I recall reading somewhere that useful ideas in computer science were frequently thought up by graduate students under the influence of certain hallucinogens.
My question is whether this is factually, historically true. If so, what are some examples?
I’m not asking about whether being drugged out of your skull is in fact helpful for developing innovative ideas. That’s an interesting question, too, but since we’ll never know for sure, that’s IMHO not a GQ. I don’t mean to start a debate about whether we advocate or use drugs. Rather, I’m hoping this thread can promote a better understanding of the conditions under which humanity taught itself to make computers.
I’ve been a CS grad student and prof. The “stoner” grad students were the least productive ones without question.
Doing anything really good involving computing requires a fairly lucid mind. A firm grasp of logic is quite important. Drugs don’t assist in logical thinking.
CS is sort of a weird field that many people don’t understand. It is highly creative, but also highly restricted. Standard MMP personality tests are designed under the assumption that the two are inherently incompatible. They aren’t. Logic and beauty can go together.
From my “OTOH” file (note that I don’t even drink so this is as close as I can come):
When I was a graduate student, I had an “almost working” algorithm. I had some ideas on small changes to make it work. But I got sick and my mind went fuzzy. So I coded up a program to check correctness of these variations and ran my changes thru until I found one that work. That lead to my first conference publication.
A few years later, someone wanted a simpler version of that for a less restricted model. What happened then was my most famous result (it appears in textbooks).
But I think that if I hadn’t gotten sick I would have still come up with it, and in far less time as well.
There’s a widespread belief that drugs aid you in creativity, and while that may or may not be true, they’re not going to help you in math or CS. The thing is, as creative as your ideas are, they also have to fit in the logical framework that exists. They have to be internally consistent and relate well to what’s already out there. No matter what effect drugs have on your creativity, they really mess up your logical thinking skills, so I’m doubting this.
Another tale from the Old Days in CS, before people caught on to the general personality type of computer geeks.
For a conference, the organizers would negotiate with the hotel for rates on the rooms, banquets and so on. We used to get really low rates because the hotels thought they would make up for it on high alchohol markups. Boy, did they lose money! One hotel manager got incensed and ended up yelling at the local organizers. Hardly anyone was drinking!
Keep in mind, that these were top-of-the-line research conferences with some of the biggest names of all time in attendance.
Top CS researchers drink very little, so I hardly think drugs made much impact on the field.
I do know that drug usage by various computer weenies is keeping me employed. Seven of the last ten contracts that I’ve had, I was brought in to replace someone who had gotten fired because of drugs. Each one of them ran longer than expected because there was so much repair work to be done, cleaning up crappy, drug-fogged code and the like.
Well, this isn’t an all-or-nothing issue. Sometimes working a problem to an impasse, “expanding your mind”, and coming back to check that it makes sense once sober can get somewhere, but not often enough to rely on it. My advisor admitted to having done this on a take-home final exam as an undergrad when asked to construct a nonstandard model for arithmetic.
Another interesting anecdote is a friend of mine who had been reading about Goedel’s theorems and could see the logic but couldn’t quite grok them. On an unrelated note he tried some drug (I forget which) that is known to inspire a lot of self-referential thought. By the time he came down, Goedel just clicked.
DISCLAIMER: I am by no means advocating the use of any substances illegal under American law or prohibited as “performance-enhancing drugs” by the governing board of Mathematics.
Nope. But it did jog my memory as to where I read about this.
Back in (I think) 1996, Time magazine did a cover article on whether consciousness can ever be replicated by machines. The article said (approximately) that early AI researchers were reputed to make many important discoveries in late night pizza parlors, whilst their minds boiled on an assortment of illegal substances.
I don’t believe the article confirmed or denied that early AI researchers did in fact do this.
At the time I read the article, I believed that all AI researchers are computer scientists. Even though I now know that this isn’t always true, I had conflated AI with CS. And thus the story came back to me last night, when some guy in a bar told me that it was “no coincidence that LSD was discovered at the very beginning of the information age.”
In any case, I’m fairly well convinced that this is wild nonsense, thanks in part to these posts describing the mindset and discipline required to be a computer scientist.
Hmm. Brian Kernighan would be interested in hearing that. Not that I think Bill Joy was/is much of a stoner.
Bars aren’t the only ones disappointed by computer people. I went to the National Computer Conference in Houston in 1982. Very big show, filling up the Astrodome convention center. At the end of the day when you left, you had to pass a line of hookers, who looked extremely disappointed in the lack of attention. Maybe they did better when the sales guys left.
Some video game ideas may have been conceived while under the influence, but I guarantee you if you try to actually code a game (or anything) while drunk or high, it’ll never even compile, much less make it to market. I haven’t been drunk or high, but I have been dead tired from all night coding sessions, which isn’t such a hot idea either.
Statements like “drugs do this” or “drugs do that” are pretty vague, and can be applied to my personal mantra, “all generalizations are universally untrue”.
As for marijuana, the level of creativity or lethargy induced by the drug varies widely from one batch to the next (I can vouch for this based on personal experience). Stoners speak of this as a range of effects from “clean” to “stony”–“clean” describing weed that gets you “high” (creative, inspired, etc.), and “stony” describing weed that gets you “stoned” (lethargic, lazy, tired, etc.). Although stoners generally don’t speak of such things around Muggles, if you listen closely you may hear us saying things to each other like “this stuff is a really stony mec” or “the stuff I scored on Wednesday gives a nice clean high”.
Anyway, the whole point of that is that marijuana isn’t universally sedating or thought-clouding, and that a computer scientist who wishes to write inspired code or to expand his mind to think around a problem in his code might partake in some especially clean chronic and end up seeing solutions he wouldn’tve thought of otherwise. I myself have doped my way around code impasses and, on other occasions, doped msyelf into pumping out shitty, worthless code.
I would imagine that LSD and coding have generally been considered seperate pastimes by computer scientists. I wouldn’t know–I’m not a computer scientist (although I’ve taken a couple of programming classes) and I’ve never dropped acid (nor do I ever plan to).
As for other drugs, I used to know a CSist cokehead, who probably mixed his two favorite pastimes quite often. I can see how stimulants like cocaine could help you code better. I’m sure everyone’s familiar with the All-American tradition of late-night amphetamine-fueled cram sessions, and it’s not much of a stretch to figure some CSists take their code for a “spin” every once in a while.
“Dead tired” and “high” (on weed, anyway) are two very different states of mind, let me tell you. I’ve had code that wouldn’t compile at all until my good friend Mary Jane got in on the action.
By the way, I have ADHD, so I’ve lived most of my waking hours (most of my in-school hours, anyway) under the influence of one drug or another, usually something not far from amphetamine. Most of my coding has been under the influence of drugs–some recreational, some prescription. The only difference between the two is a bottle and a label with my name on it.
I don’t think he ever was, either - he seems to have been too wrapped up in the joys of coding to get involved in anything else. Having said that, though - remember how he looked in the early days? One look at his picture and you were convinced that he was a wierdo-hippy-stoner-freak.