Jun. 21st, 2010 @ 01:07 pm
Just a little thinking out loud about thinking.
A common refrain in reporting about the internet, and other related social trends, is that it makes us stupider. Google makes us stupider, multi-tasking makes us stupider, using tiny text formats to communicate makes us stupider (or at least erodes our ability to write, etc).
At the same time we keep having computer demonstrations of things that we used to consider really clever. We used to think chess grandmasters were the heights of human cleverness, but computers are very good at chess now, and being a chess expert has lost a little sparkle. IBM has a computer that plays Jeopardy
, which is pretty much a classic competition of human cleverness. Right now, this computer is huge and expensive — but we all expect that this decades huge and expensive is next decades affordable and the decade after thats cheap and ubiquitous (if it even takes that long). Once it is ubiquitous, well, it still might be kind of clever, but it is not a kind of clever we value. Multiplying two large in seconds was once considered astonishing, and highly valuable, now it is a curiousity as a human ability (but the foundation of much modern infrastructure as automation) — human being mimics $2 calculator.
In the medieval era, cleverness was all about the sheer amount of information you could keep in your head, and organising and accessing it. They developed intricate clever techniques for it, such as Memory Palaces
. And in that era, a very smart man could theoretically master all academic knowledge. The last man reputed to have mastered all the worlds academic knowledge of his era
died in 1680, and was already thought a little outdated then, and the idea is now absurd — but we all have wikipedia at our fingertips, and can find out an awful lot about almost anything fairly quickly, and the idea of cleverness being defined mostly by prodigious feats of memory now seems quaint and somewhat pointless. Great feats of memory are something we associated with autistic savants, not the worlds great thinkers.
It is not that the internet is making us smarter or dumber. But the internet is changing what it means to be clever, as each new set of cognitive skills is 'outsourcable' to machines, the cognitive skills that we value changes, our ideas about what is really
clever are updated.
|Date:||June 21st, 2010 06:47 am (UTC)|| |
|(Link)|Multiplying two large in seconds was once considered astonishing, and highly valuable, now it is a curiousity as a human ability (but the foundation of much modern infrastructure as automation) — human being mimics $2 calculator.
And yet, people should retain the skill of being able to do basic arithmetic without needing a device to do it. It's not about being remarkable it's about competency.
|Date:||June 21st, 2010 07:19 am (UTC)|| |
Intuition about how numbers interact is still totally critical to being able to estimate costs, improvise construction, and tune complex systems.
It's not enough merely to know that two adjustable parameters have a multiplicative effect on price, size or output: it's also necessary to be able to think out the approximate result of specific choices for each.
Relatedly, I don't believe there are many who have poor grammar but can still put together a solid, logically ordered argument.
|Date:||June 21st, 2010 08:31 am (UTC)|| |
|(Link)|If I could think 5 * 3 at some sort of assistant computer and get back 15, wouldn't that be wonderful and liberating?
Or it enslaves you do our new silicon overlords ;-) The premise of Dave's original argument seems to be that technology frees us from memorizing stuff to be able to think about higher level problems. I don't agree with that analysis. I think we're being freed from life skills so we can worry about whether Britney is wearing underpants.
My premise was not particularly that technology frees us from memorising stuff to concentrate of cleverer things -- more, that technology has made memorising stuff a vastly less necessary skill than it used to be, and so other skills have become relatively more valuable, and this is just as natural as no longer valuing the ability to lift heavy things or good horsemanship, as much as we used to. In general I think anything that aids human productivity is good, but I don't think being freed from some onerous mental task implies that we have to spend the same effort on developing other worthy skills.
And I think we always worried about whether the Britney equivalent of the time was wearing underpants (just read Suetonius on Valeria Messalina, or Procopius on Theodora). It is actually a good thing that a large percentage of the population is able to spend a large amount of their time worrying about that sort of thing rather than where their next meal is coming from, or how to avoid being stabbed by gangsters, or whether or not they will make it through the winter. We may not be doing the most awesome job of giving people interesting things to fill their leisure time with, but it is fantastic (and historically quite new) that a large percentage of the population actually has excess leisure time.
We may not be doing the most awesome job of giving people interesting things to fill their leisure time with, but it is fantastic (and historically quite new) that a large percentage of the population actually has excess leisure time.
Also, we have moved on to Miley Cyrus's underpants.
|Date:||June 21st, 2010 08:55 am (UTC)|| |
"If I could think 5 * 3 at some sort of assistant computer and get back 15, wouldn't that be wonderful and liberating?"
We have said assistant computers. They're called brains.
There are certainly circumstances where your point is valid, but many more where it helps not to need a spreadsheet.
Let's say you're renovating. If you have to boot a computer or track down a calculator to work out to the nearest 4L tin how many square metres of paint coverage you're going to need, or how many 30x30 tiles you'll need to get to the end of the skirting, that's a pain in the arse.
I'd go further than that, however, and suggest that many management decisions are made on an ad hoc basis choosing between estimates of the total cost of alternative paths forward. In these cases due diligence may be nice for reporting, but is also slow and painful.
Nevertheless, I certainly agree that the understanding required to conceive of an analytical approach is far more useful in a human than number crunching.
But being able to perform complicated arithmetic yourself is essentially irrelevant to having a sense of intuition about numbers. For the purposes of intuition, being able to perform calculations to one or two digits of precision is generally fine. I think basic arithmetic skill is useful and important - but being able to perform exact calculations for a large number of digits is essentially a redundant skill. Not only do I never do long division, I think it is positively a bad idea for me to do it - I trust calculators more, and even if I wanted to right a long division function I wouldn't trust my memory on the perfect algorithm, I'd look it up in Knuth.
FWIW, my best friend is dyslexic and his spelling is dreadful. He has two B.A.s and a PhD - because for a while we though spelling was an important part of literacy, but we have now got the technology to outsource the difficult parts. We can't do the same for grammar quite as efficiently, but we can significantly improve mediocre grammar using software.
|Date:||June 21st, 2010 09:06 am (UTC)|| |
Point by point:
I don't believe there are many
, not any
... I also know a couple of brilliant people with language difficulties. There are also legless people who can run exceptionally fast, and brilliant surgeons with Tourette's syndrome.
To me, two digits of precision implies reasonable ad hoc numerical ability. I certainly can't multiply random three digit numbers accurately - in order to apply my "numerical intuition" I'd bracket my estimate using the upper and lower two-digit approximations of the multiplicands.
I believe spelling is
an important part of literacy. Misspelled documents result in lost respect and sometimes misinterpretation; automated spell-checking is still error prone; human-assisted spell-checking is grunt work that isn't fit for people who can actually think.
Software-assisted grammar checking in current common-or-garden tools is still utter shit.
I disagree fairly strenuously with the idea that trained and habitual intuition and precision working both with numbers and language is not a highly valuable trait. Although we are more and more able to rely on machines to handle these basic challenges, their centrality to our day to day tasks is also on the rise.
|Date:||June 21st, 2010 09:08 am (UTC)|| |
Of course, I can count a few grammatical errors in that comment.
While being clever and having good facility with language are correlated, that doesn't tell us that much. In particular, we train people in written language skills and value them highly, so of course clever people will tend to have them, but are they really necessary for a facility with abstract thought, or would spoken language skills do just as well in a culture that valued formal written language less? I'm inclined to think it is cultural bias from those of us who already have such skills.
I'm certainly unconvinced that spelling is that important a part of literacy in the age of the spell-checker. I mean, it helps, but fairly marginally. And some grammar rules I'm convinced are little more than class markers so we can tell the nice people from the oiks.
But it is all largely irrelevent I disagree fairly strenuously with the idea that trained and habitual intuition and precision working both with numbers and language is not a highly valuable trait.
But I was certainly not expressing that idea. I think precision working with numbers is a highly useful skill, I just think we now do it using machines to assist us rather than performing tedious large arithmetic calculations ourselves, and so the ability to perform said large arithmetic calculations has become largely a historical curiousity. If anything, this should develop our skills at habitual working with numbers, as we need spend less time as children rote learning tedious arithmetic methods.
Similarly I think precision working with language is a valuable skill - but we have tools to assist us with that too, and many of our hard learned language skills are becoming redundant. A lot of my early schooling was based on the idea that good handwriting was an essential skill, and I was rewarded for my ability to spell complicated and unusual words -- neither seems particularly valuable today (complicated and unusual words are precisely the ones that spelling checkers excel at, and I haven't handwritten anything of any length in years). Excellent handwriting was once an essential skill for some professions, now it is an oddity. Precise use of language continues to be a valuable ability -- but what exact individual skills are most valuable in pursuit of that goal can, and will, change as we develop machines to assist us. We will continue to value those skills that genuinely can't be replaced by machines, and naturally cease to care as much about those skills that can. With each iteration, as the need to learn and exercise the duller parts of the process drop away, we are able to (if we choose) use the cognitive resources freed up for other skills (spend the time we would have spent checking spelling on graphic design, etc) or not (and have more leisure). Either seems valuable.
|Date:||June 22nd, 2010 01:51 am (UTC)|| |
Dave, I believe I understand your arguments, but I think you're wrong. I'll address your point about the correlation of cleverness with written language skills on two fronts.
Firstly, that correlation's partial basis in society's high valuation of writing skills makes the skills themselves no less significant: unless you propose that having skills that society values highly isn't important.
Secondly, I'd argue that clever people tend
to be highly adept manipulators of language not just because of social pressure, but because that's what defines them as clever. There are certainly people who are better at using language to communicate, and they are useful for that reason. Training them further in precise communication helps them contribute to everyone's benefit.
I personally believe written language skills are crucially important to the process of abstract thought. But it's clear to me that these skills are even more crucial to communicating the abstract things that are thought, abstractly, by abstract thinkers.
This can be deduced from the increasing sophistication of the discourse and notation that accompanies the communication of increasingly abstract concepts - as in advanced mathematics, or literary theory.
I agree that spoken language, pictures and diagrams - all of which we now have this fantastic capacity to record, transmit, and render - can ably assist written communication, but I sincerely doubt there are many, if any sufficiently specialised disciplines, in which there is a genuine overall trend towards reducing the use of written communication. There is currently no better, more efficient way to make many thoughts persistent and allow them to be disseminated than to simply type them out.
I agree that we have useful tools to mitigate against the absence of habits of accuracy in writing and arithmetic, and I agree that some skills have fallen, or are falling by the wayside. But there's also a burgeoning number of tasks in which precise communication is even more important than it has been in the past. Contract drafting, tender responses, liability statements, budgets, programming itself (of course), the aforementioned academic disciplines are all examples, in the case of all of which misplaced, poorly chosen, badly ordered or misspelled words can have significant unpleasant ramifications.
You personally know so many people whose livelihoods depend on precise communication, and whose productivity and pleasure at work would be greatly diminished by having less facility as writers that I really find it hard to believe you think its importance is on the wane.
|Date:||June 22nd, 2010 02:06 am (UTC)|| |
Have to apologise for talking past you to some extent - I don't mean to be infuriating. I do appreciate many aspects of your argument. I think this statement:
"We will continue to value those skills that genuinely can't be replaced by machines, and naturally cease to care as much about those skills that can. With each iteration, as the need to learn and exercise the duller parts of the process drop away, we are able to (if we choose) use the cognitive resources freed up for other skills (spend the time we would have spent checking spelling on graphic design, etc) or not (and have more leisure). Either seems valuable."
is quite accurate. However, I think the evolution of programming languages and tools alongside the evolution of the uses to which programming is put demonstrates in a very concrete way that as tools to amplify the effectiveness of written language are improved, there are correspondingly greater expectations placed on the capability of systems based on written language, without any real diminishment of the need for habits of precise expression. The more we automate, whether explicitly in software or hardware, or implicitly in the systems of convention in which all humans are embedded, the more complex and ambitious the overarching process of specifying that automation becomes.
|Date:||June 21st, 2010 09:10 am (UTC)|| |
Of course, people who actually know stuff have the best filters. To go to extremes, imagine someone with no vocabulary trying to come up with effective search phrases.
|Date:||June 21st, 2010 09:27 am (UTC)|| |
*nod* I'd agree with that.
A classic example, for me, was being faced with a sensitive buffer level control problem, finding that a bang-bang wasn't adequate, and remembering, ever so vaguely, that PID feedback controllers existed.
That recollection was the first step of "applying half-remembered knowledge to solve a problem", and got me as far as adapting a cut and paste of a basic implementation to my circumstances.
The second step was vaguely recalling something about the significance of the integral "windup" term to minimising the settling time. But in order to fumble that far I'd had to draw on a fair bit of actual, albeit latent, knowledge, all of which I'd obtained in a formal educational context.
Bravo, la frase ГЁ venuto solo a proposito
Hmm. People say I'm intelligent a lot, but this is only because I have some sort of weird mnemonic happening where I remember stuff. Just random stuff. I think it's still considered a feat of cleverness.
Being able to use the outsourced information and digest it and then come forth with your own theories and use information to develop new stuff--that's a mark of intelligence, or at the very least a sign of creative and/or logistical thinking. Now logic: that wrangles my brain, wrings me out, and leaves me weeping--and with only one question of logic!