## The Email Event Horizon

I know I’ve been gone from the shtetl too long—I even stood by as a P=NP goon performed a drive-by shooting through my comments section.  Part of the explanation, I’m ashamed to admit, is that I’ve been procrastinating by proving theorems and writing papers, rather than building up the massive corpus of blog entries on which my tenure case will undoubtedly rest.

But most of my absence has an unhappier source.  At an unknown time about three weeks ago, I crossed the Email Event Horizon—defined in General Unproductivity as the point beyond which you could literally spend your entire day answering emails, yet still have more emails at the end of the day demanding immediate attention than you had at the beginning.  Not spam or crank mail, but worthy missives from students, prospective students, high-school students, secretaries, TAs, fellow committee members, conference organizers, visit hosts, speakers, editors, co-editors, grant officers, referees, colleagues … everything, always, requiring you to do something, commit to some decision, send a title and abstract, pick dates for the trip, exercise Genuine Conscious Thought.  No one ever writes:

Please respond to the situation described above by cracking a joke, the less tasteful the better.  You will never need to deal with this matter again.

I don’t know the precise moment when I crossed the EEH—there was nothing to herald it, it felt like any other moment—but it’s obvious now that I’m in a new, unfamiliar causal region (and that, while I might have thought I’d crossed years ago, I hadn’t).  Communication from inside the EEH to the external universe is theoretically possible, but like Hawking radiation, it tends to be excruciatingly slow—and when it finally arrives, might simply regurgitate the incoming information in garbled form.

When I was a student, I used to wonder constantly about the professors who’d ignore my long, meticulously-crafted emails or fire off one-word replies, yet who might suddenly have an hour for me if I walked into their offices.  Were they senile?  Rude?  Did they secretly despise me?  Now I get it, now I understand—yet I doubt I could explain the warped spacetime Gmailometry I now inhabit to my own past self.  On the other hand, the recognition of what’s happened is itself a sort of liberation.  I’m starting to grasp what’s long been obvious to many of you, those who crossed the EEH before I got my first AOL account in seventh grade: that it’s useless to struggle.  By definition, the speed required to escape the EEH exceeds that of typing, while the mental energy required to accelerate a massive, resting theorist to such a speed is infinite.  So there’s nothing to do but blog, goof off, prove theorems, let the starred-but-unanswered inquiries pile higher and higher, and await the Email Singularity in my causal future.

### 58 Responses to “The Email Event Horizon”

1. Carl Says:

This is why rich people have secretaries.

2. Daniel Reeves Says:

Amen! I think social conventions and expectations are seriously lagging the technology and something is going to have to give.

In the meantime, here’s a desperate measure:
First, archive everything in your inbox. Next, add a filter to mark everything as read, always. Now checking for new mail becomes brutally tedious, since it’s lost in the sea that is your inbox… unless you don’t let your inbox become a vast sea. If you hold fast on the nothing-unread filter then the only way to keep your sanity is to pretty much empty your inbox every time you check your mail. This will both force you to prioritize and quickly cure you of obsessively checking for new mail.

A less drastic solution is to gradually “kibotz” your inbox down to size: kibotzer asks you every so often the size of your inbox and you make sure to keep the number decreasing. E.g., kibotzer.com/dinbox. (Disclosure: I’m a co-creator of kibotzer. It’s in private beta, but we’ll bump you to the top of the list if you mention Scott Aaronson!)

3. Mark Wilson Says:

I guess this comment is more likely to be read than an email to you. Read Getting Things Done by David Allen, or a summary of it. Read Merlin Mann’s series Inbox Zero and apply it.

I have noticed a huge increase in email as my career has progressed. But most of it really doesn’t require much action. I am sure the above references will help.

4. Greg Kuperberg Says:

If you think that it’s bad now, try having children.

5. rrtucci Says:

You need to learn more about upper bounds 🙂 Maybe you should force yourself to give only X talks a year that require air travel.

6. rrtucci Says:

I think you should answer some of the emails by: I’m not sure. Maybe my friend Greg Kuperberg can help you. Here is his email.

7. Raoul Ohio Says:

All,

Speaking of singularities, event horizons, etc., a recent article in Scientific American maintains, among other things, that quantum mechanics and relativity are incompatible. I need to read it again at least twice, with more coffee. Anyone betting on which is wrong, or at least familiar with the argument?

8. knausrich Says:

I always wanted to read “Getting Things Done”… but I never got to it.

9. hawk Says:

“Please respond to the situation described above by cracking a joke, the less tasteful the better. You will never need to deal with this matter again.”

rofl i’ll try doing that.

10. Frank Says:

Well put… Dude, your the man! Ditto to Mark Wilson’s advice: READ GTD by Allen, it will change your life. OR-declare email bankruptcy- send one apology to all on your list and only answer the most important messages.

11. Scott Says:

Raoul, for the reasons explained in the post, I’ll leave it to others to pick apart that article. I’ll just say I read it and found it a carnival of the confuseniks. Firstly, it seems strange to present any of this as news—these questions have been debated since the 1930s, and (contrary to the article’s claim) almost nothing new has been said about them for 40 years. Secondly, maybe two-thirds of the way in, you understand the semantic game the authors are playing: once you define special relativity in such a way that it has to be classical, voila—you can make the earth-shaking discovery that it’s incompatible with quantum mechanics! But superluminal signalling is the thing that (when combined with SR) would lead to bizarre physical consequences like backwards-in-time causation—and the fact that quantum mechanics doesn’t allow superluminal signalling is brushed aside in this article with incredible indifference.

Look, the notion that you can have effects that

(1) can’t be explained by local hidden variables, but
(2) don’t entail faster-than-light signalling,

and that moreover such effects are actually realized in the physical world, really is profound and important, and 99.9% of people who’ve read popular accounts of quantum mechanics don’t get it. But rather than explain this crucial point, this article aims (and will no doubt succeed) at confusing it further.

Having published in SciAm before, having worked with excellent editors there, I have to admit I’m disappointed. This is something that belongs in New Scientist.

12. Greg Kuperberg Says:

Having published in SciAm before, having worked with excellent editors there, I have to admit I’m disappointed.

Their editors may indeed be excellent, but their mandate is pretty bad. This became apparent 15 years ago, by the time of a dubious article by John Horgan entitled, “The Death of Proof”.

Meanwhile there is a more optimistic article today in the New York Times on the same subject as your post.

http://www.nytimes.com/2009/03/05/technology/personaltech/05basics.html

13. svat Says:

Concur with others’ recommendation of the “Inbox zero” program. In short, the idea seems to be to read email once a day (others will just deal with it), and try to keep zero emails in the inbox each time, using Delete (“Archive”) ≥ Delegate ≥ Respond (briefly) ≥ Defer (actively, not leaving it there!) ≥ Do (quickly).

[Random comment: the CSAIL event calendar has this talk for next week:
“On P vs NP, Geometric Complexity Theory, and the Riemann Hypothesis”
Speaker: Scott Aaronson, CSAIL, MIT
Host: Ketan Mulmuley , University of Chicago

A simple error probably caused by the EEH, but would be awesome if it were true :)]

14. Crystal Turjillo Says:

Scott,

What constitutes a “goon?” In all fairness, wasn’t it always about figuring out the truth? Isn’t the whole idea about testing the bounds?

15. KaoriBlue Says:

I thought “New Scientist” was supposed to be higher quality than SciAm? Kind of like a journal for teenagers vs. something more sensationalist?

And it looks like you’ve also hit an event horizon for a deluge of kinda scary and incoherent blog comments.

16. Scott Says:

KaoriBlue: Unfortunately, yes. “Martin M. Musatov” has been posting dozens of bizarre, incoherent comments at a time, now mostly under assumed names. He has been banned permanently from this blog.

17. asdf Says:

You could follow the trend of falling into email bankruptcy:

http://gawker.com/tech/trends/declaring-e+mail-bankruptcy-254608.php

I guess it fits in with the general economic collapse. Perhaps Obama could launch a special stimulus package just to help people catch up with email.

18. der Says:

Beyond the horizon, Structured Procrastination” looms.

19. Scott Says:

That’s one of the better essays on procrastination I’ve read! Don’t wait—read it now!

20. Jonathan Vos Post Says:

“I thought ‘New Scientist’ was supposed to be higher quality than SciAm?”

IMHO, ‘New Scientist’ is roughly equal to SA in quality of Biology, Ecology, Psychology, and some other subjects. But ‘New Scientist’ is far, far behind Scientific American when it comes to Physics. Greg Egan and John Baez have blogged about that extensively.

I am mulling over — right now — whether or not to renew the ‘New Scientist’ subscription in my son’s name for another 3 years.

I made a too long or to incoherent prior submission on the emerging interplay of Facebook, LinkedIn, Twitter, and Email. But the CEO of LinkedIn was quite coherent on Charlie Rose last night. Their goal is to have 1/4 of the world’s population in their social network. They admit that this requires acceleration, as they only acquire one new professional per second.

Networks of Trust help to solve the email crisis that Prof. Aaronson describes. Want to get a message to the CEO of Facebook or LinkedIn? This can only be done if someone one link away from them vouches for you.

The social network is NOT a random graph. Random graphs provide asymptotic bounds. The WWW is NOT a random graph. Google indexes 10^12 pages now. Is quantum computing necessary for the Semantic Web? I don’t know. But polynomial time algorithms are surely needed for grappling productively with interacting social networks.

21. Greg Egan Says:

I thought “New Scientist” was supposed to be higher quality than SciAm?

At New Scientist they don’t even believe in conservation of momentum. (Actually it’s worse: they don’t believe in it, but they understand it so poorly that they don’t believe they don’t believe in it.)

22. Abel Says:

Well, the essay reminds me of a theory I came out with a couple of years ago, thinking about my days in high school… I could remember the names of most Pokemon then, while I could not remember most things from courses.

I thought this was because every time I let my mind wonder (when studying, or doing menial things), I ended up thinking about video-games, so maybe it would be more effective trying to do what you do not need to do, while thinking about what you need to do, instead of trying to concentrate on what you have to do, but really think about other things… (A friend I told that actually managed to raise his high school GPA by about 10% by doing it)

23. denis bider Says:

Er,

I suggest a market-based approach to stem the flow of incoming email. Granted, it would require convenient micropayments to be integrated into email clients, but that’s but a trifle compared to the benefit of being able to impose a cost barrier to people emailing you.

You could then just raise the cost of an email to where you can keep up with them comfortably. 🙂

24. Job Says:

I don’t have this problem. Sometimes i go into the spam folder just to read something.

25. cool-RR Says:

Really Scott, why not hire a secretary?

26. Scott Says:

I have a secretary (shared with a few other people), and she’s extremely good, and she handles what she can.

27. Job Says:

There are days when i don’t even get spam.

28. Greg Kuperberg Says:

I have a secretary (shared with a few other people), and she’s extremely good, and she handles what she can.

And you’re still complaining? Geez, Scott, what happens if you don’t get tenure and you have to go to a peasant university?

29. sd Says:

This… explains… everything! Being a grad student, I am often on the receiving end of these ignores and one-word responses. I must say I instinctually find it upsetting. I often sit down and say to myself: “Calm down. He/she doesn’t hate/secretly dislike/not care about you, but is just overwhelmed with too much to do and is just waiting till he can give your email/plea/proof/request the time and attention it deserves.” However, I would rarely succeed at completely convincing myself of this. Your post puts everything in perspective, and should be required reading for all new grad students.

On another note, is it a bad sign if, as a grad student, you find yourself inching ever closer to the EEH? Is this normal?

30. Greg Egan Says:

From the troll-your-own section of the Wikipedia article (though doubtless soon to be deleted):

… any input string of length Failed to parse

Who knew that Wikipedia had a built-in bullshit detector?

31. Scott Says:

is it a bad sign if, as a grad student, you find yourself inching ever closer to the EEH? Is this normal?

sd: Alas, yes. I’m so sorry.

Greg K.: I thought MIT was a peasant university. I mean, we’re not Harvard. 🙂

32. Mar Says:

p=npL: “Hey, Mom, are we there yet?

33. .m. Says:

oracle: p=np=the Internet

34. W Says:

Neo: M of Y (bg) i>

35. Trinity Says:

Scott, you’re so handsome.

36. Trinity Says:

Will you help out my friend here, tell him what he’s doing wrong? http://www.scribd.com/doc/12694681/A-Solution-to-P-Equals-NP

37. Trinity Says:

3:33, see it’s time…

38. Morpheus Says:

We gotta move…

39. Operator Says:

“17 Responses to “The Matrix Is On Line Two”. David Poe Says: … Don’t even get me started on the autodialers who call us to extend our car’s warranty. … “

40. Neo Says:

“We’re live.”

41. Morpheus Says:

“Get the site.”

42. Scott Says:

“Martin Musatov”: I’ve logged your IP address, and reported your repeated abuse to your hosting provider. Please just go away, so that I don’t need to pursue this matter further.

43. matt Says:

Lots of people (especially professors) say things like this, and mostly I think it’s due to bad time management. Let me mention 2 scientists I know. One is an outstanding soft matter theorist, who moved to industry and acquired some very high-up managerial responsibility. Another is also an outstanding soft matter theorist who has a permanent academic-type job with no teaching requirements. The one who complains the most about being too busy is the one with the academic job, and yet by any reasonable standard of their job duties, he has got orders of magnitude less responsibility.

Do this: if you get email, you go through it in the morning, delete the junk, hit reply-to on the stuff you know you should reply too, and leave the reply window open. Now, come back later and just reply to each open window….should take a couple minutes max per mail. Do it, and it’s done. If you can’t reply for a while for specific reasons (i.e., need to know something else to make your plans), just say that, and again, it’s done. Similar for refereeing papers: if you get a paper to referee, decide if you will do it, and see if you happen to have an hour or two free that day. If you do, sit down, read it, send in your comments, if not, decline to review. Done.

Anyway, for your own good, I want to say that I’d bet the problem is not too much to do, but bad organization. Work at a national or industrial lab for a while and you will learn the art of disposing of things quickly (bet you never have to take safety training at MIT, lucky you).

44. matt Says:

Also, learn the term “action item”. That’s key…sounded like some weird industry jargon when I first heard it years ago, but it works.

45. Scott Says:

Matt, I completely agree that the problem is bad time management. Alas, I spent four summers working at Bell Labs (where I did have to watch the safety video, and also take a drug test), and it didn’t cure me.

46. Jonathan Vos Post Says:

As with many very successful people that I know, I maintain a daily “To Do” list. Usually, there are about 20 to 25 tasks on my list.

After I write them all down, I rate each according to two criteria: how important, and how urgent.

There are 4 combinations to consider:

(1) Unimportant and non-urgent: then why bother? I’ll get around to it if time and mood permit.

(2) Unimportant and urgent: got to do it by a deadline real soon, but it’s only important because it’s important to some else whom I want to please, or because if I don’t do it that could lead to something important and bad.

(3) Important and not urgent: maybe I’ll spend a few minutes getting the notes and references or emails or snailmails or whatever in a folder, for when I need to get to it and finish it, but not do it all right now.

(4) Important and urgent: these get marked with an “A” on my To Do list. I intend to get every single “A” task done today. Usually no more than 2 or 3 of these. Thank God!

Then go through the rest of the list, and mark the ones that don’t rate “A” but are important or urgent enough to rate as “B.” I’ll do most of those, and let the rest slide until tomorrow.

I’ve toyed with adding a 3rd bit for cost/benefit in estimated dollars or hours, but then I might as well use formal Decision Analysis or get all Bayesian. So the 2-bit filter is okay for juggling research in 3 or 4 fields of academe, plus Appellate and Supreme Court procedures, plus financial/tax management of multiple small businesses, plus homeowner/spouse/parent duties, plus writing and publishing science fiction and poetry and stuff in the so-called “free time.”

There’s a good thread on time management in Terry Tao’s blog.

47. Dave Bacon Says:

Scott I recently learned a new word: “[deletia]” (from “microserfs” by Douglas Coupland.) When someone sends you an email you delete the whole text and send back an email with “[deletia].” It stands for everything that’s been lost.

I’m happy to see that I’m not the only one who just wasted precious moments of my life deleting entries by “Martin M. Musatov.”

48. Job Says:

I have been following structured procrastination for years, unknowingly. I actually described this process in a software blog a while back as being very productive for personal projects – i believe i put it as “do the easier stuff first”. I got replies of the sort “You’ll never get a job in my software team”, etc.

It works for me. Just last month i started and finished a whole site (http://www.youtensils.org) that i had no reason to work on, just because i didn’t feel like working on another project, which i’m currently working on actually because i don’t feel like paying my bills or doing my taxes. Eventually it all gets done, that’s the beauty of it. 🙂

49. Job Says:

Stupid bills & taxes.

50. Jonathan Vos Post Says:

Nothing is certain but death and taxes and science writers asserting that Quantum Computing is magic.

There was the year when I owed no taxes, and the California Franchise Tax Board wrote demanding (I can’t make this up) $0.00 under threat of garnishment. I paid$300.00 to file a return that proved that I owed nothing. Then I wrote a letter beginning: “Dear Computer: I know that no human looks at your output..” and cc’d my Congressman. The Tax Man wrote me a letter “We apologize. Errors were made. File closed.”

I spend roughly 100 hours per year doing taxes. Since it saves about $10,000 a year (all those deductions we professional writers take) it’s worth$100/hour — but still hideous.

51. David Says:

All I can say is good luck. It is interesting to watch people make the transition from grad student/post doc to professor. As a soft money guy, I tend to feel compelled to write long careful emails with detailed bulleted points and I tend to get no replies from students and replies from hard moneyed bosses that say “good.” I appreciate the one word boss replies and don’t like radio silence. I tend to give radio silence to my mom which isn’t so smart of me.
At the moment, I’m sacrificing other bureaucratic stuff (getting reimbursed, safety forms…) rather than long-winded emails and blog comments.

52. gowers Says:

Great article, and comforting to see that it’s not just me. One technique that I occasionally manage to apply is performing tasks suboptimally. I reason as follows. Suppose that somebody wants me to write a reference for a paper. The optimal response would be to send, within a short time, a decision about the paper accompanied by a detailed justification of that decision. But if it’s a choice between a quick decision that’s not adequately backed up and a slow decision that’s backed up very carefully and convincingly, then it’s pretty obvious what the journal will prefer.

Using this obvious principle has slightly improved my efficiency (which is a polite way of saying reduced my inefficiency), but there’s a long way to go.

Actually, what I often end up doing is the worst of all: eventually a time comes when I feel so guilty about not having done the task that I then do it quickly and suboptimally. It’s as though the probability distribution that determines what I do and when I do it is memoryless.

53. Pat Cahalan Says:

@ Greg Kuperberg

Try a Kinesis. I had the same problem. The Kinesis is not only ergonomically terrific, it’s programmable. It’s one of the few geek bits of hardware that I’ll endorse pretty freely.

It is damn expensive, I’ll admit 🙂

54. Greg Kuperberg Says:

Well, my solution is a programmable keyboard, but I did it the Linux way, for free. This solution has the advantage that I don’t have to tote a special keyboard with me. For instance right now I am on my laptop with no special hardware.

I have been trying to get the attention of wrist specialists on the Internet so that my point can be added to general medical wisdom. Namely (a) that pinky RSI is a special condition which is more common than standard medical advice suggests. And (b) it’s partly caused by a very specific keyboard layout problem; it is not well addressed by generic solutions like “cut back on your typing” or “try one of these five special keyboard brands”. But I have not been able to get any expert interested.

55. John Sidles Says:

Greg Kuperberg, your web site’s aphorism “Many problems in the real world go unsolved because they are eclipsed by more interesting but less pressing problems.” is admirable … I immediately added it to my database of quotations.

To oversimplify, you have crisply expressed the reason that a cultural emphasis on rigorous proof is well-advised in mathematics and physics, yet ill-advised in engineering and medicine.

Also, may I say that your summary of tendonitis and the benefits/perils of surgery is medically right-on too? If that’s what your physicians are telling you, then IMHO you have good physicians!

56. Greg Kuperberg Says:

The Kinesis is not only ergonomically terrific, it’s programmable.

Well, my solution is a programmable keyboard, but I did it in software, for free.

57. Patrick Says:

If anything is important enough, the person will eventually call or stop by.

Let human persistence be your spam filter.

58. Jack in Danville Says:

Scott,