Book Review: Liars and Outliers 68
First time accepted submitter benrothke writes "It is said that the song Wipe Out launched a generation of drummers. In the world of information security, the classic Applied Cryptography: Protocols, Algorithms, and Source Code in C by Bruce Schneier may have been the book that launched a generation of new cryptographers. Schneier's latest work of art is Liars and Outliers: Enabling the Trust that Society Needs to Thrive. For those that are looking for a follow-up to Applied Cryptography, this it is not. In fact, it is hard to classify this as an information security title and in fact the book is marked for the current affairs/sociology section. Whatever section this book ultimately falls in, the reader will find that Schneier is one of the most original thinkers around." Keep reading for the rest of Ben's review.
In Applied Cryptography Schneier dealt with the pristine world of mathematical cryptography where aspects of pure mathematics could be demonstrably proven. For example, non-repudiation is absolutely provable. | Liars and Outliers: Enabling the Trust that Society Needs to Thrive | |
| author | Bruce Schneier |
| pages | 384 |
| publisher | Wiley |
| rating | 10/10 |
| reviewer | Ben Rothke |
| ISBN | 978-1118143308 |
| summary | Brilliant book on trust and society, and it complex interrelation |
In Liars and Outliers, Schneier moves from the pristine world of mathematics into the muddy world of human trust. Non-repudiation is no longer an absolute in a world where a Windows kernel can be compromised and end-users can be victims of social engineering.
The book addresses the fundamental question of how does society function when you cant trust everyone. Schneier notes that nothing in society works without trust. Its the foundation of communities, commerce, democracy, in truth — everything. And Schneier deals extensively with social and moral pressures that effect trust.
Liars and Outliers is very similar to books by Umberto Eco, that have a Renaissance feel to them; bringing myriad and diverse topics together. Schneier does this here and intertwines topics such as game theory, evolution, surveillance, existentialism and much more. Schneier's brilliance is that he is able to connect seemingly disparate dots around information security and society, and show how they are in truth tightly coupled.
In the book, Schneier makes note of those that don't follow the rules. He calls these people defectors, and these are the liars and outliers of the book. The book notes that everything is a trade-off, and these defectors are the ones that try to break the rules.
An overall theme of the book, in which Schneier touches and references sociology, psychology, economics, criminology, anthropology, game theory and much more, is that society can't function without trust. He writes that in our complex interconnect and global society, that we need a lot of trust.
Schneier makes frequent reference to Dunbar's number, which he first references in chapter 2. Dunbars number was first proposed by British anthropologist Robin Dunbar and is a theoretical cognitive limit to the number of people with whom one can maintain stable social relationships. It is generally in the area of 150. So when someone sees a person with 3,000 Facebook friends, something is clearly amiss.
In chapter 9 on institutional pressures, Schneier takes a very broad look at threats facing society today. One of the biggest perceived threats we have today is terrorism, and the book astutely notes that we can never ensure perfect security against terrorism.
If Schneier had his way, the TSA budget would be measured in the millions, not billions of dollars. He incisively observes that all the talk of terrorism as an existential threat to society is utter nonsense. As long as terrorism is rare enough (which it is), and most people survive (which they do), society will survive. He writes that while that observation is true, it is not politically viable for our leaders to come out and say that.
While the book is heavy on the people focus, Schneier also acknowledges that sometimes and for some people, the incentives to commit crimes are worth the risk. To deal with those, that is where security technologies come into play.
An interesting observation made in chapter 10 around technology is that sometimes the technological changes have absolutely nothing to do with the societal dilemma being secured. For example, he notes that between the ubiquity of keyboards and the tendency for teachers to focus on standardized tests, cursive is no longer being taught that much in schools. The result is that signatures are more likely to be either printed text is an illegible scrawl; making them easier to forge; which in turns creates new security risks.
In the book Schneier makes scores of astute observations on how society functions around security. He notes in chapter 16 that we are currently in a period of history where technology is changing faster than it ever has. The worry is that if technology changes too fast, the attackers will be able to innovate so much faster than society can that the imbalance become even greater; with failures that negatively affect society.
In many of the examples in the book, Schneier paints a dark picture given the advantage that the attackers and defectors have. But he also notes that we are in a period of history where the ability for large-scale cooperation is greater than it has ever been before. On that topic, he refers to the book The Penguin and the Leviathan: How Cooperation Triumphs over Self-Interest by Yochai Benkler where he writes that the Internet can and has enabled cooperation on a scale never before seen. And that politics, backed by science, is ready to embrace this new cooperation.
On the lighter side, in chapter 17, Schneier notes that Mussolini didn't make the trains run on time; he just made it illegal to complain about them.
Schneier notes at the end of the book that its lesson isn't that defectors will inevitably ruin everything for everyone. Rather that we as a society need to manage societal pressure to ensure that they don't.
Liars and Outliers is an absolutely fascinating and groundbreaking book. In this election year where the candidates attempt to make sweeping simplistic promises to fix complex problems, Schneier simply answers that in our complex society, there are no simple answers.
In Applied Cryptography Bruce Schneier demonstrated he was quite the smart guy. In Liars and Outliers, he shows he is even smarter than most of us first thought.
Ben Rothke is the author of Computer Security: 20 Things Every Employee Should Know.
You can purchase Liars and Outliers: Enabling the Trust that Society Needs to Thrive from amazon.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
Not Quite Ben's First Rodeo (Score:5, Informative)
First time accepted submitter benrothke writes
Nahh, he just switched from brothke to benrothke. He's been reviewing as far back as 2008 [slashdot.org] with a review of The Tangled Web [slashdot.org] not even a month ago. Maybe he forgot his password?
Haters gonna hate [slashdot.org] on book reviews (including me) but keep 'em coming, brothski!
Re:Not Quite Ben's First Rodeo (Score:5, Insightful)
It's pretty good, as short reviews of books on technical subjects go. As someone who has analyzed texts my entire professional career, he has clearly reviewed books before, his user name aside.
I do enjoy the book reviews here on Slashdot. I've gotten some pretty good reads and reference books.
I'd like to see more reviews by lots of different people. Reviews of things besides books, too. I think there are a lot of Slashdot readers who have interesting and informative perspectives. RogueyWon's game reviews, for example, are as good as any I've found on the Internet.
Re: (Score:1)
Not so much a "reviewer" as a "talent scout" ... if you only write positive reviews, you're in the business of promoting books you think deserve it. On the whole that seems like a better occupation than writing screeds that pan some hapless author who probably already knows he should try another line of work.
Re: (Score:2)
Re: (Score:1)
perverted religion teaches deceptions, lies and murder
Yeah, that really narrows it down to ... all religions (except Buddhism). Oh wait I see you mentioned one by name in the first line. Sry.
Re: (Score:1)
Nonsense, trust is always limited, and is not particularly bounded by societal lines. I trust my own family to a certain extent (less that absolutely), close friends to a lesser extent, acquaintances to an even less extent, strangers of my religion to a different extent, strangers of my own profession yet another way, strangers from my hometown or ethnicity in a different way, etc. Other people will trust these groups in a different way than I do, etc.
Re: (Score:2)
Actually all of the Abrahamic religions teach deceptions, lies, and murder.
Are you saying that they teach about them? As in...they tell about people who told lies? Or are you saying that they teach their followers to do them? The reason I ask is that in Judaism the first question you are asked in the afterlife at judgement is, "were you honest in your business dealings?" In that respect honesty could be said to be one of the most important tenets of Judaism. Similarly in Christianity Jesus admonishes i
Re: (Score:2)
Actually all of the Abrahamic religions teach deceptions, lies, and murder.
Are you saying that they teach about them? As in...they tell about people who told lies? Or are you saying that they teach their followers to do them? The reason I ask is that in Judaism the first question you are asked in the afterlife at judgement is, "were you honest in your business dealings?" In that respect honesty could be said to be one of the most important tenets of Judaism. Similarly in Christianity Jesus admonishes its adherents to be completely truthful, as in not being more truthful when swearing an oath.
~Loyal
It is only Islam that teaches followers to lie and to be fair it is in the Hadith rather than the Qur'an.
Re: (Score:2)
Christianity is the following of the teachings of Christ. However, being a Christian simply means a belief or faith in those teachings, not one's ability to follow them.
As the old story goes, our church may be full of hypocrites, but there's always room for one more.
Parsing Error (Score:5, Funny)
The result is that signatures are more likely to be either printed text is an illegible scrawl; making them easier to forge; which in turns creates new security risks.
A couple related sentences seem to have collided here...
Re: (Score:3)
Re: (Score:1)
Defectors aren't all bad (Score:5, Insightful)
While Trust definitely lowers the cost of transactions between people/entities, I think that having a small amount of defectors is actually beneficial on a macro level. Without a couple people willing to take advantage of process flaws, it would be very easy for society to become stagnant and complacent.
Doesn't mean we shouldn't crush those defectors to dust when we find them, though.
Re: (Score:2)
You just described the plot of "The Invention of Lying", a clever Ricky Gervais movie about a society where trust is 100%, and he discovers he can abuse this by lying.
Re: (Score:3)
Which is precisely the problem.
How do you protect yourself from liars? Do you begin lying yourself? The short term advantages of one-skilled liar can be so detrimental to others, that by the time the long-term disadvantages of lying catch up, everyone else has converted to lying to keep themselves alive.
As principled as you may imagine yourself to be, duress, like most forms of torture, can force a more pragmatic approach.
A recent study with machines, posted on reddit, showed that when they were forced to c
Re: (Score:2)
How do you protect yourself from liars? Do you begin lying yourself?
Actually, this has been dealt with by the game theory folks. It has also been tested in Real Life by contests pitting implementations of various strategies in games that pairs strategies against each other, and rewards/punishes them depending on various payoff functions. The most popular have been Repeated Prisoner's Dilemma games, in which the PD payoff matrix is used, and "players" are allowed to remember what each opponent did in previous encounters.
The simplest strategy that consistently wins again
Re: (Score:2)
Read the book - this is exactly one of Schneier's points.
It's even covered in the available-on-the-web Chapter 1 [schneier.com].
First Post! (Score:1)
Nah. I am a liar.
If I am the last to claim FP, does that make me an outlier?
Re: (Score:2)
Dunbar and his number (Score:3)
Dunbar gets made fun of sometimes because of the whole correlation/causation thing. Its just as sensible that social size and neocortex size depend on food pressure.
There are also the weird equivalents of the squared/cubed law
Also primate behavior is pretty boring compared to human, so Dunbar gets made fun of because of what amounts to FTE full time equivalent accounting. Surely the relationship I have with my cube neighbor is not exactly the same as yours, resulting in a larger or smaller chunk of the brain necessary to process... So 150 average full time equivalents, but someone really shallow might do the mentioned 3000 FB friends because each, on average, is only 1/20th of an average relationship. This is actually a well known FB problem, for example I don't do the 1/20th of a real relationship thing, so I found FB incredibly annoying, I don't give a F about some kid I sat next to in study hall 20 years ago, so I deleted my account from all the friendspam. A social media network 2.0 thingy that understands that would be interesting... G+ and its circle sliders is pretty close.
Re: (Score:2)
My understanding is that this is exactly what FB has been trying to accomplish with their Top Stories/whatever crap they keep doing in each new update. Even though someone has 3000 friends, only the forty or so that person interacts with show up in the feed.
Of course, that's what pisses a lot off us off about FB - we are already control freaks, and if we don't want to hear about a person, we just block them. My 92-year-old great aunt through marriage is on my FB friends list, and even though I never comme
Slower not faster (Score:5, Insightful)
In the book Schneier makes scores of astute observations on how society functions around security. He notes in chapter 16 that we are currently in a period of history where technology is changing faster than it ever has. The worry is that if technology changes too fast, the attackers will be able to innovate so much faster than society can
Historically haven't attackers always innovated slower at least on the net? I was on the net for years before the invention of spam around '93 or so. Most "attacks" seem to be the same old social con artist crimes, or finding dumb coding mistakes... but for those in the biz, those dumb mistakes are defined as dumb, not insightful. A buffer overflow is freaking magic to a noob, but to a guy who knows C its a parlor trick.
Technologically enhanced stupidity on the victim side seems to be a bigger issue than technologically enhanced criminality on the attacker side.
In 20 years on the net, I've seen the victims get stupider, but I haven't seen the attackers get smarter or dumber. You'd think the same demographic pressure would apply to each, but...
Re: (Score:2, Insightful)
You're looking at it from the wrong perspective. "Attackers", manipulators, and conmen have always and will always be trying the best schemes they can find. When they find one to be profitable, they don't NEED to move on until the scheme has been foiled and its chances of success drop to the point where it's more profitable to move on to a new scheme than to continue entertaining an existing, [semi-]successful one.
Find an easy way to make money, and you will find the place swarming with conmen, abusers, a
Re: (Score:1)
actually, attackers generally innovate faster than society or "good guys" do. that's the whole reason that zero-day vulnerabilities occur before the security loopholes are patched. plus, if you don't think attackers are getting smarter, you haven't been paying attention (e.g., Stuxnet, Anonymous eavesdropping on FBI conversations, etc.).
Re: (Score:3)
Re: (Score:3)
Look at OWASP and SANS. Every entry in their databases indicates someone who got smarter and figured out a new hole to exploit. And their databases are growing as attackers learn more and exploit more.
And yes, coders are still making mistakes that others have made. They have failed to learn, and that's what OWASP is trying to prevent. But there are certainly new attacks coming out. Why? Because as systems grow in complexity, unforeseen interactions between the components create vulnerabilities. And t
Re: (Score:2)
Re: (Score:2)
Every entry in their databases indicates someone who got smarter and figured out a new hole to exploit.
Moving to a different hole in a colander is not the same as getting smarter.
Trust requires accountability (Score:3, Funny)
Common Law America used to have quite a bit more social trust. And people used to live with more integrity.
Of course, we had better social enforcement. Rude curs would be shunned as "no-account trash", or dueled if they continued to deliberately give offense.
Being free, any citizen was allowed to reject the law and opt not to be held accountable to it. Of course, they were also not protected by the law either, and if these "outlaws" began encroaching against society (robbing stages, kidnapping and the like) they had only the protections of natural law (kill or be killed, the law of the wild) to protect them. Society's members were free to contribute to a large bounty and hunt them down like rabbits.
It's these social costs for giving offenses that prevented everyone from doing that. While I can't agree with the puritanical basis in whole, most of those social costs have now been removed from our society. With the penalties gone, acting out and giving offense has become prevalent. And trust correspondingly takes a hit when we no longer concern ourselves with whether we've transgressed against another person, and merely worry about whether we've transgressed against the State and are liable for punishment by it.
Re: (Score:1)
I'm sorry, when exactly was this? You seem to have mistaken mythology for history...
Outliers (Score:1)
This sounds like a cross between Kevin Mitnick's "The Art of Deception" and Malcolm Gladwell's ("The Tipping Point") "The Outliers". Probably should take an interest in any of these if you take an interest in any of them.
Re: (Score:2)
Remember the Morrisons in _Scientific American_? (Score:2)
I guarantee you they were in no sense cheerleaders or shills. They jus
Re: (Score:2)
Here you go http://www.utoronto.ca/ethnicstudies/Putnam.pdf [utoronto.ca]
Is this racist dribble? Dunno, maybe.
Basically Mr Putnam advances a theory about a postulated metric of social capital and how much you claim that you trust your neighbor and how much that socally isolates you. The data is mostly extrapolated from aggregate data and the author even admits that many of the speculations do not follow from data that passes statistical significance as he makes the leap from corellating trust with aggregate neighborhood
Re: (Score:2)
Cost of mistrust (Score:1)
This is my observation having spent some time in IT security of financial firms, and I wouldn't be surprised if those with experience in other security arena (physical, law enforcement, etc.) come to make similar observation.
In the end, increasing the general level of trust in the wh
Not unrelated (Score:3)
I have not read Lies and Outliers, expect for the excerpt online, but I have followed the writing on the blog. It seems to me that this is another book that promotes and explains rational security. Really that was all that Applied Cryptography did. Explain what to do with the tools and knowledge we had. Most security systems are merely only façades. Door locks are not that hard to disable. Checks are not all that hard to forge. They exist to put a layer of ritual between others and our secrets, and it is beneficial for most of us to respect those rituals. What is left, then, is what to do with the those that do not respect ritual, or, even more dangerous, appear to respect the ritual but really are just abusing the rituals to maximize personal returns at the expense of the community.
The end, as always, is an efficient security that does not cause more trouble than it is worth. I think of all the alarms that used to go off every time a car was approached. Of course alarms did not really stop car theft. Most people just ignored them. Modern methods that do not destroy the civil tone of society tends to be more productive.
Re: (Score:3)
Follow-up to Applied Cryptography?This it is not! (Score:2)
I love Yoda's book reviews!
For those that are looking for a follow-up to Applied Cryptography, this it is not
(On a more serious note - this is an great review of a really interesting book - thanks for posting it! :) )
we're all defectors (Score:1)
It's all a matter of degree (Score:2)
Re: (Score:2)
It seemed germane to me—how society deals with defectors is an important part of the whole.