« Incomplete Analysis | Main | Anthropomorphic Optimism »

August 04, 2008


Should governments perhaps force all microbiologists to be evaluated to determine if they are sociopaths?

How exactly do you imagine this can be accomplished? Check their blood for metabolic byproducts of evil? Have trained phrenologists examine their heads for unusual bumps in the 'supervillain' area? Inspection of tea leaves? Ask them about their mothers? Show them inkblots and see if any are perceived as mankind dying in a megaplague?

Psychology doesn't work that way. Psychiatry certainly doesn't.

The whole story still seems *highly* suspect to me. Extremely convenient that he committed suicide when he did. Not to mention the fact that the feds originally tried to character-assassinate Hatfill before that fell through and they turned to Ivins. Just sayin'

The whole story still seems *highly* suspect to me. Extremely convenient that he committed suicide when he did. Not to mention the fact that the feds originally tried to character-assassinate Hatfill before that fell through and they turned to Ivins. Just sayin'

Caledonian - The article I linked quotes Ivins therapist as saying that Ivins "has been forensically diagnosed by several top psychiatrists as a sociopathic, homicidal killer. I have that in evidence. And through my working with him, I also believe that to be very true." So apparently psychiatrists do think they have some skill at identifying sociopaths.

psyche screening seems prudent. there's a psychopath checklist that i believe is used by psychologists.

what's the long run solution though? it seems you'd need a power of considerable strength and international reach to prevent psychopathic types from getting dangerous educations in things like microbiology or physics.

one might consider people with powerful brains in certain fields to be a bit like WMDs themselves, and you might look at their proliferation as you would proliferation of nukes, et c.

I suspect the therapist may have been influenced by contact with the FBI. "Forensic diagnosis" by "top psychiatrists" sounds like FBI profilers to me. It would be pretty easy for an FBI interviewer to drop hints or info about their investigation to the therapist and the therapist in turn to become alarmed and go for a restraining order (which is where these statements appear). Then the restraining order becomes evidence against the suspect. Then again, it might all be true. But the FBI has a track record, with Richard Jewell, Hatfill, etc. of creating suspicion around innocent (and even heroic) people.

Also, on the subject of screening, we need to start with politicians, lawyers and law enforcement. Just sayin'.

So apparently psychiatrists do think they have some skill at identifying sociopaths.
Sure - by looking at what a person is known to have done, after the fact. Identifying people ahead of time? We can't do it. And not for lack of trying.

Even when the categorization can be made, it's because the sociopath was caught engaging in seriously maladaptive actions - usually because they showed incredibly poor judgment both in committing the actions and covering them up again. Some psychologists claim poor judgment is an inherent property of sociopathy; others point out that the traits referenced by the concept frequently appear in highly-successful people in all sorts of careers, and such people are likely to evade detection by any means. Sociopathic traits make it easier to fool people with a facade. Combined with high executive functioning...

This is the same issue as with people who demand that we prevent criminals likely to offend again from being released. Recidivism can be calculated for large groups by looking at past statistics, but we have no generally reliable method for individuals - despite many years of looking and vast resources expended on the task.

Other sources that I've seen suggest that Ivins had gotten a lot worse over the past year, although he had always talked a lot about wanting to kill people.

DSM-IV 295.3

Well, he couldn't have actually been a killer, because therapists are not obliged to protect people who have committed violent crimes, and the guy would have already been in prison.

Let me project mild skepticism that a lone crazy happened to work at a secure bioweapon research facility, had the clearance to check out enough weaponized anthrax to conduct multiple (admittedly pathetically ineffective) attacks without anyone noticing (or being responsible to notice, apparently).

This is hardly the first time in history where an intelligent scientist could use their knowledge to harm people. I don't really feel it's enough of a threat to be worried about.

Although, on the plus side: you know what comes hand-in-hand with evil mad scientists bend on mass destruction? Superheroes. Truly we are at the dawn of a new and glorious age.

This is somewhat irksome to me, even if it is untrue... I know a group that is currently trying to take advangtage of the adenovirus's ability to link to dynein, which is a motor protein that takes cargo directly to the nucleus, and would be a free ride for a virus if not for evolved cell counter-dynein mechanisms... If this group makes a sucessful 'adaptor protein' to prevent the cell from detatching the virus and splices the gene for said protein into the adenovirus genome... well, transfections will work much better-that's for sure- BUT.... fill in the blank.

Now, I had the thought, 'Gee, these people are just too much big idiots to do this successfully.' But what if I'm only half right?

So... After hearing this talk, I was much less comfortable just letting anyone work with viruses. On the other hand, I'm not sure regulations would help anything, or what kind of people/policy makers we could realisticly get to enforce them.

Reapproaching Caledonian's objection: how big are the Type I and Type II errors in that screening process, and at what cost? False positives are highly problematic, but false negatives are just inefficiencies. If there is a simple way of preventing the low-hanging apocalypses, definitely, do that. Past that, we are talking about some trade-offs with a potentially poor test. Versus an existential threat.

See also the problems of "will this Large Hardon Collider destroy the visible universe?" and "how do you know this AI will be friendly?"

> If the blind idiot God of evolution can develop microorganisms that kill millions, what could an observant, intelligent microbiologist accomplish?

What could a top-notch (perhaps world-class), inventive, very well-supplied biologist do?

Kill fewer people than your random mouthbreather with a gun, it would seem.

Yeah, the data suggests that an observant, intelligent microbiologist could only manage to kill 5 people before being stopped.

Gwern and Tim - You can't make strong inferences from a sample size of one.

The last thing this world needs is a stronger government hand in the academic community. How long will it be before our ever knowledgeable overseers decide an up and coming microbiologist's appreciation for Dawkins is unhealthy.

You might want to read "The White Plague" by Frank Herbert. In that novel, an Irish-American biologist faces a tragedy when his wife and two daughters are killed by terrorists on a visit to Ireland. His revenge is terrible. He uses his knowledge and wealth to develop a virus that kills only, but all, females; he unleashes it on the three countries he blames for the deaths. Of course, the virus escapes, and the scientific community spends the rest of the story looking for a vaccine before the human race is finished.

In this novel, the biologist is insane as opposed to sociopathic. But the dangers are the same: knowledge of ways to make cheap WMD in the hands of a small number of angry people.

We got off lucky on with the anthrax guy; we might not be so lucky next time.

Tim: No, I can't make a strong inference from one example.

Nor could I from Aum Shinrikyo spending millions to kill 12 people, or Saddam killing thousands with more trouble than just shooting or starving them would've been, nor...

I'd rather make inferences from real evidence than continue to deal in vague 'we're all gonna die!' hypotheticals.

Gwern - Aum Shinrikyo stopped developing bio weapons in 1995. Microbiologists have learned a lot since then. Saddam, to my knowledge, never used bio weapons.

Argh. For a blog devoted to overcoming bias, you folks sure have a tendency to report badly-sourced crap as if it was pearls of wisdom from the mouth of God.

The credibility of Jean Duley, the therapist who is the source of these allegations about Ivins, is very much in doubt. See here for some attempts to actually get at the facts of the matter.

But that's not to say there's not something to be nervous about, given that we are currently at a point where biohacking techniques and components are accessible to a bright high school student. Don't know how we're going to screen all of them.

"Should governments perhaps force all microbiologists to be evaluated to determine if they are sociopaths?"

I would expect governments to do a lot of damage with that power without much positive results at all, especially because sociopathic microbiologists are probably so rare that the number of false positives would drown them out completely.

Next stop, thoughtcrimes.

Criminalizing sociopathy is creating a thoughtcrime. It depends on sources, but some claims of prevalence place as many as 15% of the population as sociopaths. Also, sociopathy seems to be merely the lack of what could best be called "moralistic intuitions", they can still see and reason that killing is not in their best interest for various reasons.

Seconding Gwern and MTraven here.
Cells have lots of defenses against adenoviruses other than counter-dynein mechanisms, organisms have defenses other than those of cells, society defenses other than those of organisms, and species are ultimately defended by the logic of epidemiology (optimal for lethality != optimal for contagion or optimally fit) in any event.

Smart people are dangerous. No telling what they'll come up with. Keep smart people away from resources that could be dangerous.

Dumb people are dangerous too. They might do extremely bad things by accident. Keep them away from potentially-dangerous things too.

Average people combine the dangers of the other two groups.

If we want to be safe from this stuff, we need to keep people from doing science. Period. Science and technology give people new methods, new power. Power is dangerous. Keep it away from human beings.

B.H. - "The Moral Virologist" by Greg Egan covered a similar idea. The virologist's motivations were religious, and he engineered a virus that killed adulterers and gay people, with catastrophic unintended consequences.

You'd be surprised what a top-notch microbiologist could do--it's not all that hard, using genetic recombinant technology, to produce a "super bug." The hard part would be scaling it up for military purposes (just like how making weapons-grade anthrax is the difficult stage).

Researchers have already resurrected the 1918 influenza strain, and everything else may be simple in comparison. One could imagine a clever graduate student using current technology to mix and match (i.e. artificially re-assort) genes in an influenza genome to produce a strain that is highly effective at infecting humans AND highly virulent.

The silver lining is that this hasn't been done before, which tells you something about the people doing this sort of research (i.e. they aren't, on the whole, sociopaths).

Oh, don't be another of those anecdote-driven dumbasses, James. Defending against specific threats like this is a fools game. Next thing you'll be raving about how dangerous long-distance buses are in Canada.

We won't know how dangerous this sort of thing actually is, until somebody does it.

Virulence is complex. It tends to involve getting somebody sick enough to be infectious, and still well enough to go out and infect people. That window needs to be open for a significant time. Also it matters how many infectious particles are needed for a clinical infection. For anthrax by inhalation that's somewhere between 1000 and 10,000, if you get a small enough dose you probably won't develop symptoms. For Ebola it might be somewhere between 1 and 10.

When you create a new virus by patching together pieces of existing ones, it needs a lot of fine-tuning. Think what you'd get if you tried combining pieces of a Harley-Davidson motorcycle with a Jaguar sports car. You might get something fast that ran on 2 wheels with really good doors and windows -- but it would take a lot of design work. Without knowing how to design for virulence and infectious ability, evil molecular biologists currently have to take pot luck. It takes extensive testing on human subjects to be sure what you have, and currently only governments can provide the prison populations to do that testing.

So if you want to be a mad-scientist this way, you'd need to make one thing after another and release them, and wait until one of them did something. You wouldn't get to choose how it would behave, you could only find out what you had after it was already spreading.

Kind of like the old meme:

1. Release random viruses into the population.
2. ....
3. Profit.

How sure are you that you don't have a sociopath running your sociopath-detection program?

The more power the program has to affect people's lives (and if such a thing is developed, it's not going to stop with bioweapons researchers), the more attractive the job is for sociopaths, both for self-protection and as an opportunity for blackmail.

I'm not even going to mention the possibilities of putting together a super-villain organization. That's just for fun, and good enough for fiction.

True sociopaths (people with antisocial personality disorder) are very rare, ~2% of the population. They usually aren't very dangerous, because they can't plan ahead, they behave impulsively. I am extremely doubtful that someone who can't keep a job for more than a few weeks could ever manage a super plague.

>True sociopaths (people with antisocial personality disorder) are very rare, ~2% of the population.

So in a high school with 1000 people, there'd be 20 sociopaths. Somehow that doesn't sound very rare.

As additional evidence comes in, it looks more like Ivins was in fact the culprit and that he had some serious mental problems, although he doesn't look like a "sociopath", just a depressive with paranoia, which is bad enough. It is pretty scary that someone like this was working in a secure government lab, but it is more scary to realizae that advanced biotechnology is going to be available outside of military labs, and soon.

The archetypical sociopath ever has to be 'Johnny Fair Play' from the 'Survivor' TV reality series. This was the name he gave himself, but his play was anything but fair.

Any number of ultra-technologies could attract sociopaths like flies to honey. Watch out for the 'Johnny Fair Plays' of AGI, preaching about morality whilst scheming to take over the world - in fact, anyone giving us long speeches about morality in regards to artifical general intelligence should immediately be suspect.

There are any number of crack-pots and meddling dabblers associated with artificial general intelligence. We can only hope that the moderators catch all of these bastards.

I think Ivins was a patsy.


Norman, when the attacks first came along with crude anti-israel slogans it made some sense to suppose they were done by somebody who was anti-israeli.

Then when the material came from an american bioweapons facility, and the arabs who worked there were cleared, it made sense to suppose they were done by somebody who wanted to support israel by stirring up resentment against arabs etc. The purpose of the attacks was not to spread disease -- only a few unimportant people died, mostly by accident, and the death toll could easily have been far far worse. The obvious purpose was to make it look like people who hated israel attacked us, so we would support israel.

But instead we get a dead man who was mentally unbalanced, who could have done it for any stupid reason at all.

In the first case we would have had the question why we allowed arab terrorists access to our bioweapons labs. In the second case we would have had the question why we allowed fanatical zionist murderers access to our bioweapons labs. Instead the question is why we allowed a garden-variety madman access to our bioweapons labs.

Once it turned out to be an official US bioweapon that got aimed at the US Congress, our security guys looked real bad no matter who it was they slipped up on.

Here's a side question: Does the USA need these bioweapons? Note that when we invaded iraq and we were concerned that iraqis might use anthrax on us, we didn't threaten them with our own anthrax. We threatened to nuke them.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30