LESSWRONG
is fundraising!
LW
$

Comment Permalink
Cyan30

This is a good question for one to ask one's self. That said, juliawise nailed it. If I had seen your reply before juliawise posted, I would have said, "'Disorder' imples the individual's map doesn't match the territory (well ok, consensus reality) in conspicuous ways, e.g., schizophrenia with paranoid delusions."

The killer apparently identified himself to police as "the Joker" (with hair dyed red or orange instead of the comic book character's green). He also rigged his apartment with booby traps but then told police about them. I'm not getting a strong sense of coherent goal-seeking here, in spite of the reported fact that the venue of the attacks was apparently chosen for maximum killing.

Reply
Jayson_Virissimo30

Yeah, I see it now. I'm pretty sure I would anticipate them having a significantly more faulty web of belief than non-serial killers.

Reply
2Decius
If you think that the venue was chosen for maximum killing, you haven't considered what someone who was optimizing for killing would do.
See in context
by James_Miller
1 min read
117

7

 

A Ph.D student in neuroscience shot at least 50 people at a showing of the new Batman movie.  He also appears to have released some kind of gas from a canister.  Because of his educational background this person almost certainly knows a lot about molecular biology.  How long will it be (if ever) before a typical bio-science Ph.D will have the capacity to kill, say,a million people?

 

Edit:  I'm not claiming that this event should cause a fully informed person to update on anything.  Rather I was hoping that readers of this blog with strong life-science backgrounds could provide information that would help me and other interested readers assess the probability of future risks.  Since this blog often deals with catastrophic risks and the social harms of irrationality and given that the events I described will likely dominate the U.S. news media for a few days I thought my question worth asking.  Given the post's Karma rating (currently -4), however, I will update my beliefs about what constitutes an appropriate discussion post.

 

Personal Blog

7

Mass-murdering neuroscience Ph.D. student
New Comment
117 comments, sorted by
top scoring
Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings
[-]RobertLumley390

In other news, over 91,000 people have died since midnight EST.

Reply
2Shmi
Most everyone dies sooner or later, artificially and knowingly making it sooner is where the ethical and legal issues start.
[-]RobertLumley140

It's where the legal issues start, certainly. But I would argue that ethically, what matters is how easily any of those 91,012 lives could have been saved. And many could have been saved very easily with malaria nets.

Reply
6FiftyTwo
Just did, thanks for the reminder. Maybe we should put together an LW donations page on there to link to and encourage donation via peer pressure?
2A1987dM
I'm pretty sure that it's a hell of a lot easier to avoid shooting at people in a cinema than to earn enough money for the AMF to save a dozen lives. I do the former all the time -- in fact, I'm doing that right now as I'm typing.
8RobertLumley
Yes, but it's a lot harder for us as a society to prevent people from doing all random acts of violence like that.
0DanArmak
It's much easier to directly save one person from malaria, than to save one person from a mad gunman. Not just on the societal level, as RobertLumley stated, but as simple individual actions.
0A1987dM
I guess I misunderstood his point, then. I took "ethically, what matters" to mean 'what matters to the question how bad a guy the gunman was, compared to how bad a guy or gal the typical person is'. There was an action X the gunman could have done such that, if counterfactually that day the gunman had done X instead of what he actually did, at the end of the day there would have been 12 fewer dead people -- namely, staying out of the cinema. There was no such obvious action in my case -- at least, none which wouldn't have left me in several thousand dollars of debt.
0A1987dM
How do you know it wasn't 91,011, or 91,013? :-)
6RobertLumley
I was just adding the 90,000 figure I used and then the 12. It was rhetorical.
1A1987dM
Of course! [slaps forehead]
-4Shmi
Seems like you have an ax to grind and so are getting completely off-topic. Time for me to disengage.
1RobertLumley
I'm curious as to why this has been downvoted. Isn't tapping out generally considered the polite thing to do around here?
1Shmi
Hmm, I'm guessing that it is because yours and EY's stance (if I understand it right) is along the lines of "every life is sacred, every life is great" and is a common sentiment on LW; and that's why your comment was upvoted and mine downvoted (probably misunderstood as "he has no valid argument to offer and so disguised this fact by tapping out"). Again, this is only a guess.
2DanArmak
I certainly don't think consciously, or act as if, "every life is sacred, every life is great". Nevertheless the people I care about personally, and myself, are still far more likely to die from some disease that is curable but is not eradicated due to lack of funds - including most or all causes of natural death - than due to the actions of madmen, gunmen, evil biology professors, or their tiny intersection. Which is why when I read news like in this post, I think: "why am I wasting my time thinking about this?"
0RobertLumley
Hm. Well, FWIW, I don't think it should have been.
-10Decius
[-]CarlShulman170

I think this is only a very small update to our picture of spree killer or terrorist demographics. We already know about the enrichment of terrorism with engineers, and Aum Shinrikyo had access to some smart generalists (neuroscience is not a directly deadly field). We also know that such folk are much more likely to succeed at super-simple plans, like this one, than at doing complex technological endeavours for the first time in the face of countermeasures.

How long will it be (if ever) before a typical bio-science Ph.D will have the capacity to kill, say,a million people?

That doesn't have to ever happen. Ubiquitous DNA sensors that identify deadly agents, DNA vaccines, ubiquitous surveillance at the planning stage, tripwires in synthesis machines, and so forth can proliferate. The typical bio-science PhD is pretty crappy (Sturgeon's Law), and likely to fail at something like making the smallpox virus unless there is a cookie-cutter script kiddy style approach pre-packaged. That might happen, but we also might see varied government interference that makes failure much more likely, just as interference at every stage of potential nuclear terrorism makes it impracticable (although that is eased by the rare materials).

Reply
4buybuydandavis
It doesn't have to happen. Certainly, if you take defensive measures, it's less likely to happen. But warfare has always been an offensive/defensive battle. Offense has seemed to have the advantage for a while, and random violence in a free society has a huge advantage as well. 1K seems easy. 10K takes some skill. 1000k seems unlikely without a very contagious bug with just the right incubation time. "Ubiquitous surveillance at the planning stage" - that sounds rather ominous there, Big Brother. What's more amazing to me is how little damage crazies do, when it just doesn't seem like it would be that hard.
3wedrifid
1k seems easy? That is, killing 1k people? That seems possible, but not easy. At very least it deserves a "takes some skill and a huge amount of dedicated planning".
2buybuydandavis
I started replying, but I just don't think it's worth it. I have no particularly brilliant idea here, and likely you'd still just disagree, but I don't see mileage in giving people ideas about this on the internet just so we can chat.
2[anonymous]
Sensors and vaccines all rely on that the pathogen is known, even coded. You could have a register of virulent genes, in order for it to be effective against hybrids, but even then it would be very hard to say whether a something is dangerous or not.
1CarlShulman
It is much harder to make and test a new deadly pathogen than to create smallpox from the genome, and avoiding even known dangerous genes forces evildoers to do real science.
0DanArmak
The whole premise of the original question was that the evildoers are capable of doing real science, within the limits of one or a few people, years, and personal budgets and tools.
0[anonymous]
"the genome" wouldn't help really be of much use, in order creating pox from a genome you need to transplant the genome into a similar bacteria, which is no small feat, it would be much easier just to get your hands on the pathogen somehow, and you don't need to know anything about bio-science to do that, just have a source. In this matter advancing technology is will not likely result increase risk.
2CarlShulman
I know, it's hard to restore smallpox without a sample, harder still to create a new and unrecognizable disease, harder still to make one without using identifiable dangerous already-studied genes.
[-]VincentYu130

Noun and Chyba (2008, "Biotechnology and biosecurity") reviewed the risks of hostile uses of biotechnology in Global Catastrophic Risks (eds. Bostrom and Ćirković).

An excerpt from section 20.7 ("Catastrophic biological attacks"):

[M]odern society’s experience with bioterrorism has, fortunately, so far been limited to a small number of events that were neither intended to, nor did result in high mortality figures, so they may not serve as good indicators for what a successful major attack would look like. The 2001 US Anthrax scare that caused five deaths, for instance, involved a non-contagious pathogen, and although milled into a fine powder, the bacterial spores were initially contained within envelopes that resulted in only local dissemination. By contrast, the Aum Shinrikyo cult, seeking to stage a mass-casualty attack in order to realize a prophecy, attempted to disperse Bacillus anthracis, from a building rooftop onto the dense urban population of Tokyo. The Aum, which later succeeded in dispersing Sarin nerve gas in Tokyo subways, was, fortunately, unsuccessful both in efforts to procure a pathogenic strain of Bacillus anthracis, and in its attempts to ef

... (read more)
Reply
[-]FiftyTwo120

To look at it in another way, it is surprising that someone with that level of knowledge intent on killing people didn't kill far more people. There are lots of simple ways someone with decent chemical knowledge and access to lab equipment could kill a lot of people in a confined space, but instead he chose to use guns primarily.

My leading hypothesis would be that people in this sort of mental state are not motivated by maximising the number of people they kill but by fitting into the mold of a gunman or fulfilling some other psychological desire. So if that is the case we should be comforted that even if the access of people to dangerous chemicals increases they won't use them. What would be really dangerous is if someone psychologically normal decided to kill a lot of people.

Reply
[-]asparisi80

Since most of the damage seems (thusfar) to have been by gun rather than by gas this particular event, while tragic, does not seem to be evidence for a particular timeline on this sort of risk question.

Reply
1James_Miller
What's relevant is that someone capable of getting into a Ph.D. bio-science program had such a strong preference for mass murder that he was willing to spend the rest of his life in jail to achieve this ends.
[-]Cyan200

preference for mass murdering

My best guess is that this individual has an organic mental disorder, e.g., schizophrenia with paranoid delusions.

Reply
[-]Jayson_Virissimo100

My best guess is that this individual has an organic mental disorder...

My best guess is that this individual has an organic mental preference...

What possible experiences would you anticipate if the first statement was true that you wouldn't also expect if the second statement was true?

Reply
[-]juliawise170

I would expect to find people with very skewed senses of reality (as seen in schizophrenia). I don't consider that the same as a preference. What's currently called antisocial personality disorder, also known as psychopathy or sociopathy, I might consider more a preference (in that it deals with how they value other people's wellbeing, not from their perception of reality).

I wouldn't be surprised to hear that someone who attacked strangers for no apparent reason was experiencing delusions or hallucinations. I would be surprised to hear that someone with sociopathy did so, because they normally hurt people only for personal gain, and there's nothing to gain from opening fire on a crowd.

Hearing voices is not a preference.

Reply
8Desrtopa
People with schizophrenia usually do not attack people for no reason either. The independent association of schizophrenia with violent behavior is low, and most of the difference in rates of violence between schizophrenics and non schizophrenics seems to be attributable to the higher rate of substance abuse among schizophrenics. If you know that someone is a mass murderer, it should give you a high posterior probability assessment for mental illness, but not a high probability assessment for schizophrenia. Most sociopaths are not mass murderers or serial killers, but as best I can determine (I've found articles that allude to it, but none that give an actual percentage, and wikipedia pages for individual mass murderers seem to support it) most mass murderers and serial killers are sociopaths. However, most mass murderers and serial killers are not schizophrenics, although it seems that a significantly greater proportion of serial killers and mass murderers are schizophrenics than the proportion of the population in non mass murderers or serial killers.
1juliawise
Sorry, you're right. I spent last year working on a psych ward, and I agree that most people with schizophrenia are unlikely to hurt others. My guess is that mass murderers with some ideological or practical reason for choosing the people they murder are more likely to be sociopaths. I can't think of a reason to target people at a movie theater, which makes me put a higher prior on delusions or hallucinations in this case.
4Desrtopa
Well, given the specific evidence that the culprit dressed up and identified himself as the Joker and informed the police that he had booby trapped his apartment, I'd assign a high probability of schizophrenia, but I wouldn't write off preference based reasons for a sociopath to gun people down at a movie theater in general. It could be motivated by a fantasy rather than a delusion of the perpetrator. Winston Moseley, for instance, sexually assaulted and killed Kitty Genovese and two other women because he had violent sexual fantasies. Alyssa Bustamante (who I read about recently while looking up information on young female murderers,) committed premeditated murder by her own testimony because she wanted to know what it felt like. Personally, I sometimes find myself frustrated at the ineptitude of both terrorists and the Department of Homeland Security, and have thoughts along the lines of "that's pathetic, I could show them how it's done," (I consider this to be one of the manifestations of my imp of the perverse,) but I'm not inclined to do it because I don't want to terrorize the population. If I were a sociopath, on the other hand, I might actually be tempted to do it. Other people's preferences are not necessarily going to be relatable. Even if there's no potential for profit or cause for vengeance, for the right sort of person, murder could be a thrill activity, like skydiving (another example of something people do for pleasure that other people can't imagine why one would ever want to do it.)
4Jayson_Virissimo
Thanks for the serious response. I see what you mean. My unease about these kind of explanations is the asymmetry between the kinds of terms used to explain why someone would choose to commit a horrible act and the kinds of terms used to explain why someone wouldn't choose to commit a horrible act. After rethinking it though, I agree that making reference to beliefs in addition to preferences does add explanatory power, but I'm not very certain of this. After all, we make reference to the same physical laws to explain why a bridge stays up as when it falls down; but when we explain "normal" behavior, we talk about preferences and constraints, without saying anything about having non-schizophrenia. And yet, fame is a very common preference among humans and this guy is now world famous (he even has his own thread on Less Wrong!). Depending on how strong his fame-preference is, we can't rule out that these tactics weren't instrumentally rational in his particular case. Maybe, IDK. Do you never hear voices? I hear a voice (it sounds very similar to my normal speaking voice, but not exactly). As far as I can tell, this is just what linguistic thinking feels like from the inside. I can even make it go away by using my visual imagination, doing strenuous physical exercise, or practicing mindfulness meditation. The extent that I do make it go away seems to be determined by my personal preferences and the constraints imposed by my environment.
2prase
I don't understand the analogy. Explanations why a bridge stays up usually point out different physical laws than explanations why it had fallen down. You rarely hear "the bridge stays up because the construction hasn't corroded and the engineers made no mistake", unless there is a reason to suspect something wrong with the bridge.
3Cyan
This is a good question for one to ask one's self. That said, juliawise nailed it. If I had seen your reply before juliawise posted, I would have said, "'Disorder' imples the individual's map doesn't match the territory (well ok, consensus reality) in conspicuous ways, e.g., schizophrenia with paranoid delusions." The killer apparently identified himself to police as "the Joker" (with hair dyed red or orange instead of the comic book character's green). He also rigged his apartment with booby traps but then told police about them. I'm not getting a strong sense of coherent goal-seeking here, in spite of the reported fact that the venue of the attacks was apparently chosen for maximum killing.
3Jayson_Virissimo
Yeah, I see it now. I'm pretty sure I would anticipate them having a significantly more faulty web of belief than non-serial killers.
2Decius
If you think that the venue was chosen for maximum killing, you haven't considered what someone who was optimizing for killing would do.
-2Jayson_Virissimo
Another possibility is that Cyan and James Holmes are not as creative or intelligent as Decius.
-4[anonymous]
http://www.amazon.com/Schizophrenia-A-Very-Short-Introduction/dp/0192802216/
[-]Dr_Manhattan120

I'm not sure if that's surprising. I've seen somewhere that psycopaths have higher than average IQ, and there were some serial killers who certainly could have managed a PhD program... Wait, Kaczynski?

Actually I'd be more scared of a Kaczynski than of Holmes. Holmes seems to enjoy his mass murder up close and personal. Kaczynski was driven by ideology; that type seems much more dangerous.

Reply
4mwengler
Kaczynski killed 3 people in 20 years. Holmes killed 12 people in a few minutes. I can't easily get away from an estimate that I am around 4X as likely to suffer from someone like Holmes than from someone like Kaczynski, and so if I had my fear under any sort of rational control and thought it was helpful in lowering my chances of harm, I would fear Holmes more, myself.
6fubarobfusco
I suppose it depends on whether you expect to be the target of someone's ideology. Kaczynski targeted mostly people in technical fields (science professors, grad students, and people in the computer business). Anders Breivik, who copied parts of Kaczynski's manifesto in his own, targeted mostly members of a particular political party.
3asparisi
Why is the category of 'capable of getting into a Ph.D. bio-science program' relevant? I can kindof see where you are going with this; it's easy to say that someone who can design something that could hurt a lot of people has better access and is therefore more likely to do so. But while this is evidence that someone of the category you've named can want to engage in mass murder, it isn't evidence that he had better access or was more likely to do so given that he may have had better access. In actual fact, since the harm here largely involved a firearm and we don't know if the canister was even related to his bio-science program you should be updating downward on someone in that category relying on a biochemical agent rather than just using a gun. You have evidence that, even with a (perhaps strong) bio-science background people who engage in mass murder for what appear to be their own preferences just use a damn gun like everyone else with the same preference.
0Decius
People who kill with guns aren't trying to kill lots of people. Mass murderers prefer armies, or if armies aren't available, bombs. A backpack-sized IED have killed more people than a gun.
2asparisi
Generally I agree, for certain values of "a lot of people" and with the restrictions of a given person at a given time. Fortunately, not all mass murderers are optimal mass murderers.
0Decius
Very fortunately, there have been very, very few optimized murderers. The most effective ones use armies and countries or religions, while the ones who use bombs are many orders of magnitude lower in effectiveness. Part of the lesson is that guns in the hands of citizens helps prevent mass murderers, even though it facilitates group murders.
[-]gwern50

The technology already exists for hundreds or thousands of deaths, you will grant; but they are not obviously being used, and instances where unusual methods are being used are low body counts (the 2001 anthrax attacks). Given that the spree killers are not already using them, why would we expect this this change?

Are you arguing either that even a small probability of a spree killer using them is too much when the damage could run into the hundreds of thousands, or that the increasing capabilities themselves will increase the probability?

Reply
4James_Miller
I'm wondering how much damage a guy like this could do in the future if he decided to kill as many people as possible. I figured that some readers would have a strong enough life science background to be able to make a reasonable estimate.
[-]aelephant160

This is assuming that his goal is just to kill as many people as possible. He could much more easily have set up some kind of bomb in the movie theater and killed as many, if not more, people. My impression is that he wanted the visceral rush of murdering all of those people first-person.

Reply
[-][anonymous]20

How long will it be (if ever) before a typical bio-science Ph.D will have the capacity to kill, say,a million people?

I certainly think there is a considerable risk, as bio-engineering becomes routine even in smaller labs. One could rely on brute force methods, like creating an array of random compositions of influence viruses for example, attempt to infect a few people, hope something sticks(selection). And then several different strains with non-overlapping surface antigens - you get no overlapping immunity - emerge from your batch, instead of one super pathogen.

Reply
[-]Shmi20

How long will it be (if ever) before a typical bio-science Ph.D will have the capacity to kill, say,a million people?

How hard is it to get one's hands on some of the select agents?

Reply
2Decius
The second one on the list can be created accidentally in the kitchen, using ingredients already present in most kitchens. I'm not sure if there is a safeguard against slipping it into some portion of the food pipeline, such that it struck as many as a million prior to a recall being issued.
[+][anonymous]-160
Moderation Log
117Comments