My greatest predictive failure

by Doctor Science

– occurred late in the spring semester of my senior year at college. I looked up from the paper I was writing, out the window to where the sky was lightening over Princeton with the dawn.

I said to myself: "When you're out there in the Real World™, you won't be able to procrastinate and then pull an all-nighter to get it done."

Never have I been so comprehensively wrong. I got into that Real World™, and here I still am, pulling the occasional all-nighter to get something done. My husband does it too, but then we're both self-employed so strict 9-to-5 isn't really what we've signed up for.

So I pulled one last night, but I got everything up and running and to the client today. I'm a website designer/builder, and this was a project to basically build a small content management system. I decided to upgrade my skills and learn more about PHP/mySQL/jQuery/Ajax -- I've tended to do more Microsoft-centric stuff in the past -- but it was a really exasperating learning curve that I don't feel right billing them for.

Anyway, for those of you who don't know what I'm talking about, it was yet another illustration of one of the more accurate predictions I've encountered in the last couple decades: back in the late 90s, our ISP had some tutorials up, saying something like "in the Internet Age, we'll be going back to school for the rest of our lives."

So tell me: your favorite accurate and/or failed prediction? Do you still pull all-nighters? Do you see dawn more often from the back side or from the front?

Dawn in St. Petersburg by Fyodor Vasilyev.

Comments

[thompson]

Easy:

'Academia is where you are judged by the quality of your ideas and your work, not like the political BS you have to put up with everywhere else.'

That was me as a fresh graduate student, many years ago. It was very quickly made apparent that there is as much political BS in academia as there is in every other human endeavor.

For a prediction that has never steered me wrong:

Wherever you set the bar, about 50% of your students will slip in just below it, and about 50% will slip in just over it. So you might as well set it high.

[Stephen Kriz]

I haven't pulled an all-nighter since college. Sorry. Accountants don't do all-nighters. We are too boring.

[wj]

I wouldn't mind the occasional all-nighter. But in my part of the IT world, what gets me down is when I end up spending all weekend on something as well.

My biggest predictive failure was that I thought, if I wanted to keep learning new stuff, I was going to be spending lots of time with University Extension courses. Then I discovered that, if you work with computers, you can never stop learning new stuff. The whole field is in such a constant state of flux that, if you stop learning, you drastically narrow your options -- to close to zero.

All of which is fine for me. But there are a lot of people who were really comfortable with not having to learn much of anything new after they hit 25 or so. And they bitterly resent it when they have to.

[hairshirthedonist]

I don't pull all-nighters, but I do procrastinate, somehow managing never to let anything slip long enough for anyone to notice. I find my work is a lot like school in that way for me. I'm not entirely disciplined or organized, but I land on my feet.

Accountants don't do all-nighters? Not any accountants, of any type, ever? I could see not many doing them, but I imagine some field where lots of work shows up at once, maybe last minute, and it all has to get done at the same time. But that's not really based on anything other than my personal musings in an accountancy vacuum.

[joel hanes]

In my current job (CAD scripting, in perl, for a large chip design team), all-nighters are ultimately counterproductive. About every other year I try one, but the things to be done can't, in general, be done in an 18-hour push; the basic unit of productivity takes about 50 hours of mindful work to complete if I'm doing a good job -- and given the difficulty of remaining mindfully in-flow, that's usually the content of a month of regularly going to work.
Also, the work I produce during an all-nighter is likely to contain errors of analysis and judgement that must be repaired later.
Also, I'm too old to gracefully recover from an all-nighter by sleeping in a couple hours extra the following day.

[bobbyp]

I occasionally pull near all-nighters in the local card room playing hold'em...an old vice. As they say, there are only three things that will kill a confirmed poker player: Three meals a day, exposure to bright sunlight, and stab them through the heart with the ace of spades.

Unfortunate prediction: Standing on the porch of the collective one sunny day in the early, early 70's, looking out over the scenic vistas of Pullman, WA, it came to me in a blinding flash that there would be, much to my dismay, no imminent proletarian revolution.

But there is always hope. Keep on truckin'.

[Marty]

CFO's for public companies often pull all nighters getting the a Board Package ready. They are accountants usually.

[thompson]

Joel:

all-nighters are ultimately counterproductive.

I agree, in general. I find bursts of activity are occasionally quite helpful (grants, paper deadlines, etc), as long as it's in a backdrop of regular hours and regular sleep.

I can do one and be productive and useful. I can't do half-nighters for any stretch of time and do good work.

I know people that are constantly working long hours and I feel they don't really get more done. Like you said, they seem to spend a lot of time fixing what they've already done.

[Roger Moore]

There's a reason big business was willing to accept a 40 hour week; they discovered you got about as much work out of someone in a 40 hour week as in a 60 hour week, but the capital intensive equipment they used was available for more shifts. Long hours can be productive in the short term and can be essential in an emergency, but shorter, regular hours are a better long term plan.

[liberal japonicus]

I used to drop the all-nighter for various things, but since I had emergency surgery for a detached retina, I find that my vision gets blurry if I work too long, so unfortunately, I can't anymore. Even more unfortunately, I have not been able to find a way to compensate yet.

Not sure what level these inaccurate predictions should be at, but I guess it still surprises me at how long things can go on even when the systems seem to be completely broken, especially here. I always think 'ok, now they are going to have to do something', but they never do.

[Countme-In]

I pulled all-nighters in college and sometimes for work in order to squeeze in MORE procrastination during the remaining desperate hours before whatever deadline loomed like a guillotine.

Dawn's pale light would find me staring me into space like a ne'er-do-well vampire.

My college library, like many do, had an all-night study room. You could stake out a study carrel for a quarter/semester, so I would stack my precious books and paper and pencils (no computers back then, which seems like five minutes ago, or maybe a million years) neatly in one the first week and then stop by several nights a week as the witching hour approached and visit them, as one might visit the tomb of Lenin, or one's favorite diorama at the natural history museum.

The carrel might as well have enclosed in glass for all the work that got done.

I might have even have made a show of sitting down for a bit. But usually, after some preliminary page-turning or straightening of the writing implements, I'd sigh and head off into the night for whatever rambunctiousness could be had.

There were couches (big mistake) so sometimes I'd have a good lie down and wake up the next morning completely refreshed for bullsh*tting my way through the next exam.

If they had made the room anything but a study space, say, a train station, or a squash court, or a bar, or even a bloody classroom, I would have been highly productive and read the night away like a Torah scholar, for to this day, I always have a book with me and read when everyone else is using whatever facility it is for its intended purpose.

I carry a book into movie theaters (AR-15s turn nary a head, but I get looks as I read during the previews). I've been known to carry a book into a baseball dugout.

Haven't taken one to the outfield ... yet, though it occurs to me if we have a particularly effective pitcher.

I met one of the two loves of my life in that all-night study room.

And therein is the story of my greatest predictive failure in life.

That and every time I've said "Bob's your uncle!

He never is.

[Countme-In]

"I might have even have made ...."

My second greatest predictive failure is promising myself that next time I will proofread more closely.

[cleek]

the only all-nighter i did in college was forced upon me by the chemicals we'd ingested. none of us could have slept if we'd tried.

since, the only time i've stayed up all night was due to time zone changes (NYC to Tokyo). so it doesn't really count.

[hairshirthedonist]

I predicted that I wouldn't get married until I was at least 30 and that I wouldn't have any kids.

I got married at 27 and I have 4 kids.

[McKinneyTexas]

the only all-nighter i did in college was forced upon me by the chemicals we'd ingested. none of us could have slept if we'd tried.

Common ground here. I never could stay up all night without chemical additives.

As for predictions: just about everything I ever thought would happen turned out differently.

[Carleton Wu]

Growing up (ie until 30-something), I serially made this wrong prediction: while the adults I could currently observe were by and large petty, vindictive, and banal- at some higher level things were being run by serious people. Maybe wise, maybe well-intentioned, but definitely way past eg screwing up a major undertaking because of a perceived slight or meaningless turf infringement.

That's not to disparage the generous, kind, etc people Ive met along the way, only noting that they were sadly outnumbered and usually outgunned.

[russell]

in the Internet Age, we'll be going back to school for the rest of our lives

oddly enough, i seem to be regressing technically. or, rather, i keep finding myself working with older and older technology.

lately it's all C / C++ on a plain old Win32 Windows server stack. proprietary protocol, no less.

just when i thought i was out, they pull me back in.

i'm basically ok with it. i actually really like C++, the money is good, and there's tons of weird old code around that nobody wants to touch so there's plenty of work if need be.

somebody's got to keep all this crufty old crap running. let it be me.

just about everything I ever thought would happen turned out differently.

ain't that the truth.

[wj]

That is the (possible) upside to having obsolescent skills: when nobody is training anyone new in them, and yet there is still lots of infrastructure which needs them. I find myself this past year coding mainframe assembler -- because my employer has stuff which is written in it, and nobody has trained anyone in it since the early 1980s.

I like to think of it as guaranteed employment for as long as I want to work. Certainly nobody (outside, maybe, India) is learning it. And yet pretty much every big business in America which has been around more than 20 years is critically dependent on programs written in it. And rewriting all that would be a mammoth undertaking.

[jeff]

All-nighters in the design lab was a hallowed tradition when I was in college.
It was some right of passage similar to what Doctors must do as interns.
I recall it as a creative and productive time spent well with comradery.

Cramming the night before an exam also proved successful, if success is not measured in how much is retained a week later.

Bad prediction was believing this sort of life style was sustainable.

[Hartmut]

Greatest misprediction? I assumed I would never have any use for this email and internet stuff and definitely not for private purposes.
Well, and there was this thing about predicting certain plot developments in the Harry Potter novels (e.g. I expected Wormtail to stab Voldemort in the back in a deliberate Tolkien allusion)

[cleek]

oddly enough, i seem to be regressing technically. or, rather, i keep finding myself working with older and older technology.

ditto

lately it's all C / C++ on a plain old Win32 Windows server stack. proprietary protocol, no less.

and, ditto.

C# is a fine language at heart, but i like not having to chase MS around, trying to learn their latest suite of soon-to-to-be-deprecated technology.

i'll stick with C/C++ as long as they'll have me.

[hairshirthedonist]

I wrote some stuff in C when I was in school. Maybe I should brush up and consider a career change.

On cramming, there were certain subject that were extremely crammable. I remember taking Linear Systems and Signals my junior year, doing very little studying until the night before the mid-term. It all just kind of went right into my head, I did very well on the exam, and it went right out of my head almost as fast.

[Brett Bellmore]

I thought the 6800 series processors from Motorola would triumph over the 8086 series from Intel, based on their much cleaner memory architecture and symmetric command set. I didn't realize that throwing huge amounts of development money at a fundamentally flawed product could beat an apparently underfunded though fundamentally better one.

After later going to the trouble of buying a transputer array, and learning Occam, in anticipation of the entry of parallel processing into the mainstream computer world, I gave up on technology predictions.

[SeeMoreGlass]

After two separate Dean's Date-induced all-nighters (also at Princeton) ended in seizures, I swore off all-nighters and have been pretty good about sticking to that. But I guess I have a stronger incentive than most.

[CharlesWT]

Texas Instruments also had some well designed processors. But, apparently, they were too focused on using them for embeded applications instead of general purpose computers until it was too late to overcome the path dependence established by Intel.

[Nombrilisme Vide]

In grad school, I'd pull 2-3 all-nighters a week and never blink an eye. At 27. At 28, it was like a switch had been flipped. One all-nighter by its lonesome would destroy me, and I was utterly incapable of getting any work done while being destroyed. This didn't stop me trying, but I see the wisdom in avoiding the wretched things.

Last fall I looked hard at a job programming C (and maybe a little of that new-fangled C++) for legacy sortation systems, but I didn't get it - quite possibly for lack of trying. I'd be happy to never touch C/C++ again. Java has corrupted me too thoroughly.

(I'm sometimes struck by the peculiar demographics quirks of the commentariat here; currently apropos would be our more than fair share of IT backgrounds...)

[Slartibartfast]

I have worked as late as 2am very recently, but I find my real breakthroughs come from better thinking rather than harder and longer work.

And sometimes I get them from other people that just plain aren't around to talk to at 2am.

Of extremely lately, I have some family issues to worry about, and so working overtime just doesn't happen. It's basically irrelevant where family is concerned, which is the whole of my current focus. It's not even something that friends or other family can help with, unfortunately, so they basically don't even know. How do you explain something that you don't yet understand to each and every one of the members of your immediate family?

You don't. You take care of what there is to take care of. Learn what there is to learn. Then, later, you may tell some of it that has been condensed through a long process of understanding and distilled down to a sentence or two.

That's an all-nighter of a different sort, maybe; that struggle to understand and deal.

[thompson]

NV:

I'm sometimes struck by the peculiar demographics quirks of the commentariat here; currently apropos would be our more than fair share of IT backgrounds...

I was noticing that. There are a few legal professionals, a CPA, and many IT professionals.

I've done very little coding, only as required. Formerly FORTRAN, now mostly MATLAB (not real programming, but useful to me). I stumble through a little C and Java (very) occasionally because of available libraries.

[CharlesWT]

I worked years as a programmer until I decided I wasn't good enough at it to justify the stress. Especially since not being good enough at it was causing a lot of the stress. Most of my experience is with the older languages like Assembly, RPG, COBOL, FORTRAN, PL1, Basic and Pascal. I never made the effort to learn C very well. Too addicted I guess to Pascal's syntactical sugar.

[bob_is_boring]

NV (again):

(I'm sometimes struck by the peculiar demographics quirks of the commentariat here; currently apropos would be our more than fair share of IT backgrounds...)

I guess I'm different; I have an IT foreground.

[dr ngo]

Completely different, hence probably irrelevant story, about wrong predictions.

When I was young I was a huge sports fan, especially of UCLA football. In 1953 the only regular season game they lost was an upset to Stanford (20-21, if memory serves). In 1954 - I was ten years old - they started the season really well, but then there was Stanford on the schedule, and I just had this feeling, this horrible intuition, that it would happen again.

UCLA won. 72-0.

Since when I have never trusted my intuition about anything. I'll listen to it occasionally, just to see if it picked up anything my intelligence missed, but I don't trust it. And I don't predict much.

I also have never been a programmer and never pulled an all-nighter (for studying purposes). I think I may be what statisticians call an "outlier" here.

[wj]

What is the relevant characteristic of IT jobs, and shared by law (and maybe CPA work, for all I know)? It is that those doing it are mostly self-supervised. That is, we take breaks when and as we please -- all that our supervisors care about is that we get the job done. And if our breaks include commenting on blogs, nobody basically cares. And that, I suspect, explains a large part of why our demographics are what they are.

[Brett Bellmore]

Not an IT professional. I dual majored in computer engineering, (The hardware side.) and human biology, planning on a career in the medical technology field, with bionics as an option if that took off.

But for a variety of reasons, (For one, had to drop out of college in my senior year to nurse my mom back to health after a really bad auto accident.) I ended up working as a tooling designer/engineer, (Apprenticeship route.) though I tried to keep up my programming skills for a while, and did some hobby robotics. (I'd still like to get into that again, but it's an expensive hobby...)

[Brett Bellmore]

But WJ has, I think, identified the commonality. I work at a computer, and nobody minds if I go websurfing on my breaks, so long as I get the work done.

[Slartibartfast]

When I was much younger, I predicted that I would invent stuff, mostly spaceships.

When I was a little older, I was thinking more that I would be flinging nuclei and even subatomic particles at other, similar objects at high speed and learning things from the results. Or working on some other kind of frontier in physics.

The reality has turned out that I am a guy who does a lot of applied math for a living, and has occasionally worked on rockets that don't carry people.

That could be disappointing, but isn't. My life is a gift. Maybe not one that you, dear reader, would want, but it's one that I am grateful for.

[Countme-In]

Slarti, I hope your family issues resolve themselves to the good.

"Pascal's syntactical sugar"

Nice turn of phrase, that, Charles.

Regarding all-nighters, I was an inveterate night-owl earlier in life, but I hit a wall maybe 15 years ago and now to bed by 10:00 pm at the latest.

Then, about five years ago, as an accompaniment to some "rearrangements" in my life, I began waking up at precisely 2:00 a.m. to mull things over. This pattern has gradually shifted a bit to 3:30 a.m. on the button maybe four or five nights a week at present.

I turn on the local classical music radio station, and get a lot done in these wee hours, whether reading, or studying and thinking about how I approach the stock market, which is a passion, and, I guess, my job.

My mind is much more settled and clear in the early hours and I wish I'd figured that out when much younger.

I do need an afternoon power nap once a week.

So, I guess I do pull something like an all-nighter most of the time. Dairy farmer hours, more like.

I couldn't program an alarm clock, but I suppose I share the commonality wj mentioned in that my boss (me) doesn't appear to mind my blogging as long as I get the work done, but I can tell he worries, the wretch.

Wait, here he comes now. Lay low.

Whew, that was a close one.

Regarding family issues, I head back today to Pennsylvania to help out with my mother sunk in dementia. Some other "issues" with the siblings have arisen as well, so I gaze eastward out my window from my perch 1400 miles to the west of the homeland, and I can see the volcano of Mordor flaring and reddening the sky with its foreboding light (a low rumble, too), as I prepare to make this trip.

Maybe a gigantic spider will descend on me at the airport and sting me unconscious, right in the solar plexus, and I'll be none the wiser.

Samwise, don't fail me now. Sam?

[CharlesWT]

"Pascal's syntactical sugar"

Nice turn of phrase, that, Charles.

Not original with me. And the more correct phrase I should have used is syntactic sugar.

[Barry]

wj "And rewriting all that would be a mammoth undertaking."

It's the replacement which is the greater threat, IMHO. When somebody has a turnkey solution, and all that the customer has to do is set the various options, and load the org tables and variable lists[1]. I'd heard that they did a lot of that ~2000, when they realized that they could (a) rewrite the old systems or (b) replace them with newer, more internet-capable systems, and spend far less work tweaking them to give the right output.

[1] Note that this is what the vendors *say*, and that if the customer has to hire a crash crew to actually make it work - oh, darn.

[thompson]

wj:

It is that those doing it are mostly self-supervised.

Agreed. I'm pretty much straight basic research, some teaching. A lot of random hours, a lot of hurry up and wait.

I try to fill my time usefully during waiting catching up on lit review or what have you. But after a few long days...I become less useful during downtime.

[wj]

Turnkey "solutions" are fine . . . if you currently have a manual system. Or if your current computerized system is just not working for you.

But if you have the sort of highly-customized systems that most large firms have in place, "turnkey" is likely to resemble "turkey" -- at least until you spend a lot of time and effort getting it customized to resemble your existing one.

Which won't keep vendors from pushing them, of course. Or keep foolish executives from buying them, and then complaining that their staff failed to magically make them instantly compatible. The lucky companies will be those which scrap the turnkey solution before they have dropped their existing one -- rather than after they have done so, and let go all of the folks who knew how to keep it running.

[hairshirthedonist]

All the best, Slart. You're a good one.

I turn on the local classical music radio station, and get a lot done in these wee hours, whether reading, or studying and thinking about how I approach the stock market, which is a passion, and, I guess, my job.

My mind is much more settled and clear in the early hours and I wish I'd figured that out when much younger.

Wasn't there a post, maybe from Doc Sci even, about how many people didn't sleep for 8 hours straight until, maybe, the Industrial Revolution? Or was that something I heard on NPR?

Either way, I think I remember reading or hearing that something like the pattern you're describing, Count, was much more the norm, at least within certain segments of society, in years past.

Does anyone else recall or know something about that? Or am I nuts? Or both?

[russell]

there was book published a few years ago called "at day's close" by roger ekirch, which presents the case for a segmented sleep pattern in pre-industrial society.

basically, it appears that people would go to bed when it got dark, sleep for a few hours, get up and do stuff for a while, then go back to bed until it got light.

ekirch was on the radio and TV a lot at the time talking about the book.

[russell]

At Day's Close, by A. Robert Ekirch.

[CharlesWT]

Segmented Sleep

[hairshirthedonist]

Thank you, boys. (in the voice of Perry Farrell)

[Slartibartfast]

All the best, Slart. You're a good one.

You don't know how much things like that mean, at present.

There's almost certainly N stages of emotional response to when your kid goes off the rails; guessing heavily, step 1 is horror, step 2 is resolution to Do Something Appropriate, step 3 is hope, and step 4 is fury.

Not going to fill in the blanks, there; sorry. Anyway, it should be easy to guess that we are in step 4, and we have no thought or care for what step 5 is. But I am dreading it.

On the bright side, everyone is still alive, and most of us are still living under the same roof.

[James Stoner]

Worst prediction (in 2000): "It will be great if the Republicans nominate George W. Bush--he can never win the general election."

All-nighters: No, though writing my blog posts (http://chinshihtang.blogspot.com) does sometimes keep me up until 3.

Going back to school: No, I've had enough. Getting my M.S. in Statistics was the hardest thing I've ever done. In business, you learn by doing.