What do you think?
Rate this book
Artificial Minds
Recent decades have produced a blossoming of research in artificial systems that exhibit important properties of mind. But what exactly is this dramatic new work and how does it change the way we think about the mind, or even about who or what has mind?Stan Franklin is the perfect tour guide through the contemporary interdisciplinary matrix of artificial intelligence, cognitive science, cognitive neuroscience, artificial neural networks, artificial life, and robotics that is producing a new paradigm of mind. Leisurely and informal, but always informed, his tour touches on all of the major facets of mechanisms of mind.Along the way, Franklin makes the case for a perspective that rejects a rigid distinction between mind and non-mind in favor of a continuum from less to more mind, and for the role of mind as a control structure with the essential task of choosing the next action. Selected stops include the best of the work in these different fields, with the key concepts and results explained in just enough detail to allow readers to decide for themselves why the work is significant.Major attractions include animal minds, Allan Newell's SOAR, the three Artificial Intelligence debates, John Holland's genetic algorithms, Wilson's Animat, Brooks' subsumption architecture, Jackson's pandemonium theory, Ornstein's multimind, Marvin Minsky's society of mind, Pattie Maes's behavior networks, Gerald Edelman's neural Darwinism, Drescher's schema mechanisms, Pentti Kanerva's sparse distributed memory, Douglas Hofstadter and Melanie Mitchell's Copycat, and Agre and Chapman's deictic representations.A Bradford Book
463 pages, Paperback
First published July 26, 1995
4 people are currently reading
93 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 4 of 4 reviews
September 16, 2025
Stan Franklin’s Artificial Minds came into my life in a way that I cannot disentangle from the memory of its gifting. I did not buy it in a shop, I did not stumble across it by chance in a library; it was pressed into my hands by a colleague in my first year of engineering, as though this were the sort of book that should belong to an engineer, that to read it was part of becoming what I was in the process of becoming. The gesture itself contained a subtle pedagogy: engineering was not only about material circuits or steel beams or controlled processes, it was also about the imagination of intelligence itself, about what it means to design systems that appear to act, learn, adapt. I took the gift in good faith, as one takes any initiation, and opened Franklin’s book expecting revelation. What I found, instead, was a catalogue. The book is an encyclopaedia of research programmes, a survey of artificial intelligence in the 1990s: symbolic systems, connectionist networks, robotics, emergent models, diagrams of architectures. It is useful, as one finds a handbook useful, in that it collects and presents. But it is thin where it should be thick, descriptive where it should be interrogative, mechanical where it should be philosophical. It was a disappointment, though a formative one, because it revealed to me not only what AI research was at the time but how impoverished it could be without philosophy.
Franklin writes as though the mind were an object that could be approached through multiple technical routes: one might try to simulate symbols, one might build neural architectures, one might design robots to embody cognition. Each chapter sketches one of these programmes, describes its methods and limitations, and moves on. To a novice, there is stimulation: names are learned, terms introduced, competing schools set side by side. But nowhere does Franklin ask what mind itself is. Nowhere does he interrogate whether “artificial mind” is a coherent category. The book is pragmatic in the narrowest sense, and in its pragmatism lies its philosophical poverty.
Here Dreyfus becomes indispensable. Hubert Dreyfus, from the late 1960s onward, posed a challenge to precisely the optimism Franklin embodies. Drawing on Heidegger and Merleau-Ponty, Dreyfus argued that intelligence is not symbol manipulation, not rule-following, not a matter of codified procedures, but skillful coping. Human beings are not machines that represent the world in order to act; they are beings who already inhabit a world, who act directly, who perceive meaning in context. To walk into a room is not to process a set of symbols about distances and coordinates but to be already oriented, already attuned, already skilled. To play chess as a master is not to compute billions of possibilities but to perceive configurations as meaningful. Dreyfus insisted that any attempt to model mind as symbol processing would fail because it ignored the embodied, situated nature of intelligence. Franklin, however, writes as though Dreyfus had never existed. Symbolic AI, neural nets, emergent models — all are treated as alternative techniques, not as philosophically vulnerable claims.
The absence matters. To describe expert systems without asking what expertise is; to describe neural networks without asking what learning is; to describe robotics without asking what embodiment means — this is not enough. Franklin reduces mind to mechanism. He confuses the description of programmes with the exploration of intelligence. And in doing so, he perpetuates the very naivety Dreyfus criticised: the belief that if only we refine our mechanisms, we will one day arrive at mind.
But what is mind? Heidegger would say that mind is not something housed inside the skull but the openness of Dasein, the being that finds itself already in a world, already surrounded by equipment that is ready-to-hand. To be intelligent is to be able to use a hammer without representing the hammer in thought, to drive a nail without calculating physics, to orient in a room without mapping coordinates. The ready-to-hand is the clue: intelligence is not in representation but in use, in practical coping, in being-in-the-world. Merleau-Ponty adds: perception itself is embodied. To see is not to process pixels but to inhabit a body that is oriented in space, that moves, that touches. Mind, then, is embodied, situated, practical. Franklin’s diagrams of architectures — boxes and arrows, inputs and outputs — miss this completely. They are accounts of systems in the third person, as if mind could be viewed from outside, as if mind were only mechanism. What they cannot capture is the lived first-person reality of intelligence.
It was precisely this absence that I felt most keenly as a young engineer reading Franklin. I was learning circuits and equations, but I was also living thought: struggling with problems, finding solutions that came not by calculation but by intuition, by a sudden shift of perspective, by what felt like an embodied grasp. Franklin’s book, by reducing intelligence to architectures, denied the very texture of that experience. It described the field but did not capture the phenomenon. In Heidegger’s terms, it mistook the present-at-hand (objects, representations) for the ready-to-hand (use, skill). It is this misrecognition that explains why the book feels hollow.
And yet, the hollowness is instructive. One begins to see that Franklin is not simply negligent but symptomatic. His book reflects the 1990s optimism of AI: that if one catalogues enough architectures, if one refines enough programmes, one will converge on intelligence. It is the optimism of engineering: that problems yield to technique, that complexity yields to more complex mechanisms. It is an optimism that continues today, in the triumphalism surrounding deep learning and large language models. But optimism without reflection is ideology. To believe that more layers of computation will eventually yield understanding is to mistake power for meaning. Franklin is an early specimen of this ideology. His book is not false, but it is naive.
The two-star rating I give is therefore not because the descriptions are inaccurate — they are not — but because the book fails at the level of philosophy. A book called Artificial Minds cannot restrict itself to mechanisms; it must ask what mind is. Franklin does not ask. And so the book disappoints, even as it informs.
In this sense, the book is valuable as a negative example. It shows what happens when engineers describe without questioning, when catalogues replace critique, when diagrams stand in for analysis. It shows why philosophy is necessary. Dreyfus asked what intelligence is; Heidegger asked what being is; Merleau-Ponty asked what perception is. Without these questions, one is left with only architectures. Franklin gave me the catalogue; philosophy gave me the critique. In that sense, the book was formative: it made me realise the insufficiency of engineering alone.
There is also a cultural diagnosis to be made. The book is a product of its time, but it is also a precursor of ours. The 1990s faith in catalogues and surveys is not so different from today’s faith in data and scale. Then it was expert systems and connectionism; now it is deep learning and transformers. Then it was Franklin promising that one of these approaches would converge on mind; now it is corporations promising that their models “understand.” The same mistake persists: the confusion of mechanism with mind, the refusal of embodiment, the reduction of intelligence to computation. Franklin’s book is worth reading today not for its content, which is outdated, but for its form, which is still with us. It is the form of optimism, the form of naivety, the form of technological faith without philosophical depth.
And so the book becomes almost tragic. It promises revelation — artificial minds — and delivers only catalogues. It initiates, but it does not satisfy. It informs, but it does not enlighten. It is valuable, but only negatively. Two stars reflect this ambivalence: it is worth knowing, but not worth admiring.
In closing, one must return to Dreyfus. His insistence that intelligence is embodied, situated, skillful coping, remains the necessary corrective. To ignore this is to repeat the error. Franklin repeats the error. His book, in its very limitations, shows why philosophy matters. To think about mind without philosophy is to end in catalogues. To confront the question of being-in-the-world is to begin to understand. It is a lesson I learned, in part, because this book disappointed me. For that, I remain oddly grateful.
Franklin writes as though the mind were an object that could be approached through multiple technical routes: one might try to simulate symbols, one might build neural architectures, one might design robots to embody cognition. Each chapter sketches one of these programmes, describes its methods and limitations, and moves on. To a novice, there is stimulation: names are learned, terms introduced, competing schools set side by side. But nowhere does Franklin ask what mind itself is. Nowhere does he interrogate whether “artificial mind” is a coherent category. The book is pragmatic in the narrowest sense, and in its pragmatism lies its philosophical poverty.
Here Dreyfus becomes indispensable. Hubert Dreyfus, from the late 1960s onward, posed a challenge to precisely the optimism Franklin embodies. Drawing on Heidegger and Merleau-Ponty, Dreyfus argued that intelligence is not symbol manipulation, not rule-following, not a matter of codified procedures, but skillful coping. Human beings are not machines that represent the world in order to act; they are beings who already inhabit a world, who act directly, who perceive meaning in context. To walk into a room is not to process a set of symbols about distances and coordinates but to be already oriented, already attuned, already skilled. To play chess as a master is not to compute billions of possibilities but to perceive configurations as meaningful. Dreyfus insisted that any attempt to model mind as symbol processing would fail because it ignored the embodied, situated nature of intelligence. Franklin, however, writes as though Dreyfus had never existed. Symbolic AI, neural nets, emergent models — all are treated as alternative techniques, not as philosophically vulnerable claims.
The absence matters. To describe expert systems without asking what expertise is; to describe neural networks without asking what learning is; to describe robotics without asking what embodiment means — this is not enough. Franklin reduces mind to mechanism. He confuses the description of programmes with the exploration of intelligence. And in doing so, he perpetuates the very naivety Dreyfus criticised: the belief that if only we refine our mechanisms, we will one day arrive at mind.
But what is mind? Heidegger would say that mind is not something housed inside the skull but the openness of Dasein, the being that finds itself already in a world, already surrounded by equipment that is ready-to-hand. To be intelligent is to be able to use a hammer without representing the hammer in thought, to drive a nail without calculating physics, to orient in a room without mapping coordinates. The ready-to-hand is the clue: intelligence is not in representation but in use, in practical coping, in being-in-the-world. Merleau-Ponty adds: perception itself is embodied. To see is not to process pixels but to inhabit a body that is oriented in space, that moves, that touches. Mind, then, is embodied, situated, practical. Franklin’s diagrams of architectures — boxes and arrows, inputs and outputs — miss this completely. They are accounts of systems in the third person, as if mind could be viewed from outside, as if mind were only mechanism. What they cannot capture is the lived first-person reality of intelligence.
It was precisely this absence that I felt most keenly as a young engineer reading Franklin. I was learning circuits and equations, but I was also living thought: struggling with problems, finding solutions that came not by calculation but by intuition, by a sudden shift of perspective, by what felt like an embodied grasp. Franklin’s book, by reducing intelligence to architectures, denied the very texture of that experience. It described the field but did not capture the phenomenon. In Heidegger’s terms, it mistook the present-at-hand (objects, representations) for the ready-to-hand (use, skill). It is this misrecognition that explains why the book feels hollow.
And yet, the hollowness is instructive. One begins to see that Franklin is not simply negligent but symptomatic. His book reflects the 1990s optimism of AI: that if one catalogues enough architectures, if one refines enough programmes, one will converge on intelligence. It is the optimism of engineering: that problems yield to technique, that complexity yields to more complex mechanisms. It is an optimism that continues today, in the triumphalism surrounding deep learning and large language models. But optimism without reflection is ideology. To believe that more layers of computation will eventually yield understanding is to mistake power for meaning. Franklin is an early specimen of this ideology. His book is not false, but it is naive.
The two-star rating I give is therefore not because the descriptions are inaccurate — they are not — but because the book fails at the level of philosophy. A book called Artificial Minds cannot restrict itself to mechanisms; it must ask what mind is. Franklin does not ask. And so the book disappoints, even as it informs.
In this sense, the book is valuable as a negative example. It shows what happens when engineers describe without questioning, when catalogues replace critique, when diagrams stand in for analysis. It shows why philosophy is necessary. Dreyfus asked what intelligence is; Heidegger asked what being is; Merleau-Ponty asked what perception is. Without these questions, one is left with only architectures. Franklin gave me the catalogue; philosophy gave me the critique. In that sense, the book was formative: it made me realise the insufficiency of engineering alone.
There is also a cultural diagnosis to be made. The book is a product of its time, but it is also a precursor of ours. The 1990s faith in catalogues and surveys is not so different from today’s faith in data and scale. Then it was expert systems and connectionism; now it is deep learning and transformers. Then it was Franklin promising that one of these approaches would converge on mind; now it is corporations promising that their models “understand.” The same mistake persists: the confusion of mechanism with mind, the refusal of embodiment, the reduction of intelligence to computation. Franklin’s book is worth reading today not for its content, which is outdated, but for its form, which is still with us. It is the form of optimism, the form of naivety, the form of technological faith without philosophical depth.
And so the book becomes almost tragic. It promises revelation — artificial minds — and delivers only catalogues. It initiates, but it does not satisfy. It informs, but it does not enlighten. It is valuable, but only negatively. Two stars reflect this ambivalence: it is worth knowing, but not worth admiring.
In closing, one must return to Dreyfus. His insistence that intelligence is embodied, situated, skillful coping, remains the necessary corrective. To ignore this is to repeat the error. Franklin repeats the error. His book, in its very limitations, shows why philosophy matters. To think about mind without philosophy is to end in catalogues. To confront the question of being-in-the-world is to begin to understand. It is a lesson I learned, in part, because this book disappointed me. For that, I remain oddly grateful.
December 6, 2025
This is a time capsule from early cognitive science, full of GOFAI architectures, neural nets, and robot stories that try to reverse-engineer what a mind is. It is strongest when it just explains these systems clearly and lets you see how many different technical routes people have taken toward mindlike behavior. The book feels dated now in terms of specific models, but that also makes it a useful reality check on how fast fashionable theories come and go in AI. As a philosopher, you come away with fewer big slogans and more respect for the messy engineering that any serious theory of mind has to talk to.
October 1, 2025
reads like a casebook so its skimmable if what you're looking for is a meditation on the more interesting phenomenological...epistemological...& qualia-logical questions which naturally undergird consciousness. as this book bears out, the allure of AI research is lackluster without cartesian traditions and.....well: philosophy. could've been enriched with even just 1 chapter assessing the viability of functionalism.
all things go back to philosophy of mind in the end: sweet functionalism.
all things go back to philosophy of mind in the end: sweet functionalism.
May 29, 2008
A long long time ago T.M.K. bought this for me at a bookstore in New Orleans, a strangely distant memory.
Displaying 1 - 4 of 4 reviews