended for anyone interested in the emergence of what we call consciousness and free will. The author takes you through journey following his axiom that it is much more difficult to try to guess internal structure just from the observation of behavior than it is to create the structure that gives the behavior.

This he calls the 'law of uphill analysis and downhill invention'. With this in mind, he adds simple rules to the design as he proceeds from Vehicle 1 to 14 that he proposes can resul Recommended for anyone interested in the emergence of what we call consciousness and free will. Something that blew my mind was how he introduces a possibility to the vehicles to have trains of thought, by adding a negative feedback system to prevent runaway activations and add stability.

The proposed emergent behavior is the ability for the vehicle's logical system to drift through concepts and associations that it has derived from the environment. People who are familiar with the ideas used in modern day reinforcement learning and artificial neural networks will probably find analogues in the kind of design that goes into the vehicles and such discoveries can be very enjoyable as well.

Feb 03, Morgan Blackledge rated it it was amazing. This book is a magic treasure chest of deep insights.

Vehicles: Experiments in Synthetic Psychology

If you are interested in the mind, behavior and evolution, you just found a secret classic. Feb 22, Arkapravo Bhaumik rated it it was amazing. Republishing my answer at quora for the question, 'What are some great mind-blowing books? All of these facets are met by this little book which spans pages, written by neuroscientist Valentino Braitenberg. The book deals with a Republishing my answer at quora for the question, 'What are some great mind-blowing books?

The book deals with artificial intelligence and cognitive science and reads more like a fable if not myth. The book sets out to model vehicles which can elicit human like behaviour and emotions using pretty bare mechanisms, and within the first few chapters Braitenberg develops vehicles which can express human like behaviour and tendencies as anger, dislike, love, attachment, desire to explore etc. However radical it may seem, each single seemingly crazy idea is strongly cemented with arguments from cognitive science.

In later sections, Braitenberg models into the vehicles; machine learning methods, Darwinian evolution, principles of association, rudiments of Turing machine, decision making, apriori knowledge and also free will. The greatest achievement of this book is that it uses jargon to a minimum and builds on to some very advanced ideas structured on trivial arguments. If somewhere in the middle of the book - lets say after the first 40 pages, if you reflect back on everything which you have read and realise that Braitenberg has been turning unimaginative matter into life with the least of effort and oodles of conjuring, then that confirms that you are truly enjoying the book.

Written in , many of these gedanken experiments have been realised with real hardware and is often part of a graduate robotics course. Aug 11, Caolan McMahon rated it it was amazing. Using a series of plausible if not always practicable machines, this book walks you through several thought experiments which you soon discover map one possible path to the development of the brain.

Valentino introduces us to the law of "uphill analysis and downhill synthesis" - the idea that we tend to overestimate the underlying complexity of phenomena observed through analysis. Instead, following the 'downhill synthesis' approach, this book attempts to produce similar phenomena through rela Using a series of plausible if not always practicable machines, this book walks you through several thought experiments which you soon discover map one possible path to the development of the brain. Instead, following the 'downhill synthesis' approach, this book attempts to produce similar phenomena through relatively simple inventions called 'vehicles'.

Each vehicle is thought-provoking and succinctly described, and the final part of the book provides a basis for their workings in neurological research - it's a great read! Oct 25, Julius rated it liked it. A friend who teaches a class in robotics gave this to me with the admonition that I read it slowly, just a few pages at a time. That was excellent advice: It's a nice way to see, concretely, how simple components sensors, effectors and rule-sets can generate surprisingly complex A friend who teaches a class in robotics gave this to me with the admonition that I read it slowly, just a few pages at a time.

It's a nice way to see, concretely, how simple components sensors, effectors and rule-sets can generate surprisingly complex behaviour--a valuable insight not only for roboticists but also for anyone interested in AI, machine learning, or neuropsychology. Jun 08, Eli Weinstein rated it it was amazing. Very beautiful especially in the beginning sections. Jun 07, Justin Smith rated it it was amazing.

A really fun book on though experiments. May 20, Hamish Huggard rated it really liked it. Makes figuring out how the brain works look both tractable and fun!

Jan 14, Josh rated it really liked it Shelves: This is the kind of book where, if you had little prior knowledge of biological and cognitive systems, you would be hard-pressed to say whether it's oversimplistic or brilliant. That ambiguity is a testament to its refreshing treatment of an age-old subject. If nothing else, the reading is worth the law of uphill analysis and downhill invention.

I didn't go through the biological mechanism section, mostly because none of the mechanisms seemed at-all implausible and because I had to return to the This is the kind of book where, if you had little prior knowledge of biological and cognitive systems, you would be hard-pressed to say whether it's oversimplistic or brilliant. I didn't go through the biological mechanism section, mostly because none of the mechanisms seemed at-all implausible and because I had to return to the library.

Mar 15, Steve Stuart rated it it was amazing. This is an amazing little book. The first half of the book is a series of thought experiments in which you're invited to imagine the behavior of a sequence of "vehicles": You can't quite call them robots, because they're purely mechanical without any software. An experimental study of imagination, in Readings in Perception , Beardslee, D. Thesis, University of Michigan, Ann Arbor.

Vehicles: Experiments in Synthetic Psychology - Valentino Braitenberg - Google Книги

Properties of Categories pp. Blood in the Gutter electronic reserve October 3 Hebb, D. Mechanisms of Learning and Development pp. October 8 Kaplan, S. Active symbols and internal models: Towards a cognitive connectionism. A model of temporal dynamics in a cell assembly.

Learning in Difficult Environments: A new look at some classical principles. The University of Michigan. Credit Assignment web web , html version. November 19 Damasio, A. Descarte's Error November 21 Chown, E. December 3 Kinsbourne, M. There are 14 vehicles in all -- a series of hypothetical, self-operating mobile machines that exhibit increasingly sophisticated behavior similar to that in the real biological or neuroscientific world.

One might assemble Braitenberg's vehicles like cunning table-top toys that might look like the fantastic Albrecht drawings. Whatever their guise, they behave like living creatures, ranging from simple light-seeking automata to vehicles that an observing psychologist or philosopher might conclude were operated by concealed human beings.

Each of the vehicles in the series incorporates the essential features of all the earlier models but represents an evolutionary advance to a higher plateau of complexity. Along the way, they come to embody the instincts of fight or flight, the characteristic behaviors impelled by love and by logic, manifestations of foresight, concept formation, creative thinking, personality, and free will.

The author shows that these attributes and patterns of behavior can be internalized into machines using the simplest parts -- a collection of environmental sensors, some wheel-driving motors, various threshold devices, a few fictional but logically and technologically plausible components with special properties.

The author locates many elements of his fantasy in current brain research in a concluding section of extensive biological notes.

• The Consequences of Decision-Making
• rock in a sock
• How to Please A Woman In Bed: A Mans Guide to Lovemaking
• Frontiers
• Performance Leadership: The Next Practices to Motivate Your People, Align Stakeholders, and Lead You