Once a week.
Subscribe to our weekly newsletter.
The 10 best books about A.I.
Here are some of the best books on the rich history, rabid speculations and intriguing fictionalized world of artificial intelligence.
Artificial intelligence has been the stuff of mad dreams, and sometimes nightmares, throughout our collective history. We've come a long way from a 15th-century automaton knight crafted by Leonardo da Vinci. Within the past century, artificial intelligence has inched itself further into our realities and day to day lives and there is now no doubt we're entering into a new age of intelligence.
Early computing technology ushered in a new branch of computer science dealing with the simulated intelligence of machines. In recent history, we've used A.I. for common tasks, such as playing against the computer in chess matches and other gameplay behaviors. Thanks to advances in computer technology, however, A.I. is advancing now 10 times faster than it had before. A.I. can also be defined as the capability of a machine to imitate human behavior. The holy grail—and what most people probably think about when they hear A.I.—is something called artificial general intelligence or AGI. With advanced AGI, robots would, in theory, be able to do anything a human could do. This usually conjures up pictures of anthropomorphic robot butlers or Westworld type androids.
One can get lost in the rich history, rabid speculations, and intriguing fictionalized world of A.I. Here are some of the best books on artificial intelligence that'll guide you and provide a multifaceted view of this incredible technology.
We have personal assistants in our pockets and on our desks. Automated factories and self-driving cars being tested on the daily. We're living in the imagined futures of our past. Here's some books to help us get to the next step.
Author and journalist John Markoff offers a detailed and rich history of the field of robotics and artificial intelligence. He's interested in the interplay between human societies and the effect that increasing automation and intelligence will have in the workplace and other areas. Markoff explores the views of the designers behind these machines and the paradoxical nature of the potentialities of this new tech. He states: “The same technologies that extend the intellectual power of humans can displace them as well"
Markoff writes about the concept of AI vs IA – or artificial intelligence versus intelligence automation. He believes that this and building a human element into the technologies is one of the more important aspects of this field.
Isaac Asimov's Robot Series
Science fiction often has a way of not only predicting the future, but preparing us for it as well. During the Golden Age of Science Fiction, Issac Asimov, along with Arthur C. Clarke and Robert Heinlein, was part of “the big three" in the sci-fi genre. These authors covered topics ranging from philosophical free-love martians to spaceship robots having mental breakdowns.
Asimov's contribution paved the way for something many robot ethicists and people are still thinking about today: The Three Laws of Robotics. These were immortalized in his anthology and Robot novel series.
1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.
The complete robot series includes all of Asimov's short fiction and novellas and if you're interested in his novels then start with Caves of Steel and follow that to his last robot novel: Robots and Empire. Almost every common robotics trope we have in our culture today was seeded by Asimov.
How to Create a Mind: The Secret of Human Thought Revealed
Ray Kurzweil was arguably responsible for popularizing the idea of the technological singularity. In which, much to the hopes of sun-deprived dreamers everywhere that the end was nigh for a super-intelligence would eventually emerge out of a self-improving runaway reaction. With it there'd be no telling what comes next in a world unknowable, ending history in its wake and bringing in a time of unfathomable change. While not all of Kurzweil's predictions are as verbose, many of Kurzweil's lesser predictions from years past have rung true.
By combining an investigation into the neocortex, study of language and the development of AI, this book dives into many thought-provoking lines of inquiry. How to Create a Mind is part twenty-first century AI treatise and part philosophical doctrine on the implications of unlocking the unbridled power of artificial intelligence.
While we can sometimes get lost within the existential or even eschatological ramifications (sorry Kurzweil!) of artificial intelligence, we're still very much in the infancy of the technology. The early seeds of this enterprise still require human ingenuity and some good old-fashioned hard coding.
Python Machine Learning
Many of the breakthroughs in current A.I. research come from applications of machine learning. Python has proved to be an excellent language to use to further one's knowledge in the field. This book includes a Github link and helps teach the fundamentals of machine learning through Python.
The book sets out to teach:
- Development of Learning Algorithms
- Transforming Raw Data into Useful Information
- Classifying Objects
- Regression Analysis
Author Sebastian Raschka sets out step-by-step examples of real-world machine learning applications. You can follow along and by the end of the book have the skills necessary to implement your own machine learning system for sentiment analysis in a live web app.
Deep Learning with R
Written by Google A.I. researcher and Keras library creator François Chollet, Deep Learning with R sets out to teach the fundamentals of deep learning in the R programming language. It serves as an excellent instruction manual for Keras and can be supplemented with more theoretical works for a fuller picture of the discipline.
- Deep learning from first principles
- Setting up your own deep-learning environment
- Image classification and generation
- Deep learning for text and sequences
Readers should have an intermediate skill level with R programming. You won't need any previous experience with machine learning, but this paired with another practical book like Python Learning is a plus.
Artificial Intelligence: A Modern Approach (3rd Edition)
You all knew we were going to get to this one. An accessible book for undergraduate or graduate level students in Artificial intelligence, this is a standard across in academia. Written by Peter Norvig, Artificial Intelligence: a Modern Approach has been used in over 1200 universities and is one of the most highly cited publications.
Many students find it an accessible read as it's also approachable for laymen, scientists, and tinkerers too. As a best seller in the field, the text offers a comprehensive and constantly updated stream of information to the theory and growing practice of A.I.
Ethics and philosophical musings
Enough movies and stories have been written about the potential pitfalls of this tech. Metropolis occult rituals and Arnold Schwarzenegger terminators have left some people weary and a bit scared. But that hasn't stopped us yet, has it?
Superintelligence: Paths, Dangers, Strategies
Nick Bostrom's highly ambitious book has already become a classic. It raises more questions than it answers – as any great philosophical work should. Bostrom wonders about what will happen in the world when machines surpass humans in general intelligence. Will these new beings be our downfall or our saving grace?
On the subject of our origins of intelligence, Bostrom writes: “We know that blind evolutionary processes can produce human-level general intelligence since they have already done so at least once. Evolutionary processes with foresight—that is, genetic programs designed and guided by an intelligent human programmer—should be able to achieve a similar outcome with far greater efficiency."
Bostrom assumes the mantle of 21st-century techno-sage, breathing life into a real possibility of a much greater intelligence in this universe.
Life 3.0: Being Human in the Age of Artificial Intelligence
A favorite book of Elon Musk and one in which Musk was inspired to describe developing advanced A.I. as akin to “summoning the demon."
Max Tegmark, an MIT professor (and Big Think guest), covers a wide range of spectrums in the potential future of AI. It's a little watered down compared to Bostrom's Superintelligence. But it gets the point across that things are changing rapidly. What it does best is it helps inform the layperson in areas of impact that will be affected most by AI in our day to day lives.
Robot Ethics: The Ethical and Social Implications of Robotics
The first book in a two-part series, Robot Ethics is a starting point for learning how to design and hard-code ethics into robotic artificial systems. As ethical concerns become even more pressing, roboticists and other technologists need to start thinking about what they can do now to stop any future catastrophic events.
Some of the covered topics include:
- Exploring Emotional Bonds With Robots
- Programming A Code of Ethics
- Ethical Military Use in War
- Liability and Privacy Concerns
Regulation often comes too late with new technologies. But the authors of this book state that we must reverse this trend. Ethics and regulations need to be on the forefront for such a game-changing tech.
Our Final Invention: Artificial Intelligence and the End of the Human Era
Author James Barrat explores the many governments and agencies around the world researching AGI and pouring billions into funding. Many scientists believe that once this lofty goal has been reached, these machines will have similar survival drives as we do.
Barratt explores this line of thinking and takes it to its logical conclusion as many others have. A superintelligent AGI would become an alien-like entity to the human race.
James wonders if we would serve any further purpose to a being as powerful as this. Would it begin to suck up as much matter and energy as it could for its foreign purposes and simply brush us aside? It's a despondent and completely apocalyptic view of what may happen, bringing up images of The Matrix and other fictional worlds bereft of biological life.
- Why aligning AI to our values may be harder than you think - Big Think ›
- 10 ways we should prepare for rise of intelligent machines – MIT study - Big Think ›
- 5 eBook bundles that can raise your earning potential - Big Think ›
- Why democratizing AI is absolutely crucial - Big Think ›
- Computing history: From government secrets to a failed tech utopia - Big Think ›
- Artificial intelligence beats humans at growing strawberries - Big Think ›
Some mysteries take generations to unfold.
- In 1959, a group of nine Russian hikers was killed in an overnight incident in the Ural Mountains.
- Conspiracies about their deaths have flourished ever since, including alien invasion, an irate Yeti, and angry tribesmen.
- Researchers have finally confirmed that their deaths were due to a slab avalanche caused by intense winds.
In February 1959, a group of nine hikers crossed through Russia's Ural Mountains as part of a skiing expedition. The experienced trekkers, all employed at the Ural Polytechnical Institute, were led by Igor Dyatlov. On the evening of February 1, all nine appear to have fled their tents into the Arctic temperatures, for which they were unprepared. None survived.
Six of the members died of hypothermia; three suffered from physical trauma. Some members were missing body parts—a tongue here, a few eyes there, a pair of eyebrows for good measure. According to reports, no hiker appears to have struggled or panicked. They were likely too quickly overtaken by the hostile environment in Western Russia.
All the members were young, mostly in their early twenties; one member, Semyon Zolotaryov, was 38. Good health didn't matter. Given the uncertain circumstances—what made them flee into the bitter cold?—the incident known as Dyatlov Pass has long been the type of Area 51-conspiracy theory that some people love to speculate about. A vicious animal attack? Infrasound-induced panic? Was the Soviet military involved? Maybe it was the katabatic winds that did them in. Local tribesmen might not have liked the intrusion.
Or perhaps it was aliens. Or a Yeti. Have we talked about Yeti aliens yet?
These theories and more have been floated for decades.
a: Last picture of the Dyatlov group taken before sunset, while making a cut in the slope to install the tent. b: Broken tent covered with snow as it was found during the search 26 days after the event.
Photographs courtesy of the Dyatlov Memorial Foundation.
Finally, a new study, published in the Nature journal Communications Earth & Environment, has put the case to rest: it was a slab avalanche.
This theory isn't exactly new either. Researchers have long been skeptical about the avalanche notion, however, due to the grade of the hill. Slab avalanches don't need a steep slope to get started. Crown or flank fractures can quickly release as little as a few centimeters of earth (or snow) sliding down a hill (or mountain).
As researchers Johan Gaume (Switzerland's WSL Institute for Snow and Avalanche Research SLF) and Alexander Puzrin (Switzerland's Institute for Geotechnical Engineering) write, it was "a combination of irregular topography, a cut made in the slope to install the tent and the subsequent deposition of snow induced by strong katabatic winds contributed after a suitable time to the slab release, which caused severe non-fatal injuries, in agreement with the autopsy results."
Conspiracy theories abound when evidence is lacking. Twenty-six days after the incident, a team showed up to investigate. They didn't find any obvious sounds of an avalanche; the slope angle was below 30 degrees, ruling out (to them) the possibility of a landslide. Plus, the head injuries suffered were not typical of avalanche victims. Inject doubt and crazy theories will flourish.
Configuration of the Dyatlov tent installed on a flat surface after making a cut in the slope below a small shoulder. Snow deposition above the tent is due to wind transport of snow (with deposition flux Q).
Photo courtesy of Communications Earth & Environment.
Add to this Russian leadership's longstanding battle with (or against) the truth. In 2015 the Investigative Committee of the Russian Federation decided to reopen this case. Four years later the agency concluded it was indeed a snow avalanche—an assertion immediately challenged within the Russian Federation. The oppositional agency eventually agreed as well. The problem was neither really provided conclusive scientific evidence.
Gaume and Puzrin went to work. They provided four critical factors that confirmed the avalanche:
- The location of the tent under a shoulder in a locally steeper slope to protect them from the wind
- A buried weak snow layer parallel to the locally steeper terrain, which resulted in an upward-thinning snow slab
- The cut in the snow slab made by the group to install the tent
- Strong katabatic winds that led to progressive snow accumulation due to the local topography (shoulder above the tent) causing a delayed failure
Case closed? It appears so, though don't expect conspiracy theories to abate. Good research takes time—sometimes generations. We're constantly learning about our environment and then applying those lessons to the past. While we can't expect every skeptic to accept the findings, from the looks of this study, a 62-year-old case is now closed.
Stay in touch with Derek on Twitter and Facebook. His most recent book is "Hero's Dose: The Case For Psychedelics in Ritual and Therapy."
New data have set the particle physics community abuzz.
- The first question ever asked in Western philosophy, "What's the world made of?" continues to inspire high energy physicists.
- New experimental results probing the magnetic properties of the muon, a heavier cousin of the electron, seem to indicate that new particles of nature may exist, potentially shedding light on the mystery of dark matter.
- The results are a celebration of the human spirit and our insatiable curiosity to understand the world and our place in it.
If brute force doesn't work, then look into the peculiarities of nothingness. This may sound like a Zen koan, but it's actually the strategy that particle physicists are using to find physics beyond the Standard Model, the current registry of all known particles and their interactions. Instead of the usual colliding experiments that smash particles against one another, exciting new results indicate that new vistas into exotic kinds of matter may be glimpsed by carefully measuring the properties of the quantum vacuum. There's a lot to unpack here, so let's go piecemeal.
It is fitting that the first question asked in Western philosophy concerned the material composition of the world. Writing around 350 BCE, Aristotle credited Thales of Miletus (circa 600 BCE) with the honor of being the first Western philosopher when he asked the question, "What is the world made of?" What modern high energy physicists do, albeit with very different methodology and equipment, is to follow along the same philosophical tradition of trying to answer this question, assuming that there are indivisible bricks of matter called elementary particles.
Deficits in the Standard Model
Jumping thousands of years of spectacular discoveries, we now have a very neat understanding of the material composition of the world at the subatomic level: a total of 12 particles and the Higgs boson. The 12 particles of matter are divided into two groups, six leptons and six quarks. The six quarks comprise all particles that interact via the strong nuclear force, like protons and neutrons. The leptons include the familiar electron and its two heavier cousins, the muon and the tau. The muon is the star of the new experiments.
For all its glory, the Standard Model described above is incomplete. The goal of fundamental physics is to answer the most questions with the least number of assumptions. As it stands, the values of the masses of all particles are parameters that we measure in the laboratory, related to how strongly they interact with the Higgs. We don't know why some interact much stronger than others (and, as a consequence, have larger masses), why there is a prevalence of matter over antimatter, or why the universe seems to be dominated by dark matter — a kind of matter we know nothing about, apart from the fact that it's not part of the recipe included in the Standard Model. We know dark matter has mass since its gravitational effects are felt in familiar matter, the matter that makes up galaxies and stars. But we don't know what it is.
Whatever happens, new science will be learned.
Physicists had hoped that the powerful Large Hadron Collider in Switzerland would shed light on the nature of dark matter, but nothing has come up there or in many direct searches, where detectors were mounted to collect dark matter that presumably would rain down from the skies and hit particles of ordinary matter.
Could muons fill in the gaps?
Enter the muons. The hope that these particles can help solve the shortcomings of the Standard Model has two parts to it. The first is that every particle, like a muon, that has an electric charge can be pictured simplistically as a spinning sphere. Spinning spheres and disks of charge create a magnetic field perpendicular to the direction of the spin. Picture the muon as a tiny spinning top. If it's rotating counterclockwise, its magnetic field would point vertically up. (Grab a glass of water with your right hand and turn it counterclockwise. Your thumb will be pointing up, the direction of the magnetic field.) The spinning muons will be placed into a doughnut-shaped tunnel and forced to go around and around. The tunnel will have its own magnetic field that will interact with the tiny magnetic field of the muons. As the muons circle the doughnut, they will wobble about, just like spinning-tops wobble on the ground due to their interaction with Earth's gravity. The amount of wobbling depends on the magnetic properties of the muon which, in turn, depend on what's going on with the muon in space.
Credit: Fabrice Coffrini / Getty Images
This is where the second idea comes in, the quantum vacuum. In physics, there is no empty space. The so-called vacuum is actually a bubbling soup of particles that appear and disappear in fractions of a second. Everything fluctuates, as encapsulated in Heisenberg's Uncertainty Principle. Energy fluctuates too, what we call zero-point energy. Since energy and mass are interconvertible (E=mc2, remember?), these tiny fluctuations of energy can be momentarily converted into particles that pop out and back into the busy nothingness of the quantum vacuum. Every particle of matter is cloaked with these particles emerging from vacuum fluctuations. Thus, a muon is not only a muon, but a muon dressed with these extra fleeting bits of stuff. That being the case, these extra particles affect a muon's magnetic field, and thus, its wobbling properties.
About 20 years ago, physicists at the Brookhaven National Laboratory detected anomalies in the muon's magnetic properties, larger than what theory predicted. This would mean that the quantum vacuum produces particles not accounted for by the Standard Model: new physics! Fast forward to 2017, and the experiment, at four times higher sensitivity, was repeated at the Fermi National Laboratory, where yours truly was a postdoctoral fellow a while back. The first results of the Muon g-2 experiment were unveiled on 7-April-2021 and not only confirmed the existence of a magnetic moment anomaly but greatly amplified it.
To most people, the official results, published recently, don't seem so exciting: a "tension between theory and experiment of 4.2 standard deviations." The gold standard for a new discovery in particle physics is a 5-sigma variation, or one part in 3.5 million. (That is, running the experiment 3.5 million times and only observing the anomaly once.) However, that's enough for plenty of excitement in the particle physics community, given the remarkable precision of the experimental measurements.
A time for excitement?
Now, results must be reanalyzed very carefully to make sure that (1) there are no hidden experimental errors; and (2) the theoretical calculations are not off. There will be a frenzy of calculations and papers in the coming months, all trying to make sense of the results, both on the experimental and theoretical fronts. And this is exactly how it should be. Science is a community-based effort, and the work of many compete with and complete each other.
Whatever happens, new science will be learned, even if less exciting than new particles. Or maybe, new particles have been there all along, blipping in and out of existence from the quantum vacuum, waiting to be pulled out of this busy nothingness by our tenacious efforts to find out what the world is made of.
- Benjamin Franklin wrote essays on a whole range of subjects, but one of his finest was on how to be a nice, likable person.
- Franklin lists a whole series of common errors people make while in the company of others, like over-talking or storytelling.
- His simple recipe for being good company is to be genuinely interested in others and to accept them for who they are.
Think of the nicest person you know. The person who would fit into any group configuration, who no one can dislike, or who makes a room warmer and happier just by being there.
What makes them this way? Why are they so amiable, likeable, or good-natured? What is it, you think, that makes a person good company?
There are really only two things that make someone likable.
This is the kind of advice that comes from one of history's most famously good-natured thinkers: Benjamin Franklin. His essay "On Conversation" is full of practical, surprisingly modern tips about how to be a nice person.
Franklin begins by arguing that there are really only two things that make someone likable. First, they have to be genuinely interested in what others say. Second, they have to be willing "to overlook or excuse Foibles." In other words, being good company means listening to people and ignoring their faults. Being witty, well-read, intelligent, or incredibly handsome can all make a good impression, but they're nothing without these two simple rules.
The sort of person nobody likes
From here, Franklin goes on to give a list of the common errors people tend to make while in company. These are the things people do that makes us dislike them. We might even find, with a sinking feeling in our stomach, that we do some of these ourselves.
1) Talking too much and becoming a "chaos of noise and nonsense." These people invariably talk about themselves, but even if "they speak beautifully," it's still ultimately more a soliloquy than a real conversation. Franklin mentions how funny it can be to see these kinds of people come together. They "neither hear nor care what the other says; but both talk on at any rate, and never fail to part highly disgusted with each other."
2) Asking too many questions. Interrogators are those people who have an "impertinent Inquisitiveness… of ten thousand questions," and it can feel like you're caught between a psychoanalyst and a lawyer. In itself, this might not be a bad thing, but Franklin notes it's usually just from a sense of nosiness and gossip. The questions are only designed to "discover secrets…and expose the mistakes of others."
3) Storytelling. You know those people who always have a scripted story they tell at every single gathering? Utterly painful. They'll either be entirely oblivious to how little others care for their story, or they'll be aware and carry on regardless. Franklin notes, "Old Folks are most subject to this Error," which we might think is perhaps harsh, or comically honest, depending on our age.
4) Debating. Some people are always itching for a fight or debate. The "Wrangling and Disputing" types inevitably make everyone else feel like they need to watch what they say. If you give even the lightest or most modest opinion on something, "you throw them into Rage and Passion." For them, the conversation is a boxing fight, and words are punches to be thrown.
5) Misjudging. Ribbing or mocking someone should be a careful business. We must never mock "Misfortunes, Defects, or Deformities of any kind", and should always be 100% sure we won't upset anyone. If there's any doubt about how a "joke" will be taken, don't say it. Offense is easily taken and hard to forget.
On practical philosophy
Franklin's essay is a trove of great advice, and this article only touches on the major themes. It really is worth your time to read it in its entirety. As you do, it's hard not to smile along or to think, "Yes! I've been in that situation." Though the world has changed dramatically in the 300 years since Franklin's essay, much is exactly the same. Basic etiquette doesn't change.
If there's only one thing to take away from Franklin's essay, it comes at the end, where he revises his simple recipe for being nice:
"Be ever ready to hear what others say… and do not censure others, nor expose their Failings, but kindly excuse or hide them"
So, all it takes to be good company is to listen and accept someone for who they are.
Philosophy doesn't always have to be about huge questions of truth, beauty, morality, art, or meaning. Sometimes it can teach us simply how to not be a jerk.
A recent study analyzed the skulls of early Homo species to learn more about the evolution of primate brains.