What do you think?
Rate this book
How to Think About AI: A Guide for the Perplexed
Revealing the unfolding story of Artificial Intelligence, Richard Susskind presents a short non-technical guide that challenges us to think differently about AI. Susskind brings AI out of computing laboratories, big tech companies, and start-ups—and into everyday life.
In recent years, and certainly since the launch of ChatGPT, there has been massive public and professional interest in Artificial Intelligence. But people are confused about what AI is, what it can and cannot do, what is yet to come, and whether AI is good or bad for humanity and civilization—whether it will provide solutions to mankinds major challenges or become our gravest existential threat. There is also confusion about how we should regulate AI and where we should draw moral boundaries on its use.
In How To Think About AI, Richard Susskind draws on his experience of working on AI since the early 1980s. For Susskind, balancing the benefits and threats of artificial intelligence is the defining challenge of our age. He explores the history of AI and possible scenarios for its future. His views on AI are not always conventional. He positions ChatGPT and generative AI as no more than the latest chapter in the ongoing story of AI and claims we are still at the foothills of developments. He argues that to think responsibly about the impact of AI requires us to look well beyond today's technologies, suggesting that not-yet-invented technologies will have far greater impact on us in the 2030s than the tools we have today. This leads Susskind to discuss the possibility of conscious machines, magnificent new AI-enabled virtual worlds, and the impact of AI on the evolution of biological humans.
In recent years, and certainly since the launch of ChatGPT, there has been massive public and professional interest in Artificial Intelligence. But people are confused about what AI is, what it can and cannot do, what is yet to come, and whether AI is good or bad for humanity and civilization—whether it will provide solutions to mankinds major challenges or become our gravest existential threat. There is also confusion about how we should regulate AI and where we should draw moral boundaries on its use.
In How To Think About AI, Richard Susskind draws on his experience of working on AI since the early 1980s. For Susskind, balancing the benefits and threats of artificial intelligence is the defining challenge of our age. He explores the history of AI and possible scenarios for its future. His views on AI are not always conventional. He positions ChatGPT and generative AI as no more than the latest chapter in the ongoing story of AI and claims we are still at the foothills of developments. He argues that to think responsibly about the impact of AI requires us to look well beyond today's technologies, suggesting that not-yet-invented technologies will have far greater impact on us in the 2030s than the tools we have today. This leads Susskind to discuss the possibility of conscious machines, magnificent new AI-enabled virtual worlds, and the impact of AI on the evolution of biological humans.
Audible Audio
Published May 8, 2025
171 people are currently reading
688 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 - 30 of 49 reviews
September 22, 2025
Richard Susskind is a British lawyer and writer specialized in legal technology, thinking and writing about the interface between artificial intelligence and society since the 1980s. His 2025 book How to Think About AI focuses on how he views our current post-GenAI era with a mix of open-minded enthusiasm and prudent vigilance. With the advent of ChatGPT and similar publicly-available free and paid tools, there are of course dozens of books being published on this topic every month in 2025 from a variety of angles. I do think Susskind's contribution is worth reading for its balanced, circumspect commentary based on the author's decades of exploration of this topic.
Further reading: additional artificial intelligence books I'd recommend (both pre- and post-ChatGPT):
Co-Intelligence: The Definitive, Bestselling Guide to Living and Working with AI by Ethan Mollick
Code Dependent: Living in the Shadow of AI by Madhumita Murgia
Searches: Selfhood in the Digital Age by Vauhini Vara
The Big Nine: How the Tech Titans and Their Thinking Machines Could Warp Humanity by Amy Webb from 2019 -- perhaps more relevant now than 6 years ago
My statistics:
Book 293 for 2025
Book 2219 cumulatively
Further reading: additional artificial intelligence books I'd recommend (both pre- and post-ChatGPT):
Co-Intelligence: The Definitive, Bestselling Guide to Living and Working with AI by Ethan Mollick
Code Dependent: Living in the Shadow of AI by Madhumita Murgia
Searches: Selfhood in the Digital Age by Vauhini Vara
The Big Nine: How the Tech Titans and Their Thinking Machines Could Warp Humanity by Amy Webb from 2019 -- perhaps more relevant now than 6 years ago
My statistics:
Book 293 for 2025
Book 2219 cumulatively
December 14, 2025
The saddest book I’ve ever read and one which has sent me into an existential crisis.
This is a fascinating first step in discussing the innumerable risks of AI, with the caveat that even where he expresses his concern that robots will replace us all, he is still a bit subservient to rampant technology for my liking. Nevertheless, this should be required reading for all policy makers along with Frankenstein and Jurassic Park - just because we can doesn’t mean we should.
This is a fascinating first step in discussing the innumerable risks of AI, with the caveat that even where he expresses his concern that robots will replace us all, he is still a bit subservient to rampant technology for my liking. Nevertheless, this should be required reading for all policy makers along with Frankenstein and Jurassic Park - just because we can doesn’t mean we should.
February 11, 2026
High-level but balanced and useful.
February 6, 2026
The one word I would use to describe AI before reading this book is “uncomfortable” and now that I read this book for the ‘perplexed’ I’ll say that I am now deeply disturbed.
“For now, humans still have the choice whether to accept or use any given AI system or not.”
…”the first ultra intelligent machine is the last invention that man need ever make.”
I disagree with thr author that the recipients of a professional service aren’t interested in the process. I understand that clients don’t want doctors;they want health and that they don’t want therapists;they want peace of mind and on and on with any profession. Yes people want the outcome and end result BUT what AI does not provide is the human element. The conversation, the discussion, the STORIES, the eye contact, the photographs on the walls of the office. There are pieces that could be eliminated that save time and money but ultimately don’t we as human beings crave connection and community (which I guess some people are still seeking from AI and dating/marrying them?).
Maybe what AI will eventually do is the true definition of a revolution. The workers at the top that earn the most will be replaced by AI and the work that we value the least, manual labor, will be what is left for humans. The women that care for our babies and children, the people that clean our homes and hotel rooms, the people that make and deliver our food, the people that walk and groom pets, the people that fix our plumbing will be the careers remaining where human beings earn money. Hmm
I like this observation, “…it’s likely that may forms of human expression, not least live performance, will continue to be valued by humans for their own sake.”
The author lists 7 categories of risks most of which we can imagine without being told explicitly. The 7th risk he lists of AI is not using AI and in turn missing “exciting economic opportunities, of neglecting credible solutions to social problems, of failing to avert all manner of disasters, of letting chances for greater human happiness slip through our fingers.” Good grief!! The hopes and dreams this man has for computers is astounding. Yes there may be improvements to processes, and tasks and even outcomes but to proclaim that if we don’t use AI then human happiness cannot be achieved to its fullest potential is ludicrous. Full stop. He then goes on to say that “we’ll need to be bold in our deployment of AI if we are to secure the benefits.” I thoroughly disagree and we MUST proceed with extreme caution. He starts off the book by saying that the people that create AI at this very moment don’t even know how it’s functioning and getting its responses. Like hold the horses. Let’s stop moving forward at this speed and figure out what exactly we are creating!!!
“We may allow that to err is human, but we’re nervous about extending the same tolerance to our machines.” Of course! If we allowed our calculators to be wrong sometimes but were not sure how often that would negate their use entirely. If sometimes our GPS sent us the wrong way, but most of the time it didn’t, well we would never use it. If we can’t “trust” the machine to be accurate, it loses all of its utility IMHO.
Also he’s saying that “the current response to AI is too reactive, too short-term, and being conducted too hastily.” DUH!! The systems were released upon us human beings from one night to the next. We didn’t have any input in the creation, or “ethics” or any kind of training and now that it’s been thrust upon the human population, we are frantically scrambling to REact!
Ok I think I found the craziest non-fiction sentence I’ve ever read in my life: “we will need to contemplate entirely new forms of legal personality (ex: people, pets, corporations) if we are to endow AI systems with appropriate entitlements, impose sensible duties on them, and grant them fitting permissions.”
Hooray I agree with the author here - “technology is too important to be left to technologists. I do not think, by default, we should allow AI engineers to become unaccountable ‘social engineers.’”
There’s an entire chapter on consciousness. What human consciousness is and how we don’t fully understand it and if AI can be conscious. I think the idea is absurd and the fact that it needs a chapter is alarming. Maybe I’m too simplistic in my thinking but we are created in God’s image and everything good and perfect is from Him. How could we create a conscious entity? AI is simply a human creation and we may be worshipping it and showing how relatable those ancient Israelites were.
Now in the chapter about virtual reality and he notes that “living much of our lives virtually may seem grotesque to some readers.” Absolutely I think that word is accurate. He then goes on to state that “the colors might be richer, the sounds purer, the smells sweeter, the landscapes more sweeping than we have experienced before.” Humans would literally be playing God and attempting to escape his creation.
Oh and then he asks us to contemplate the idea that we might be living in a digital simulation. I won’t be considering that because I have truths that I hold to be true. And that I exist physically on the planet earth in the year 2026 is a truth I believe.
Overall, I think was an important book for me to read about AI. I read it from the perspective of a Christian and so some of the questions the author posits have absolute truths to them which lead to different questions entirely. Nevertheless I do believe that people should be asking questions and be aware of what these people are creating.
“For now, humans still have the choice whether to accept or use any given AI system or not.”
…”the first ultra intelligent machine is the last invention that man need ever make.”
I disagree with thr author that the recipients of a professional service aren’t interested in the process. I understand that clients don’t want doctors;they want health and that they don’t want therapists;they want peace of mind and on and on with any profession. Yes people want the outcome and end result BUT what AI does not provide is the human element. The conversation, the discussion, the STORIES, the eye contact, the photographs on the walls of the office. There are pieces that could be eliminated that save time and money but ultimately don’t we as human beings crave connection and community (which I guess some people are still seeking from AI and dating/marrying them?).
Maybe what AI will eventually do is the true definition of a revolution. The workers at the top that earn the most will be replaced by AI and the work that we value the least, manual labor, will be what is left for humans. The women that care for our babies and children, the people that clean our homes and hotel rooms, the people that make and deliver our food, the people that walk and groom pets, the people that fix our plumbing will be the careers remaining where human beings earn money. Hmm
I like this observation, “…it’s likely that may forms of human expression, not least live performance, will continue to be valued by humans for their own sake.”
The author lists 7 categories of risks most of which we can imagine without being told explicitly. The 7th risk he lists of AI is not using AI and in turn missing “exciting economic opportunities, of neglecting credible solutions to social problems, of failing to avert all manner of disasters, of letting chances for greater human happiness slip through our fingers.” Good grief!! The hopes and dreams this man has for computers is astounding. Yes there may be improvements to processes, and tasks and even outcomes but to proclaim that if we don’t use AI then human happiness cannot be achieved to its fullest potential is ludicrous. Full stop. He then goes on to say that “we’ll need to be bold in our deployment of AI if we are to secure the benefits.” I thoroughly disagree and we MUST proceed with extreme caution. He starts off the book by saying that the people that create AI at this very moment don’t even know how it’s functioning and getting its responses. Like hold the horses. Let’s stop moving forward at this speed and figure out what exactly we are creating!!!
“We may allow that to err is human, but we’re nervous about extending the same tolerance to our machines.” Of course! If we allowed our calculators to be wrong sometimes but were not sure how often that would negate their use entirely. If sometimes our GPS sent us the wrong way, but most of the time it didn’t, well we would never use it. If we can’t “trust” the machine to be accurate, it loses all of its utility IMHO.
Also he’s saying that “the current response to AI is too reactive, too short-term, and being conducted too hastily.” DUH!! The systems were released upon us human beings from one night to the next. We didn’t have any input in the creation, or “ethics” or any kind of training and now that it’s been thrust upon the human population, we are frantically scrambling to REact!
Ok I think I found the craziest non-fiction sentence I’ve ever read in my life: “we will need to contemplate entirely new forms of legal personality (ex: people, pets, corporations) if we are to endow AI systems with appropriate entitlements, impose sensible duties on them, and grant them fitting permissions.”
Hooray I agree with the author here - “technology is too important to be left to technologists. I do not think, by default, we should allow AI engineers to become unaccountable ‘social engineers.’”
There’s an entire chapter on consciousness. What human consciousness is and how we don’t fully understand it and if AI can be conscious. I think the idea is absurd and the fact that it needs a chapter is alarming. Maybe I’m too simplistic in my thinking but we are created in God’s image and everything good and perfect is from Him. How could we create a conscious entity? AI is simply a human creation and we may be worshipping it and showing how relatable those ancient Israelites were.
Now in the chapter about virtual reality and he notes that “living much of our lives virtually may seem grotesque to some readers.” Absolutely I think that word is accurate. He then goes on to state that “the colors might be richer, the sounds purer, the smells sweeter, the landscapes more sweeping than we have experienced before.” Humans would literally be playing God and attempting to escape his creation.
Oh and then he asks us to contemplate the idea that we might be living in a digital simulation. I won’t be considering that because I have truths that I hold to be true. And that I exist physically on the planet earth in the year 2026 is a truth I believe.
Overall, I think was an important book for me to read about AI. I read it from the perspective of a Christian and so some of the questions the author posits have absolute truths to them which lead to different questions entirely. Nevertheless I do believe that people should be asking questions and be aware of what these people are creating.
November 29, 2025
I appreciate the author's sober, comprehensive analysis, some predictions and lots of suggested best practices regarding AI. The timing of my reading this book was a necessary antidote to the last book I read on AI 'If Anyone Builds It...'. I'm not saying that one book is correct or better than the other - rather this book is broader and certainly less fatalist. It's a great AI primer and palette cleaner providing a quality overview while still presenting cautionary projections.
I appreciated the 'outcome vs process' framing device the author often references along with stressing how we need to stop simply thinking of AI as a tool that has potential to automate or replace human tasks vs it's potential to completely reinvent the task or the object itself. Additionally, Mr Susskind does a great job distinguishing between examples where AI will automate, transform, or eliminate tasks, objects, etc
I can't put my fiber on it but I suspect the author leans toward optimism and willingness to place a hefty bit of faith on the market and legislators to use common sense to guide the safe development of AI - this seems naive and lacks an honest examination of historic precedence. He appears to strike a decent balance of accelerate and prices with caution... I think?
"The risk here, in myriad circumstances, is that of not using AI and, in turn, of missing exciting economic opportunities, of neglecting credible solutions to social problems, of failing to avert all manner of disasters, of letting chances for greater human happiness slip through our fingers. Risk aversion brings its own risks. Preoccupation with the downside can lead to a culture of excessive caution."
Statements like this lean toward a it's better to push forward and deal with the consequences rather than study, adapt and apply to the pace and manner as we develop. Thankfully he spends the last third of the book cautioning about AI.
I'm going to ramble next so I'll wrap up the review by stating - this is a solid book to read if your newish to the subject or want to refine what you already have read.
--------------
I found myself agreeing with him more than with other authors of AI books though I felt like he creates a straw man argument when he references how current Generative AI systems cannot explain their output in an analogous manner because they do not reason like humans. He assures the reader that this is OK because we don't ask for recommendations and explanations from many of the ubiquitous and critical systems and professionals we interact with daily - microwaves, watches, etc. and judges and psychologists. This is a stretch given that with sufficient time all of these references can be explained and understood COMPLETELY vs AI systems that are opaque.
The author assures the reader that "we have confidence that they (the devices & professionals) were developed responsibly and tested thoroughly and that the development and testing methods are open to scrutiny."
He posits that we just need better explanations and assurances on AI development models...!? AI developers have admitted that for some time they have been unable to explain how the results are generated and additional time to review the code will not magically provide clarity. We simply do not and will likely never understand what's in the box and therefore it sounds the author is in essence saying, trust us, we're experts.
I appreciated the 'outcome vs process' framing device the author often references along with stressing how we need to stop simply thinking of AI as a tool that has potential to automate or replace human tasks vs it's potential to completely reinvent the task or the object itself. Additionally, Mr Susskind does a great job distinguishing between examples where AI will automate, transform, or eliminate tasks, objects, etc
I can't put my fiber on it but I suspect the author leans toward optimism and willingness to place a hefty bit of faith on the market and legislators to use common sense to guide the safe development of AI - this seems naive and lacks an honest examination of historic precedence. He appears to strike a decent balance of accelerate and prices with caution... I think?
"The risk here, in myriad circumstances, is that of not using AI and, in turn, of missing exciting economic opportunities, of neglecting credible solutions to social problems, of failing to avert all manner of disasters, of letting chances for greater human happiness slip through our fingers. Risk aversion brings its own risks. Preoccupation with the downside can lead to a culture of excessive caution."
Statements like this lean toward a it's better to push forward and deal with the consequences rather than study, adapt and apply to the pace and manner as we develop. Thankfully he spends the last third of the book cautioning about AI.
I'm going to ramble next so I'll wrap up the review by stating - this is a solid book to read if your newish to the subject or want to refine what you already have read.
--------------
I found myself agreeing with him more than with other authors of AI books though I felt like he creates a straw man argument when he references how current Generative AI systems cannot explain their output in an analogous manner because they do not reason like humans. He assures the reader that this is OK because we don't ask for recommendations and explanations from many of the ubiquitous and critical systems and professionals we interact with daily - microwaves, watches, etc. and judges and psychologists. This is a stretch given that with sufficient time all of these references can be explained and understood COMPLETELY vs AI systems that are opaque.
The author assures the reader that "we have confidence that they (the devices & professionals) were developed responsibly and tested thoroughly and that the development and testing methods are open to scrutiny."
He posits that we just need better explanations and assurances on AI development models...!? AI developers have admitted that for some time they have been unable to explain how the results are generated and additional time to review the code will not magically provide clarity. We simply do not and will likely never understand what's in the box and therefore it sounds the author is in essence saying, trust us, we're experts.
February 8, 2026
What made me buy this book was the epigraph: “None of this book was written by AI (for better or for worse).”
As it is a brand new book, not everything is hopelessly outdated. The main reason for this is that Susskind does not attempt to define AI. He mentions that, at one conference, no definition received more than 54% approval. However, he also does not attempt to explain how AI works. For a book promising to be a “Guide for the Perplexed”, this is a bit disappointing.
He contrasts process thinking from outcome thinking. This might be a good idea, explaining what AI is capable of doing instead of trying to explain how it is doing it. There is a little bit of history of AI but not enough to give the reader an idea what was going on. For example mentioning ELIZA without telling that all its answers were canned means hiding the most important information from the reader.
He obviously knows what he’s talking about, and he’s not shy to tell us (again and again) that he was one of the early developers of AI. That was back in the days of expert systems, which means the early bronze age. He is a lawyer, and his programme could follow handwritten paths to reach conclusions on legal questions. This was undoubtedly a significant achievement, but perhaps not what we mean today when we talk about AI. To his credit, he is honest enough to admit that when machine learning emerged, he could not conceive of using it for his purposes.
Susskind is not a young man as you can guess, (but depressingly, younger than I am), so it is kind of funny that he chooses two even older men to explain the two cultures of AI, namely Kissinger and Chomsky. Kissinger was saying that AI “will redefine human knowledge [...] and reorganize politics and society.” Whereas Chomsky thinks that AI means amorality, faux science and linguistic incompetence. And is, of course, trivial compared to the human mind.
There are some interesting thoughts in the book. For example the chapter where he talks on the lack of adequate vocabulary (an obvious example is the misnomer hallucination . I liked his reply when at a conference he was asked for the first robot being able to iron shirts. He said that AI would get us shirts that do not need ironing. An in a similar vein: He was at a conference talking about the future of neurosurgery. He said that people don’t want neurosurgeons. The neurosurgeons in the audience did not like that. They wanted to hear that their jobs were safe. However, neurosurgeons are the best solution to a problem we have right now. AI will not help them do their job. AI will make their job obsolete because people will no longer suffer from damage that needs fixing.
The chapter on the question of AI consciousness is quite good. Which means not obviously nonsensical.
Overall, it's a good book, but I don’t know anyone I would recommend it to. Perhaps, given the topic, it is impossible to write a book that is neither too shallow nor too deep. However, I would say that it leans towards being too shallow.
As it is a brand new book, not everything is hopelessly outdated. The main reason for this is that Susskind does not attempt to define AI. He mentions that, at one conference, no definition received more than 54% approval. However, he also does not attempt to explain how AI works. For a book promising to be a “Guide for the Perplexed”, this is a bit disappointing.
He contrasts process thinking from outcome thinking. This might be a good idea, explaining what AI is capable of doing instead of trying to explain how it is doing it. There is a little bit of history of AI but not enough to give the reader an idea what was going on. For example mentioning ELIZA without telling that all its answers were canned means hiding the most important information from the reader.
He obviously knows what he’s talking about, and he’s not shy to tell us (again and again) that he was one of the early developers of AI. That was back in the days of expert systems, which means the early bronze age. He is a lawyer, and his programme could follow handwritten paths to reach conclusions on legal questions. This was undoubtedly a significant achievement, but perhaps not what we mean today when we talk about AI. To his credit, he is honest enough to admit that when machine learning emerged, he could not conceive of using it for his purposes.
Susskind is not a young man as you can guess, (but depressingly, younger than I am), so it is kind of funny that he chooses two even older men to explain the two cultures of AI, namely Kissinger and Chomsky. Kissinger was saying that AI “will redefine human knowledge [...] and reorganize politics and society.” Whereas Chomsky thinks that AI means amorality, faux science and linguistic incompetence. And is, of course, trivial compared to the human mind.
There are some interesting thoughts in the book. For example the chapter where he talks on the lack of adequate vocabulary (an obvious example is the misnomer hallucination . I liked his reply when at a conference he was asked for the first robot being able to iron shirts. He said that AI would get us shirts that do not need ironing. An in a similar vein: He was at a conference talking about the future of neurosurgery. He said that people don’t want neurosurgeons. The neurosurgeons in the audience did not like that. They wanted to hear that their jobs were safe. However, neurosurgeons are the best solution to a problem we have right now. AI will not help them do their job. AI will make their job obsolete because people will no longer suffer from damage that needs fixing.
The chapter on the question of AI consciousness is quite good. Which means not obviously nonsensical.
Overall, it's a good book, but I don’t know anyone I would recommend it to. Perhaps, given the topic, it is impossible to write a book that is neither too shallow nor too deep. However, I would say that it leans towards being too shallow.
October 30, 2025
There’s a line from Mortimer & Whitehouse: Gone Fishing that really stayed with me. Bob Mortimer writes:
“I don’t feel scared of death, I just feel so frustrated and sad to think I won’t see how stories end. My children’s story. My wife’s. The football. All the stories going on in the world that you’re going to miss the end of. It’s just so sad.”
That came to mind while reading this book. Susskind paints a future shaped, and perhaps ruled, by AI. It’s both fascinating and unsettling, and too many people are pretending it’s not happening.
Like Bob, I’m not scared - just sad that I might not see how the story ends.
I gave this book four stars, not because it isn’t excellent (it is), but because it took me a while to get through. I save five stars for the ones I can’t put down. Susskind writes clearly and makes complex ideas genuinely accessible - which matters, because I don’t usually enjoy business or tech books.
This one, though, is different: thoughtful, well-structured, and quietly human. A great entry point for anyone who wants to think seriously - and realistically - about what AI might mean for our future.
“I don’t feel scared of death, I just feel so frustrated and sad to think I won’t see how stories end. My children’s story. My wife’s. The football. All the stories going on in the world that you’re going to miss the end of. It’s just so sad.”
That came to mind while reading this book. Susskind paints a future shaped, and perhaps ruled, by AI. It’s both fascinating and unsettling, and too many people are pretending it’s not happening.
Like Bob, I’m not scared - just sad that I might not see how the story ends.
I gave this book four stars, not because it isn’t excellent (it is), but because it took me a while to get through. I save five stars for the ones I can’t put down. Susskind writes clearly and makes complex ideas genuinely accessible - which matters, because I don’t usually enjoy business or tech books.
This one, though, is different: thoughtful, well-structured, and quietly human. A great entry point for anyone who wants to think seriously - and realistically - about what AI might mean for our future.
September 30, 2025
A must-read for everyone interested in or affected by AI – a very balanced and readable book.
March 16, 2025
The Future with AI | A Gift or A Nightmare
This thoughtful and incisive book provokes us to think beyond our usual lives to how we can perhaps determine the impact AI will have on us as individuals and on humanity as a whole.
What is clear is we need to reexamine how we look at AI and how we build and manage our expectations of what AI can, should and will do.
A critically important book for anyone who cares about the future.
This thoughtful and incisive book provokes us to think beyond our usual lives to how we can perhaps determine the impact AI will have on us as individuals and on humanity as a whole.
What is clear is we need to reexamine how we look at AI and how we build and manage our expectations of what AI can, should and will do.
A critically important book for anyone who cares about the future.
December 20, 2025
One of Clifford Simak’s favorite narrative premises in his SF novels involves a world with no (or few) humans and a vibrant, self-perpetuating species of sentient robots. Simak’s robots demonstrate, among themselves and to any remaining humans, a kind of sincere humanity that must have appeared missing to him throughout much of the tumultuous 20th century. Richard Susskind also envisions a world full of robots, but his vision is far bleaker.
As an unchecked outcome of our current development of AI, Susskind envisions a world where humans are becoming obsolete and incapable to contributing to society in a way that is as efficient or as effective as AI-enhanced, functionally “sentient” machines. It’s a future that he envisions to have been in the cards for decades but only readily visible since late 2022, now hastened by minimally checked capitalism and debilitating ethical shortsightedness.
Maybe I’m just a naïve humanist, but I don’t buy it. I don’t buy the premise that AI, in its current form can replace all the human labor that it appears to be capable of replacing or that it could be developed to replace much of or all of our intellectual labor. Belief in that position seems to stem from a poor concept of what intelligence is and what makes it valuable.
Susskind opts for a deflationary account of intelligence that starts and stops with the ability to gather data from our sensory equipment as we engage with the world and then process that information. He imagines that machinic improvements to data collection and data processing will increase the speed and improve the efficiency of information processing so that it surpasses what humans currently provide, making them better at work requiring this kind of intelligence. But does that make AI more intelligent?
As I see it, intelligence is not just the collection and processing of information. What people bring to the intellectual work that they do is not wholly in the processing of information, it is also in the intentionality of that work. We do the work with purpose, toward some end, in service of a motive that arises out of a world that we exist in and where that work has meaning and uptake. I say something because I want you to understand, to do, to believe. Further, what I say or write I stand behind – it is a stance, a statement about my way of being in the world and how I anticipate it intersecting with your way of being in the world. This way of being in the world and acting on it are part of intellectual labor and AI are just not in the world in the same way.
Now, if you can imagine a world of only machines then perhaps there is an argument to be made that machines, via their programming, can be thought to have some expectation about how its information is processed and shared and how it will be used and by what. But is that expectation the same kind of awareness that drives intentionality in human action? Maybe in some sense of the word, but when we are talking about a world of humans and machines interacting, intentionality is a more problematic assumption. Focusing on intentionality won’t likely alter the development of AI but it feels vital to the conversation about how and why we should talk about and perhaps legislate a place for AI.
Susskind had his finger on an important part of the issue when talking about early applications of AI in the medical and legal professions (circa 1980s). He noted that AI in the professions was considered off limits for some kinds of work, like unsupervised diagnosis of medical conditions and generating legal opinions. You know, obvious stuff where human judgment is required. This is where I think Susskind overlooks the point I’m trying to make: it is important for humans to be doing this sort of work, not because they have a unique ability, or are unsurpassed in their efficiency, or are infallible, but because they are responsible for their opinions. Humans are accountable in ways that AI are not. And I imagine that few people who are responsible for taking a stance on the basis of their intelligence and for sharing that information with intention would be willing to have that responsibility delegated to AI. I’m not. Neither am I ready for others to delegate that work to AI and share it with me. If it’s about how to operate a piece of software, maybe I’m fine with it, but in other situations the stakes are too high.
There are some provocative ideas in the book that are worth reading and grappling with, but also too many unsupported or poorly supported assumptions about the rate of growth and investment in AI and the degree to which it will crowd out most forms of human labor. There are also some strangely certain assumptions about AI intelligence reaching a level that would put it on par with humans, effectively resulting in a separate sentient race.
As an unchecked outcome of our current development of AI, Susskind envisions a world where humans are becoming obsolete and incapable to contributing to society in a way that is as efficient or as effective as AI-enhanced, functionally “sentient” machines. It’s a future that he envisions to have been in the cards for decades but only readily visible since late 2022, now hastened by minimally checked capitalism and debilitating ethical shortsightedness.
Maybe I’m just a naïve humanist, but I don’t buy it. I don’t buy the premise that AI, in its current form can replace all the human labor that it appears to be capable of replacing or that it could be developed to replace much of or all of our intellectual labor. Belief in that position seems to stem from a poor concept of what intelligence is and what makes it valuable.
Susskind opts for a deflationary account of intelligence that starts and stops with the ability to gather data from our sensory equipment as we engage with the world and then process that information. He imagines that machinic improvements to data collection and data processing will increase the speed and improve the efficiency of information processing so that it surpasses what humans currently provide, making them better at work requiring this kind of intelligence. But does that make AI more intelligent?
As I see it, intelligence is not just the collection and processing of information. What people bring to the intellectual work that they do is not wholly in the processing of information, it is also in the intentionality of that work. We do the work with purpose, toward some end, in service of a motive that arises out of a world that we exist in and where that work has meaning and uptake. I say something because I want you to understand, to do, to believe. Further, what I say or write I stand behind – it is a stance, a statement about my way of being in the world and how I anticipate it intersecting with your way of being in the world. This way of being in the world and acting on it are part of intellectual labor and AI are just not in the world in the same way.
Now, if you can imagine a world of only machines then perhaps there is an argument to be made that machines, via their programming, can be thought to have some expectation about how its information is processed and shared and how it will be used and by what. But is that expectation the same kind of awareness that drives intentionality in human action? Maybe in some sense of the word, but when we are talking about a world of humans and machines interacting, intentionality is a more problematic assumption. Focusing on intentionality won’t likely alter the development of AI but it feels vital to the conversation about how and why we should talk about and perhaps legislate a place for AI.
Susskind had his finger on an important part of the issue when talking about early applications of AI in the medical and legal professions (circa 1980s). He noted that AI in the professions was considered off limits for some kinds of work, like unsupervised diagnosis of medical conditions and generating legal opinions. You know, obvious stuff where human judgment is required. This is where I think Susskind overlooks the point I’m trying to make: it is important for humans to be doing this sort of work, not because they have a unique ability, or are unsurpassed in their efficiency, or are infallible, but because they are responsible for their opinions. Humans are accountable in ways that AI are not. And I imagine that few people who are responsible for taking a stance on the basis of their intelligence and for sharing that information with intention would be willing to have that responsibility delegated to AI. I’m not. Neither am I ready for others to delegate that work to AI and share it with me. If it’s about how to operate a piece of software, maybe I’m fine with it, but in other situations the stakes are too high.
There are some provocative ideas in the book that are worth reading and grappling with, but also too many unsupported or poorly supported assumptions about the rate of growth and investment in AI and the degree to which it will crowd out most forms of human labor. There are also some strangely certain assumptions about AI intelligence reaching a level that would put it on par with humans, effectively resulting in a separate sentient race.
January 11, 2026
Although a good overview, the essential gaps in this book seriously underestimate the real dangers we see occurring with AI today. I think it is definitely possible to point out areas that should recieve focus first, beyond the meer statement that a multi functional taskforce should be assigned.
I also think the discussion on consciousness between humans and machines was glossed over too easily. Consciuousness may be impossible to prove, but it is much more than just being able to self reflect and come up with a more or less founded decision. A key omission to me was failing to adres the link to feelings and the ability to relate decisions to the emotional impact to other humans next to more logical reasoned impact. And I still believe a machine being able to come up with calculated outcome and being able to link perception to sensors is vastly different from connecting to feelings. Which to me immediately connects to the main difference of being Human. If you dismiss animals from having consciousness, why would it be reasonable to assume machines would potentially be able to develop a conscience without humans being able to detect?
Finally I think the author took a few too many leaos only supported by the evidence that he himself was one of the first focussjng on AI or ‘a group of very clever people’ having looked into it. That does not support assuming that legal teams of self driving cars must have reviewed well enough all aspect of risk and therefore the further development is warranted. No, they did not as was demonstrated in multiple occasions where we found they had not thought about certain aspects that came up.
And assuming brain computer interfaces will naturally extend to feelings from the current physical signal transmissions is making the same mistake as made with the consciousness discussion. Why would a machine be able to do that, and to what point?
Virtual reality and suggesting we might live in a Matrix type of world also feels like misleading the discussion.
Overal I think the book leans too much towards a laissez-faire for the sake of huge experted benefits that may never materialize and failing to highlight the vast negative impacts currently occurring. Consider politicians, big tech and other powerful forces openly abusing AI to manipulate, causing chaos to further their own agenda and take focus away from real problems to be solved. I think we could start with a clear statement that AI should never cause harm and define what harm means. Including: no stealing ideas from others (consider AI can only create outcome based on all historic creativeness from humans which is the data that is used), should not cause harm without evidence (no unfounded accusations or statements without provable evidence) and cannot be used in news outlets or outings covering real live events including in social media. Other harmful applications such as using AI to create nude pictures based on a fotograph should be a Criminal offense right away. Too much harm is now coming from AI and allowing it for the sake of future great wins is not acceptable. Especially since most major recent technical development only seems to lead to the rich becoming richer. And obviously big tech would not want that as it would take away their massive profits when they have to develop solutions that will support this, using the misduided reasining of freedom of expression. And what about not causing harm with your freedom of expression?
I also think the discussion on consciousness between humans and machines was glossed over too easily. Consciuousness may be impossible to prove, but it is much more than just being able to self reflect and come up with a more or less founded decision. A key omission to me was failing to adres the link to feelings and the ability to relate decisions to the emotional impact to other humans next to more logical reasoned impact. And I still believe a machine being able to come up with calculated outcome and being able to link perception to sensors is vastly different from connecting to feelings. Which to me immediately connects to the main difference of being Human. If you dismiss animals from having consciousness, why would it be reasonable to assume machines would potentially be able to develop a conscience without humans being able to detect?
Finally I think the author took a few too many leaos only supported by the evidence that he himself was one of the first focussjng on AI or ‘a group of very clever people’ having looked into it. That does not support assuming that legal teams of self driving cars must have reviewed well enough all aspect of risk and therefore the further development is warranted. No, they did not as was demonstrated in multiple occasions where we found they had not thought about certain aspects that came up.
And assuming brain computer interfaces will naturally extend to feelings from the current physical signal transmissions is making the same mistake as made with the consciousness discussion. Why would a machine be able to do that, and to what point?
Virtual reality and suggesting we might live in a Matrix type of world also feels like misleading the discussion.
Overal I think the book leans too much towards a laissez-faire for the sake of huge experted benefits that may never materialize and failing to highlight the vast negative impacts currently occurring. Consider politicians, big tech and other powerful forces openly abusing AI to manipulate, causing chaos to further their own agenda and take focus away from real problems to be solved. I think we could start with a clear statement that AI should never cause harm and define what harm means. Including: no stealing ideas from others (consider AI can only create outcome based on all historic creativeness from humans which is the data that is used), should not cause harm without evidence (no unfounded accusations or statements without provable evidence) and cannot be used in news outlets or outings covering real live events including in social media. Other harmful applications such as using AI to create nude pictures based on a fotograph should be a Criminal offense right away. Too much harm is now coming from AI and allowing it for the sake of future great wins is not acceptable. Especially since most major recent technical development only seems to lead to the rich becoming richer. And obviously big tech would not want that as it would take away their massive profits when they have to develop solutions that will support this, using the misduided reasining of freedom of expression. And what about not causing harm with your freedom of expression?
October 19, 2025
2.5 problem is that these books quickly become dated, if not obsolete. organized in five major parts: (1) Understanding AI; (2) Thinking Differently; (3) Making AI Work; (4) Confronting Risks; (5) Contemplating the Future.
1. Process-thinkers (those preoccupied with how AI works (algorithms, architectures, mechanisms) v Outcome-thinkers (those more concerned with what AI does — its effects, consequences, uses etc. He warns that process-thinkers can become overly fixated on regulating today’s techniques, rather than anticipating what’s coming next. So shift your attention (or at least balance) toward outcome-thinking — to think about how AI can reshape institutions, economies, ethical norms, human roles.
2. What we are seeing is early, primitive versions relative to what is possible in coming decades. So again, always ask: “What might come next?” to guard against being blindsided.
3 The tendency many professionals have to believe that AI will affect others but not their own field. (Lawyers, doctors, etc., often assume their domain is immune.) This mindset blinds people to potential disruption. He suggests professionals should instead ask: “What if AI built alternatives that are faster, cheaper, more trusted than the services we offer?”
As lawyer, author delves more deeply into legal domain, biggest gains may lie not in automating existing tasks (e.g. document review), but in restructuring systems — creating new ways of resolving disputes, preventing problems before they arise, or even bypassing traditional litigation.
4. AI’s perils. Ie Dual-use risks: the same technology might be used for beneficial ends (e.g. climate modeling) or malicious ones (e.g. surveillance, misinformation, weaponization).
• Power concentration and inequality: if AI capabilities become concentrated among a few entities (governments, big tech), large swaths of society may lack influence over their development.
• Loss of human autonomy: as AI systems take over decision-making, we must guard against erosion of judgment, responsibility, and control.
• Transparency, accountability, and trust: Who is responsible when AI systems err? How do we audit or regulate them? How do we maintain trust when systems are opaque?
• Misalignment and “value drift”: as systems evolve, they may deviate from human intentions or values.
emphasis that these challenges cannot be addressed purely technically — regulation, institutions, professional norms, and public engagement are necessary.
5. Futures, consciousness, and human evolution
Book ends with speculation on conscious machines, or machines with forms of agency more like ours than current ones and envisions immersive AI-enabled virtual worlds (metaverse-type environments), where human and AI agents cohabit and interact. Also speculates on how biological humans might co-evolve with AI, possibly blending with, being augmented by, or being outpaced by intelligence that is not biological.
1. Process-thinkers (those preoccupied with how AI works (algorithms, architectures, mechanisms) v Outcome-thinkers (those more concerned with what AI does — its effects, consequences, uses etc. He warns that process-thinkers can become overly fixated on regulating today’s techniques, rather than anticipating what’s coming next. So shift your attention (or at least balance) toward outcome-thinking — to think about how AI can reshape institutions, economies, ethical norms, human roles.
2. What we are seeing is early, primitive versions relative to what is possible in coming decades. So again, always ask: “What might come next?” to guard against being blindsided.
3 The tendency many professionals have to believe that AI will affect others but not their own field. (Lawyers, doctors, etc., often assume their domain is immune.) This mindset blinds people to potential disruption. He suggests professionals should instead ask: “What if AI built alternatives that are faster, cheaper, more trusted than the services we offer?”
As lawyer, author delves more deeply into legal domain, biggest gains may lie not in automating existing tasks (e.g. document review), but in restructuring systems — creating new ways of resolving disputes, preventing problems before they arise, or even bypassing traditional litigation.
4. AI’s perils. Ie Dual-use risks: the same technology might be used for beneficial ends (e.g. climate modeling) or malicious ones (e.g. surveillance, misinformation, weaponization).
• Power concentration and inequality: if AI capabilities become concentrated among a few entities (governments, big tech), large swaths of society may lack influence over their development.
• Loss of human autonomy: as AI systems take over decision-making, we must guard against erosion of judgment, responsibility, and control.
• Transparency, accountability, and trust: Who is responsible when AI systems err? How do we audit or regulate them? How do we maintain trust when systems are opaque?
• Misalignment and “value drift”: as systems evolve, they may deviate from human intentions or values.
emphasis that these challenges cannot be addressed purely technically — regulation, institutions, professional norms, and public engagement are necessary.
5. Futures, consciousness, and human evolution
Book ends with speculation on conscious machines, or machines with forms of agency more like ours than current ones and envisions immersive AI-enabled virtual worlds (metaverse-type environments), where human and AI agents cohabit and interact. Also speculates on how biological humans might co-evolve with AI, possibly blending with, being augmented by, or being outpaced by intelligence that is not biological.
January 8, 2026
After having stated my new years resolution to finally learn something about AI, I received this very thoughtful Xmas present.
As a (gulp) senior person and not professionally engaged with technology I knew that AI was something happening on the edge of my consciousness. Should I be paranoid or hopeful? Am I being left behind or is it all a storm in a teacup? Is this relevant to me or can I quietly leave it to my grandchildren?
How To Think About AI is a welcome introduction to how to actually think about AI. I found the distinction between process and outcome thinking very useful and enjoyed trying to identify my style of interaction with the idea of AI -- and I learnt, like many I imagine, that it depends. Depends on situation, need, anxiety, sensitivity to risk, what others around me say and think, and above all on my own curiosity.
Like many readers, I am the daily (even hourly) object of multiple generative AIs vying for my attention--on my phone, on my computer, social media, interactions with organisations especially (dear god no!) my bank, insurers, and increasingly everything government related.
So shifting gears I decided to learn something from the AIs that I pay for, willingly or not, on my phone (thank you Apple) and my computer (thank you Microsoft). After completing each book chapter I engaged with ChatGPT and Copilot giving each different tasks: ChatGPT personal and Copilot professional. After some stumbles, the most useful enquiry was to ask each what they needed from me so they could best assist me. This is not the place to review or compare these AIs rather to show that Richard Susskind's approach prompted curiosity and engagement that I had not experienced before. Will I keep developing my skill with these AIs--definitely. Will I keep thinking critically about AIs--definitely. Am I less perplexed as promised by the subtitle--hmm maybe not, now the more I know the more I see AI, rather like a medieval sea serpent, stirring under the dark waters. A kraken or a saviour, not sure, but not something to be ignored.
Stepping away from the virtual world, this is a lovely hardback, great cover design, very tactile cover and pages, a pleasure to hold. I liked the typeface and general feel of the book. Excellent production values.
So a very enthusiastic four stars.
As a (gulp) senior person and not professionally engaged with technology I knew that AI was something happening on the edge of my consciousness. Should I be paranoid or hopeful? Am I being left behind or is it all a storm in a teacup? Is this relevant to me or can I quietly leave it to my grandchildren?
How To Think About AI is a welcome introduction to how to actually think about AI. I found the distinction between process and outcome thinking very useful and enjoyed trying to identify my style of interaction with the idea of AI -- and I learnt, like many I imagine, that it depends. Depends on situation, need, anxiety, sensitivity to risk, what others around me say and think, and above all on my own curiosity.
Like many readers, I am the daily (even hourly) object of multiple generative AIs vying for my attention--on my phone, on my computer, social media, interactions with organisations especially (dear god no!) my bank, insurers, and increasingly everything government related.
So shifting gears I decided to learn something from the AIs that I pay for, willingly or not, on my phone (thank you Apple) and my computer (thank you Microsoft). After completing each book chapter I engaged with ChatGPT and Copilot giving each different tasks: ChatGPT personal and Copilot professional. After some stumbles, the most useful enquiry was to ask each what they needed from me so they could best assist me. This is not the place to review or compare these AIs rather to show that Richard Susskind's approach prompted curiosity and engagement that I had not experienced before. Will I keep developing my skill with these AIs--definitely. Will I keep thinking critically about AIs--definitely. Am I less perplexed as promised by the subtitle--hmm maybe not, now the more I know the more I see AI, rather like a medieval sea serpent, stirring under the dark waters. A kraken or a saviour, not sure, but not something to be ignored.
Stepping away from the virtual world, this is a lovely hardback, great cover design, very tactile cover and pages, a pleasure to hold. I liked the typeface and general feel of the book. Excellent production values.
So a very enthusiastic four stars.
September 5, 2025
🤖’Saving humanity with and from AI’🤖
Richard Susskind the author of ‘How to Think about AI’ describes this as the defining challenge of our age.
This book subtitled ‘A Guide for the Perplexed’ is excellent. A short readable book, doesn’t dive into the underlying tech but concentrates on what it can and could do and how we should respond. The negatives: I find it biased towards the good that AI may bring and has too little on the risks and damage it will bring. And assumes that AI can and will only operate within the current new liberalist society.
But if you are working in any of the professions or creative industries this book is essential reading.
My takeaways from the book:
🤖‘Not us thinking’ is everywhere. Every profession sees AI as having greater scope in other disciplines than their own
🤖We forget at our peril that ‘the competition that kills you will not look like you’
🤖The professions greatest threat may come from AI empowered clients and organisations that can do much for themselves without needing to refer out
🤖Any role or work that is of itself ‘not intrinsically valuable’ is vulnerable to AI
🤖We don’t care about the process only the outcome. We don’t want surgeons, we want to be healthy. We don’t want a lawyer arguing our case in court, we want justice.
🤖AI has started by taking over our current roles and doing them faster, more accurately and cheaper. But AI will then find a better way of achieving the same outcome (innovation) and eventually could remove the problem at source (elimination) meaning no need for the role at all.
🤖We already don’t understand how our most advanced AI systems actually work! Terrifying.
🤖If our only question about AI is ‘what jobs will be left to our children’ we are missing the point. Everything is going to change.
The book’s conclusion is that the AI genie is only ‘half out of the bottle’. That we have a decade to come up with an international regulatory framework that will ‘contain our systems ….to prevent a digital merger or takevoer’. Again terrifying.
I don’t want to leave decisions about what society will look like to the tech billionaires and the politicians in their pockets.
Reading this book will help us take part in the debate.
Richard Susskind the author of ‘How to Think about AI’ describes this as the defining challenge of our age.
This book subtitled ‘A Guide for the Perplexed’ is excellent. A short readable book, doesn’t dive into the underlying tech but concentrates on what it can and could do and how we should respond. The negatives: I find it biased towards the good that AI may bring and has too little on the risks and damage it will bring. And assumes that AI can and will only operate within the current new liberalist society.
But if you are working in any of the professions or creative industries this book is essential reading.
My takeaways from the book:
🤖‘Not us thinking’ is everywhere. Every profession sees AI as having greater scope in other disciplines than their own
🤖We forget at our peril that ‘the competition that kills you will not look like you’
🤖The professions greatest threat may come from AI empowered clients and organisations that can do much for themselves without needing to refer out
🤖Any role or work that is of itself ‘not intrinsically valuable’ is vulnerable to AI
🤖We don’t care about the process only the outcome. We don’t want surgeons, we want to be healthy. We don’t want a lawyer arguing our case in court, we want justice.
🤖AI has started by taking over our current roles and doing them faster, more accurately and cheaper. But AI will then find a better way of achieving the same outcome (innovation) and eventually could remove the problem at source (elimination) meaning no need for the role at all.
🤖We already don’t understand how our most advanced AI systems actually work! Terrifying.
🤖If our only question about AI is ‘what jobs will be left to our children’ we are missing the point. Everything is going to change.
The book’s conclusion is that the AI genie is only ‘half out of the bottle’. That we have a decade to come up with an international regulatory framework that will ‘contain our systems ….to prevent a digital merger or takevoer’. Again terrifying.
I don’t want to leave decisions about what society will look like to the tech billionaires and the politicians in their pockets.
Reading this book will help us take part in the debate.
February 7, 2026
A compact, non-technical primer that reframes the conversation around outcomes, institutions, and the second-order effects of AI. I have really enjoyed this book. Key insights for me are:
Process-Thinking vs Outcome-Thinking
Automation, Innovation and Elimination
Elimination, not automation, is the real threat
Developers aren't immune
Process-Thinking vs Outcome-Thinking
When you listen to any conversations about AI it's worth to keep in mind that there are two main camps: people who are process-orientated and those who are outcome-orientated. The process-orientated people are the ones who usually complain about the imperfections of the current technical approach. The outcome-orientated people are the ones who are focused on the end result and and don't really care about the technical details. So it helps me to understand the discussion and where people are coming from.
Automation, Innovation and Elimination
There are three ways the AI will affect the society and the jobs: Automation, Innovation and Elimination. People who are protective of the current state of their work usually focus on the Automation piece, forgetting the other two, and this is a big problem.
Elimination, not automation, is the real threat
Lawyers think their jobs are safe because legal reasoning is hard to automate. They're right, but they're asking the wrong question. The question isn't "can AI do a lawyer's job?" It's "does society still need lawyers?" Most people don't want adversarial legal proceedings. They want fair, predictable outcomes. What if minor crimes don't go to court at all? "You did X under circumstances Y, therefore consequence Z" - a deterministic system, not a trial. No judge's discretion, no lawyer's persuasion, just algorithmic justice. Sounds dystopian? Maybe. But also: faster, cheaper, more consistent. The demand for lawyers doesn't decline, it collapses.
Developers aren't immune
We think our jobs are safe because "AI can't architect complex systems." True. But what if AI makes systems so simple they don't need architecting? What if generated code is so reliable that your debugging skills become irrelevant? Elimination, not automation. That's the risk nobody's pricing in.
July 25, 2025
Richard Susskind has written the definitive guide for anyone—leader, strategist, or citizen—trying to make sense of artificial intelligence. How to Think About AI is not another technical manual or a breathless account of the latest chatbot. It is a masterfully clear, deeply insightful, and ultimately profound exploration of the questions we should all be asking about the most consequential technology of our age.
Susskind's greatest strength is his ability to cut through the noise. He brilliantly dismantles the flawed, human-centric ways we typically think about AI, introducing concepts like the "AI Fallacy" (the mistaken belief that AI must copy human thought) and distinguishing between "process-thinking" (how it works) and "outcome-thinking" (what it does). This framing alone is worth the price of admission, as it clarifies so much of the confused public debate.
He then pushes the conversation far beyond the usual talking points of job automation and ethics, forcing the reader to confront the truly transformative potential of AI:
- Innovation and Elimination: The real disruption won't come from automating old tasks, but from innovating entirely new ways of working and eliminating the problems that required those tasks in the first place.
- The Inevitability of AGI: He makes a compelling case that we should be planning for Artificial General Intelligence—machines that can rival human intellect in almost all domains—not in some distant future, but within the next decade.
- The Cosmic Gamble:The final section is a philosophical tour de force, examining machine consciousness and the "AI Evolution Hypothesis"—the idea that we may be birthing our own successors.
What makes this book so powerful is that it is both pragmatic and philosophical. Susskind provides a clear-eyed framework for leaders navigating the immediate challenges of AI while simultaneously elevating the discussion to the highest possible stakes: the future of humanity itself. This is not a book about technology; it's a book about us. An essential, and frankly, terrifying read.
Susskind's greatest strength is his ability to cut through the noise. He brilliantly dismantles the flawed, human-centric ways we typically think about AI, introducing concepts like the "AI Fallacy" (the mistaken belief that AI must copy human thought) and distinguishing between "process-thinking" (how it works) and "outcome-thinking" (what it does). This framing alone is worth the price of admission, as it clarifies so much of the confused public debate.
He then pushes the conversation far beyond the usual talking points of job automation and ethics, forcing the reader to confront the truly transformative potential of AI:
- Innovation and Elimination: The real disruption won't come from automating old tasks, but from innovating entirely new ways of working and eliminating the problems that required those tasks in the first place.
- The Inevitability of AGI: He makes a compelling case that we should be planning for Artificial General Intelligence—machines that can rival human intellect in almost all domains—not in some distant future, but within the next decade.
- The Cosmic Gamble:The final section is a philosophical tour de force, examining machine consciousness and the "AI Evolution Hypothesis"—the idea that we may be birthing our own successors.
What makes this book so powerful is that it is both pragmatic and philosophical. Susskind provides a clear-eyed framework for leaders navigating the immediate challenges of AI while simultaneously elevating the discussion to the highest possible stakes: the future of humanity itself. This is not a book about technology; it's a book about us. An essential, and frankly, terrifying read.
June 22, 2025
90%
Richard Suskin, I recommend, recognize that name, But not in a historical lens. From the AI perspective in field, Which Makes me curious because he's speaking. Like he's spending. If he feels for a very long time, Magic artificial intelligence is relatively new field. It's only been around for seven pounds 70 years And the retail application commercial application of that has been much smaller window, I believe. And it has changed dramatically since then, but that's the thing. We're always referring to AI as llms. It's, like, no AI, are not a new thing. Guys, we've been around for so long. We use them all the time and back in propagations, Python scripts, the software that we use. It's all AI you. Do you have a thermostat of smart thermostat? That's AI, Uh, there are so many machines that automate and regulate things around us all the times, like an outlet timer. For freak's sake. I mean pharmacological interventions, things that are artificial to change us even with from within. And that's one thing that actually made me think about The fact that if you don't want to be synthetic, then you shouldn't take medications because they do have Permanent alterations on your body Because they do At least partly integrate frequently. So, Yeah, functionally. A biological cyborg. We essentially all are Kick vaccines. That's the kind of shit that we're talking about here, too. It's just not the gross wet, where stuff that's like metal and wires and Plastics throughout us. But I mean, Hey, I specifically Are just artificial intelligence automated programs that we can run. I love it. The thing is like a psychologist that, I know, even had thought that it was automated intelligence And not artificial intelligence. And like, wonderful, yes, that's a great way to look at it. So, this book largely covers a moral framework, Which is just fine. I didn't really learn anything new here. However, It was A nice take.
Richard Suskin, I recommend, recognize that name, But not in a historical lens. From the AI perspective in field, Which Makes me curious because he's speaking. Like he's spending. If he feels for a very long time, Magic artificial intelligence is relatively new field. It's only been around for seven pounds 70 years And the retail application commercial application of that has been much smaller window, I believe. And it has changed dramatically since then, but that's the thing. We're always referring to AI as llms. It's, like, no AI, are not a new thing. Guys, we've been around for so long. We use them all the time and back in propagations, Python scripts, the software that we use. It's all AI you. Do you have a thermostat of smart thermostat? That's AI, Uh, there are so many machines that automate and regulate things around us all the times, like an outlet timer. For freak's sake. I mean pharmacological interventions, things that are artificial to change us even with from within. And that's one thing that actually made me think about The fact that if you don't want to be synthetic, then you shouldn't take medications because they do have Permanent alterations on your body Because they do At least partly integrate frequently. So, Yeah, functionally. A biological cyborg. We essentially all are Kick vaccines. That's the kind of shit that we're talking about here, too. It's just not the gross wet, where stuff that's like metal and wires and Plastics throughout us. But I mean, Hey, I specifically Are just artificial intelligence automated programs that we can run. I love it. The thing is like a psychologist that, I know, even had thought that it was automated intelligence And not artificial intelligence. And like, wonderful, yes, that's a great way to look at it. So, this book largely covers a moral framework, Which is just fine. I didn't really learn anything new here. However, It was A nice take.
January 4, 2026
Richard Susskind, a professor of law, seems like an odd person to write about how we should think about AI. But he also happens to be a Fellow of the British Computer Society and President of the Society of Computers and Law, and did his doctorate on AI and Law in the 1980s. As it turns out, he has some pretty astute ideas on how we should actually be thinking about AI.
Susskind aggregates and summarises ideas from multiple texts, guiding us through decades of AI development and thinking, and then projects the possible futures from where we are now, with a generous sprinkling of is own opinions. And in my own opinion, it is nearly all spot-on.
The most important observation Susskind makes is of process-thinking vs outcome-thinking, which is an observation suited to many types of innovative technology, not just AI, but is definitely overlooked in most of the contemporary AI discussions. It is a brilliant point and one that I am now always trying to apply in my own thoughts. The book also goes on to consider various risks of AI and the possible structural shifts it may cause.
My only critique is the final chapters, which veer towards very philosophical meta-thinking, which contrary to the subtitle "A guide for the Perplexed" left me a little more Perplexed that I was at the start (although it is partly because I wasn't sure about the relevance to the topic).
All in all however, it is something that anyone who wishes to have a constructive discussion about how AI will affect us should read, in order for them to be more articulate and reasoned in structuring their thoughts. You had just better read it in the next couple of years otherwise it might be too late.
Susskind aggregates and summarises ideas from multiple texts, guiding us through decades of AI development and thinking, and then projects the possible futures from where we are now, with a generous sprinkling of is own opinions. And in my own opinion, it is nearly all spot-on.
The most important observation Susskind makes is of process-thinking vs outcome-thinking, which is an observation suited to many types of innovative technology, not just AI, but is definitely overlooked in most of the contemporary AI discussions. It is a brilliant point and one that I am now always trying to apply in my own thoughts. The book also goes on to consider various risks of AI and the possible structural shifts it may cause.
My only critique is the final chapters, which veer towards very philosophical meta-thinking, which contrary to the subtitle "A guide for the Perplexed" left me a little more Perplexed that I was at the start (although it is partly because I wasn't sure about the relevance to the topic).
All in all however, it is something that anyone who wishes to have a constructive discussion about how AI will affect us should read, in order for them to be more articulate and reasoned in structuring their thoughts. You had just better read it in the next couple of years otherwise it might be too late.
June 24, 2025
A useful and illuminating guide to AI, pitched at an accessible level for most general readers, which suits me down to the ground. As AI's tentacles begin to weave their way ever deeper into life, it's harder for even those on the luddite end of the scale to ignore it, in work and personal life. I have to assess various AI tools and apps, for possible assistance in work and I'm barely someone you'd consider tech savvy. It's coming and it's not going away. Noting the truism that many of these new apps and features are "the worst they're ever going to be" is a good way to remind readers not to judge on the present capabilities alone.
Susskind guides the reader through the current landscape but with a firm eye on the present, and invites multiple perspectives on the opportunities and threats to our normative practices that are going to come, whether we like it or not. Which makes it all the more valuable to think about the possible more dramatic consequences now, so as to be prepared and to give governments, and citiziaens and businesses time to decide what they want from this technology.
It's a quick and breezy read, designed mostly to prompt ways of thnking rather than a technical examination of the nitty gritty. Those already familiar with it, may not find much new here, but it is still instructive and a useful call to arms to think about all the possibilities now.
Susskind guides the reader through the current landscape but with a firm eye on the present, and invites multiple perspectives on the opportunities and threats to our normative practices that are going to come, whether we like it or not. Which makes it all the more valuable to think about the possible more dramatic consequences now, so as to be prepared and to give governments, and citiziaens and businesses time to decide what they want from this technology.
It's a quick and breezy read, designed mostly to prompt ways of thnking rather than a technical examination of the nitty gritty. Those already familiar with it, may not find much new here, but it is still instructive and a useful call to arms to think about all the possibilities now.
July 3, 2025
The 1st half had some interesting sections but 2nd half legal bits and it's complicated not useful to me. I do agree with him that many folks think or hope it won't impact their profession. A good bit on trying to understand AI like it's a human. I don't understand it, it doesn't work like I do, humans have some special sauce when we mostly don't understand how the brain works either. When as long as it works it doesn't not have to work like a human or "It doesn't matter whether the cat is black or white as long as it hunts". Recently read where google is trying to understand how it works on a simple thing like adding two numbers. AI did 4 queries including two estimates to get the answer or to be clear did not work like a child would be taught at all but got the right answer! It's all over sports fans that AI will be a huge thing where we either add it as a brain extension or we give up even trying. Try ChatGPT 4.0 for a couple weeks having a dialogue about any complicated thing to see that it is better than trying to find someone that 'knows' more or may not with 5.0 etc coming. To quote ChatGPT BATAR
Better
Athan
The
Available
Right-now
You name it in 5 years everyone will be using AI to save $, time because it is BATAR. We used to tease you have been amazoned but soon it will be you have been machined.
Better
Athan
The
Available
Right-now
You name it in 5 years everyone will be using AI to save $, time because it is BATAR. We used to tease you have been amazoned but soon it will be you have been machined.
December 7, 2025
EN: Susskind gives a clear and calm look at AI, and the book really made me think. I liked that he stays realistic and does not use fear to make his points, which is very different from writers like Harari who often sound more dramatic.
Personally, I'm not worried that AI will replace me, but I understand why people with children might feel unsure about the future. They will have to help their kids grow up in a world where AI will be part of almost everything.
To me AGI still feels like a marketing idea. The real danger right now is how easily people can start to trust AI too much. These systems are built to please us because we pay for them like any other product. This can lead to real harm, and we have already seen sad cases where teenagers followed harmful advice from an AI.
AI is just another tool. It can help or it can hurt, but it is not going away. Susskind explains this in a simple and sensible way, which is why I think the book is worth reading.
Personally, I'm not worried that AI will replace me, but I understand why people with children might feel unsure about the future. They will have to help their kids grow up in a world where AI will be part of almost everything.
To me AGI still feels like a marketing idea. The real danger right now is how easily people can start to trust AI too much. These systems are built to please us because we pay for them like any other product. This can lead to real harm, and we have already seen sad cases where teenagers followed harmful advice from an AI.
AI is just another tool. It can help or it can hurt, but it is not going away. Susskind explains this in a simple and sensible way, which is why I think the book is worth reading.
April 28, 2025
For those that know Richard Susskind's previous works on the future of the law and lawyers, you will not be disappointed since this book has a similar personal tone and call to action to be a part of the discourse. I'd like a bit more details in many areas of the book but I appreciate the brevity and not getting into the technical aspects of AI which you can decide (not) to read elsewhere. He has done the hard work of plowing through the many books on the subject of where AI is heading. He justifies his predictions that AI progress will continue unabated and we need to at least plan for Artificial General Intelligence (AGI) already between 2030-2035 by engaging with cross functional experts to have a public debate about what future we want which I wholeheartedly agree with! (I was lucky enough to be at the European launch of the book before its release)
October 11, 2025
I was awed by what Susskind writes about AI. His discussion of the evolution of AI--slow and piecemeal at first; lightening fast immediately prior to and since the launch of ChatGPT, was very informative and enlightening. However, I was confused and perplexed when his discussion went beyond what I envision as the limits of AI; that is, what I personally can envision AI capable of performing. Should AI's capacity ultimately transcend the abilities of humanity's "best and brightest" (as it already has against single individuals deemed the very best, i.e., against the world's best chess master), what is to stop AI from then establishing its own detrimental goals and succeeding in achieving those goals? This is but one of many perplexing and frankly scary issues that Susskind posits. An important read for any of us perplexed by AI. And who, other than AI gurus, isn't?
August 17, 2025
The book is completely non-technical and offers very few components about the core of ai.. I liked how Richard Susskind constantly stayed on the border about AI’s boon and bane, like most people!
I don’t find the book eye opening but it’s a good read and I did have moments that slightly changed the way I thought about ai in some way.. like the way they explained about stages of ai improvement and its gradual yet unavoidable impact on society (I mean our near future)
And this book surely kinda insists me to learn and be in touch with the changing technology era and it kinda scared me that we’d be left behind as boomers if we don’t catch up with its pace!
I don’t find the book eye opening but it’s a good read and I did have moments that slightly changed the way I thought about ai in some way.. like the way they explained about stages of ai improvement and its gradual yet unavoidable impact on society (I mean our near future)
And this book surely kinda insists me to learn and be in touch with the changing technology era and it kinda scared me that we’d be left behind as boomers if we don’t catch up with its pace!
November 14, 2025
Excellent introductory read into the rapidly accelerating world of AI, explained simply, but with some serious philosophical questions on where this trajectory can take us.
Susskind puts across various possible scenarios of how the future may unravel with the assumption that machine just keep becoming smarter. The potential for humans in general to gain - from machines getting smarter - is unquestionable. However, one outcome (I wont spoil the book for others) is the possibility of machines replacing us entirely. He quotes a line from Vladimir Nobokov's autobiography's opening line "our existence is but a brief crack of light between two eternities of darkness".
Haunting.
Susskind puts across various possible scenarios of how the future may unravel with the assumption that machine just keep becoming smarter. The potential for humans in general to gain - from machines getting smarter - is unquestionable. However, one outcome (I wont spoil the book for others) is the possibility of machines replacing us entirely. He quotes a line from Vladimir Nobokov's autobiography's opening line "our existence is but a brief crack of light between two eternities of darkness".
Haunting.
December 26, 2025
This felt like it was a rushed release. The editor needs to be shot because there were numerous typos, and lots of jumping around with metaphors. This was surprising because, generally, Susskind’s writing is very good. I feel that for a book on AI, there was very little added to the current discourse. Susskind implores the reader to think “what if AGI?” But that is exactly what the last 5 years of LLM development has been. Susskind fails to address any of the relevant criticisms of whether AGI is something humanity should be aiming for. This was surprising, given he has a PhD and doesn’t seem to be ignorant to ethics discourse, he would be familiar with normative frameworks.
March 22, 2025
Melhor livro dos últimos tempos!
O livro traz uma reflexão sobre como estamos lidando mal com o tema da IA , deixando a discussão somente na mão dos técnicos e das empresas de tecnologia, quando, na realidade, o assunto deveria estar sendo discutido pelas mentes mais brilhantes, os filósofos. Traz vários cenários de futuro e tudo isso com uma linguagem acessível que mescla tecnologia, direito e filosofia. Ele está sempre citando os melhores filósofos , tanto antigos como contemporâneos, para mostrar que a questão da IA vai muito além do que muitos podem imaginar.
O livro traz uma reflexão sobre como estamos lidando mal com o tema da IA , deixando a discussão somente na mão dos técnicos e das empresas de tecnologia, quando, na realidade, o assunto deveria estar sendo discutido pelas mentes mais brilhantes, os filósofos. Traz vários cenários de futuro e tudo isso com uma linguagem acessível que mescla tecnologia, direito e filosofia. Ele está sempre citando os melhores filósofos , tanto antigos como contemporâneos, para mostrar que a questão da IA vai muito além do que muitos podem imaginar.
May 18, 2025
Susskind is among the best and most qualified to write about AI for laypeople, and should be read by the experts too. Most useful for me was the focus of the discussion on outcome (what can it do) over process (how does it work). I recently attended a conference on AI and GenZ, and the result of the conversation is always about what it means to be human, what are the skills we need to stay human, and how do we leverage AI whilst also reigning it in. This book is a superb aid in this journey of how to think about it. I hope it’s not too late to act.
February 2, 2026
In one of the better books on the topic, Susskind reminds us that our core challenges with AI aren’t technical—they’re ethical, institutional, and philosophical. Though less meaningful in some sections than others, this brief book offers a wonderful survey of the core AI questions of our time. If you assume that the technology will advance and accelerate rapidly—today it’s the worst it will ever be—then this is the time to think about how it will dramatically alter our jobs, our society, and our species.
February 10, 2026
This is not the AI book you are looking for... probably.
Some ideas are good, but gives credence to dualism despite complete lack of evidence for it and overwhelming evidence against it. He also brings up the simulation hypothesis and cites it as having a 25% probability based on literally one guy's opinion and again, without evidence.
The outcome vs. process thinking seems like a good distinction, but there are many instances where they seem to overlap or he identifies something as being an example of one but it fits equally well with the opposite.
Some ideas are good, but gives credence to dualism despite complete lack of evidence for it and overwhelming evidence against it. He also brings up the simulation hypothesis and cites it as having a 25% probability based on literally one guy's opinion and again, without evidence.
The outcome vs. process thinking seems like a good distinction, but there are many instances where they seem to overlap or he identifies something as being an example of one but it fits equally well with the opposite.
Displaying 1 - 30 of 49 reviews