So we (Richard Hollerith and me) tried out my anti-akrasia idea. Actually we've been doing it for more than a week now. Turns out it works just like I thought it would: when you know an actual person is checking your screen at random intervals, and they will IM you whenever you start procrastinating online, and they expect the same from you... you become ashamed of procrastinating online. You get several "clean" hours every day, where you either do work or stay away from the computer - no willpower required. Magic.

Proofpic time! Once we both left our VNC windows open for a while, which resulted in this:

The idea isn't new. I first got it this winter, Alicorn and AdeleneDawner are apparently doing similar things unilaterally, and even Eliezer has been using a watcher while writing his book. I don't know anyone who tried the Orwellian mutual screen capture thing before, but I won't be surprised if a lot of people are already quietly practicing it.

Being watched for the first time didn't make me feel as vulnerable as you'd think, because, realistically, what can the other person glean from my monitor while I work? Random screenfuls of source code? Headings of emails? We don't realize how normal the little details of our lives would look to strangers. In the words of McSweeney's, "chances are, people will understand. Most people are pretty understanding." The experiment did feel weird at first, but it was the expected kind of weird - the feeling you should get when you're genuinely trying something new for the first time, rather than just rehashing. It feels normal now. In fact, I'm already ever-so-slightly worried about becoming dependent on remote monitoring for getting work done. You decide whether that's a good sign.

Passing the microphone to Richard now:

I had to set a timer (for between 5 and 11 minutes depending on circumstances) to remind me to check Vladimir's screen (resetting the timer manually after every check).  If I did not, I either spent too much time looking at his screen or let him go too long without monitoring.

I tend to think that if I continue to monitor people in this way, I will eventually come to use software (particularly software running on the monitored computer) to reduce the demands on my time and attention, but my more immediate concern is whether the technique will remain effective when it is continued for another month or so or whether, e.g., everyone who volunteers to be monitored comes to resent it.

Because of technical problems, Vladimir has not yet been able to monitor me in my "familiar software environment" and consequently the real test what it is like for me to be monitored has not yet been done. Vladimir has monitored my using a borrowed Windows machine, but I am unfamiliar with Windows, and in general, when I am taken out of my familiar environment, I usually gain temporary freedom from my usual patterns of procrastination.  I did feel embarrassment at how ineffective my use of Windows probably seemed to Vladimir.

In conclusion, the technique seems to help me a lot, even though it's shifting my sleep pattern to somewhere in between Moscow and California. My current plan is to keep doing it as long as there are willing partners or until my akrasia dissolves by itself (unlikely). The offers I made to other LW users still stand. Richard is in talks with another prospective participant and would like more. We want this post to actually help people. Any questions are welcome.

UPDATE one month later: we're still doing it, and everyone's still welcome to join. Won't update again.

New to LessWrong?

New Comment
118 comments, sorted by Click to highlight new comments since: Today at 2:46 AM
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

My roommate and I had a no-internet day last week. It worked surprisingly well. One of the key reasons I didn't crumble and turn our router back on was that I wanted my roommate to be productive. It's easy for me to play off my own need to be productive, but I'm less cavalier with the productivity of someone else.

Just one day? Do you plan to repeat it?
I plan on repeating it again, if I can convince my roommate!
I know that personally this would require a day or two of preperation, but it might not be typical of everyone else that in order to do what I ought to do I need to download the PDFs, watch the videos, consult Wolfram Alpha, etc. On the other hand, the idea of a no-internet day is pretty exciting, as that same workflow is too easy to derail with Facebook or Less Wrong.

Procrastination is the soul rebelling against entrapment.

Nassim Nicholas Taleb, Facebook status update, Sunday Sept. 26.

Procrastination is the soul rebelling against entrapment.

My soul doesn't always have my best interests at heart.

This guy has a lot of stupid beliefs. It's too bad we have to hear about them because he made one good call.

Even if you call it rebellion of the soul, procrastination doesn't seem to be good for my life.
The useful clue in there might be to build a more unified identity and/or more reasonable goals so that there's less tendency to rebellion.
Hm. I parsed this as a joke but others seem to be taking it seriously. Who's confused?
Jokes are often used as a way to present ideas in a form that can bypass social filters. A tendency that I've noticed is more common here than in other places is to consider both the humorous aspect and the literal component of the joke. People sometimes interpret that tendency to be 'not getting it' but that's because they don't get it. It's getting all of the song. The words and the music both. (My reading: The above applies to Tyrrell's comment perfectly. Jonathan is just responding to the name - the content is only superficially relevant. Cousin_it is no doubt aware of the Taleb's intent but taking a far more serious stance. I'm applying a style of thinking analogous to the above paragraph but applied to the kind a query that would otherwise beget a simple social judgement.)
Thanks for clearing my confusion. I didn't expect to get such an insightful answer. It further encourages me to develop my practice of noticing subtle confusion.

I know some christians have used similar practices for years to avoid Internet porn.

There's probably more low hanging fruit there, if you can stand the forest.

(Edit: "Accountability Partner" is the phrase they use.

Thanks to rhollerith_dot_com)

The best way to learn not to look at internet porn is to look at some of it once.

I rather suspect that isn't a universally applicable solution. Breasts look nice.

This only has a template for a clever comeback, without actually making sense. Did I miss something?
I suspect this is a reference to some of the more unseemly parts of the internet. If you want to gouge out your eyes, look up 4chan.
URL or other clue would be appreciated.
1JamesAndrix13y "Accountability Partner" is their key phrase I should have mentioned.

It's about time someone made an antidote to TV Tropes.

This is like pair programming except you don't have to cut your productivity nearly in half.

To clarify: I've always believed that most of the benefit from the practice comes from the fact that work is actually being done for an increased portion of the time in front of the monitor, and that the costs and benefits of discussion roughly balance.

Anecdotal, but my experience of pair programming is that it's incredibly useful for picking up bugs as they are laid down rather than having to dig them up later. Not to say that being monitored working doesn't help, but finding and removing bugs is by far the hardest and most expensive part of programming.

Seems reasonable and is supported by some study linked in Wikipedia. Code review is supposed to do this more efficiently, albeit with higher latency.
Pair programming reduces variability in productivity rate, so looking back on a pairing session you should expect to remember occasions when you felt that you'd been slowed down, and should not expect to have noticed occasions when you'd saved time. Alistair Cockburn has a study up somewhere showing (from memory) a small increase in work hours to first thunking you're finished, a large reduction in bugs, and an improvement in subjective measures if code quality. Trike's experience matches this study - we think pairing is awesome. .

In the spirit of my, I've written a script to sleep randomly 0-10 minutes and take a screenshot & webcam shot. Obviously many of the values are specific to my own computer (such as the names in /proc, where in ~/ the pictures are sent to, and the use of ImageMagick, jpegoptim, and OptiPNG), but it should be easy to adapt to your own computer:

set -e

if grep open /proc/acpi/button/lid/LID?/state > /dev/null
    CURRENT=`date +%s`;
    SLEEP=$(( $CURRENT % 10 ))
    sleep $SLEEP$TIMEOUT
... (read more)
Thank you for posting this. I want to do almost exactly this, and set up another task running that will zip up all the images and send them off to my referee. Unfortunately, I haven't gotten this to work. I'm not very familiar with cron, and I don't know how to debug it. My script works as intended when run manually, but halts (dies? I don't know) when run from cron (at import). I haven't managed to get any error reporting, either by redirecting the script output or the import output. I also haven't been able to get import to run directly from cron, even by using absolute paths and no variables. I suspect these are problems you haven't run across, but if you have, I'd appreciate your advice.
Cron does its error reporting through local email, so when I'm debugging, I'd check /var/mail/gwern. You may need to've chosen 'local delivery' when configuring your install on a Debian system.

I'm not sure how ironic I should find it that my procrastination site is now actively discussing anti-procrastination methods. -rela

You get several "clean" hours every day, where you either do work or stay away from the computer - no willpower required. Magic.

How do you feel about working now compared to before? Are you more enthusiastic, or more reluctant? Are you staying away from the computer as a form of procrastination instead of going to it but not working?

1) As far as I can tell, my feelings about work haven't changed. 2) Yes, now I sometimes walk away from the computer if I don't feel like working. But these pauses are much shorter (because online procrastination makes time pass really quickly, compared to offline), and I always end up doing something productive like writing or playing the guitar. This was actually an unexpected benefit of the practice.

You... you can write without a computer? WITCHCRAFT!!!

Vladimir is sleeping now. I will leave your first Q for him, but answer your second Q because we had a discussion about it. We are in our apartments when we are being monitored, neither of us owns a TV and the internet has been the primary way for us to procrastinate or distract ourselves. We would both continue to consider the experiment a success if all it does is to continue to prevent us from using the internet to procrastinate or to distract ourselves from our work -- it is not necessary for us to work at our computers the whole time we are being monitored. So for example, I spent part of the time that Vladimir monitored me shaving and talking to my apartment manager because that was the most productive use of that time. (Of course, I told Vladimir that that was my plan.)
I see. I had thought your experiment "when I want to be working on the computer, ensure I am not procrastinating on the computer instead" but it is actually "try to ensure I am not procrastinating on the computer".
You have understood correctly. That is what Vladimir and I have defined to be our goal. (But we can probably help people achieve the other goal, too.)

I read this and it immediately shouted at me:

"Chatroullette clone"

I'm not entirely sure how this might work. Maybe something along the lines of :

Every 15 minutes a screenshot of your work is sent to a complete stranger. If they think you're procrastinating, they can hit a "horn" button which causes a loud noise to occur on your computer, and broadcasts 5 seconds worth of your reaction.

This would need some impressive anti-troll precautions. Or at least a way to temporarily disable the horn, so nobody can blast it when you're in the middle of a video chat with a client, but then that'd be a source of temptation.

Alright, imagine that you as the "watched" have a choice on the type of alert you get hit with. And on the number of strangers that have to want to alert you before it happens - maybe your screen gets sent to 10 strangers and they have to achieve a quorum. Or 10 friends. This could be a facebook app. If anyone decides to work on this, please let me know in advance.
One possible solution is to build a set of training data, which with a tagged series of sceenshots and fake responses. So the watcher wouldn't know for sure whether they were getting a real person or old data. heck, you could use that element to moderate previous watchers.
Hi Adelene, while writing the post I hoped that you'd describe the things you're doing among these lines. Please?
My response quickly attained top-post length. I expect to have it posted sometime tonight. ETA: Given the choice of doing it well or doing it quickly, I'm doing it well. I still expect it to be up sometime today.
I'm going to nag you till you post it.
Please do. :) I currently have 7 paragraphs, mostly of specific implementable advice, and am organizing my thoughts for the 8th, which is the part I think you're most interested in. A friend also suggested a few more concepts that I should probably at least mention, too, which might mean shuffling things around a bit.
I am waiting for it, too.
Almost done. I'm contemplating the final paragraph now. I'm probably going to post it without much editing; I'm starting to have trouble staying awake.
Alright, imagine that you as the "watched" have a choice on the type of alert you get hit with. And on the number of strangers that have to want to alert you before it happens - maybe your screen gets sent to 10 strangers and they have to achieve a quorum.

It seems like it'd be obviously worth investing a couple hundred dollars in a second monitor for anyone who wants to do this.

In fact, it seems like it might be worth investing in 2 or 3 monitors anyway... screen space is cheap and might make you more productive? I find myself switching between windows a lot because even minimized windows can take up a lot of screen space.

Here's a small (and probably cherry-picked) collection of studies, which I haven't examined closely, but which seem to agree with my subjective impression: ~50% productivity improvement on tasks involving multiple large simultaneously-relevant windows and tight interaction loops (data analysis, in my case), less but still substantial benefit on programming.
Maybe I am nosier than the average person, but if Vladimir's desktop were constantly in my field of vision, Will, I would spend too much time looking at it. I hide my Viewer window after verifying that Vladimir is not procrastinating. Then I set a timer to remind me to look at the Viewer window again in 5 or 10 minutes. If Vladimir and I had founded a start-up together or I was paying him $8000 a month to work for me, then more likely than not I would prefer for his desktop to be constantly in my field of vision, but the way it is now I do not have that much of a stake in his performance at the keyboard.
It occurs to me that Vladimir probably saw parent, so to keep him on his toes, I plugged an external monitor into my laptop, and am using one of my monitors to keep Vladimir's work constantly in my field of vision (but I tend to keep the monitor showing my work brighter than the one showing his work).
Oh, so that's why you react so quickly!

Enhancement suggestion: time-shifting. Make periodic screenshots and deposit them on the server. Of course this will not work with IM, but the shame factor might just be enough. Plus you can occasionally use "privacy privileges" and delete x% of the screenshots, in case of the obvious.

I could actually see this as a product.

I'm more interested in using the technique than enhancing it or spreading it. Writing special software, making it a product or even evangelizing feels icky because I don't want to become a self-help guru like pjeby (who used to be a well-known programmer in the Python world), I just want to achieve my own goals. It's strange and mysterious to me why many other people don't instinctively avoid roads that are "bad for the karma".
I was also going to mention pjeby here, in the sense that your experiment is not focused on understanding or curing your procrastination, but a way to punish yourself with shame until anticipating the shame of procrastination-by-internet feels so bad that you avoid procrastination-by-internet. Which might work, but my other question here was wondering if it merely pushed your procrastination elsewhere - after all you haven't improved the pressure to work, or removed the pressure to procrastinate, you have only made the computer unavailable for procrastination. Whereas pjeby's approach is in removing the mystery of why you are procrastinating, and then fixing it directly. You did author this post, included "We want this post to actually help people", and commented with contact details to help spread it to others. That shows some nontrivial interest in spreading it. [Edited for less aggressive wording, last paragraph].
Shame is not an essential ingredient in the method. To see this, suppose the monitor has a button which freezes all the applications on the remote machine except for the application for communication between the parties. I do not know about Vladimir, but the only reason I have not been doing it that way is to avoid the cost (in buying or developing software) of implementation of the button. Before the internet, the average office contained no distractions more potent than Minesweeper, Solitaire, annoying office mates and gossip with coworkers. I think of the method as a way to create a place with internet access as free of distractions as a pre-internet office. There are thousands of software developers, web designers, film directors and other very bright people around the world constantly dreaming up more potent ways of using the internet to grab and keep our attention. Probably nothing in human cultural or genetic evolution prepared us to resist that kind of temptation. Certainly, being able to override an instinctual or habitual response was important in human evolution, but the sort of constant temptation represented by on-demand video sites like Hulu is probably unprecedented. Consequently, I am disinclined to believe pjeby or any other self-help or productivity expert if they claim that my procrastination stems from a pathology that can be (your word) "cured" and am inclined to believe that procrastination is what you get when you give a normal pathology-free person access to the modern internet.
I claim quite the opposite, actually.
Want me to delete my "pjeby or"?
s'ok; I did claim it at one point, even sold a product called The Procrastination Cure. (I do not sell it any more, though: despite being 6 CDs in length, it covered only a relatively narrow part of the procrastination spectrum that I've discovered since.) To be clear, though, this is really a language problem. You cannot cure "procrastination", but you can cure an instance of procrastination, such as an ugh field, an unclear goal, etc. The monitoring method in the article will work well for some sources/instances, and not for others. It appears to me to be a digitized version of the ADD "body double" tactic, where having another person in the same room (even if not in any way observing or interacting with you) can improve your focus. So, if you have that type of problem with maintaining focus, one would guess that this technique would work well for it. Despite the increasing forms of addictive distraction on the internet, I have personally observed that when I have some goal that actually interests me, I can go for days without visiting my usual addictive haunts. So, I therefore choose to interpret my desire to visit, say, LW, as an indication that I need to step back and either notice what I'm avoiding, or find a more engaging goal. Speaking of which... gotta go. ;-)
See any downsides to the ADD "body double" tactic besides the cost of keeping the body double cooperative?
If it doesn't work (because your procrastination has different or additional causes), shame could create an additional ugh field or magnify the one you already have. (But that risk is common to most anti-akrasia tactics, and the antidote is the same: realize that until this procrastination instance is fixed, you don't (and can't) really know what's causing it.)
That hasn't helped me see it. It still seems as though, if you didn't care one bit about someone seeing you timewasting, then the method wouldn't work, and no amount of screen locking buttons would change that bit. Even if I consider it a friendly cajoling instead of a remote monitoring, it still wouldn't work if you felt no qualms about ignoring the viewer's cajoling and carrying on timewasting. The unprecedented temptation is a nice idea. I want the difference between me and people who can easily ignore it to be a learned behaviour, so I could potentially learn it, but I have no solid support for suggesting that it is. Maybe it isn't and we are being trapped by addictive superstimulus. [Edit: I see that I am procrastinating now, but I am not on Hulu, iPlayer or any other attention trap site (as I sometimes do), instead I am alternately reading Less Wrong and pacing up and down the room. This is anecdotal support that it is not the attractiveness of certain carefully designed attention hooks which is causing my (current) procrastination, but instead a repulsion away from what I should be doing.]
"The monitor has a button" was a poor choice of words, so I will rephrase. Suppose the person monitoring me has a way to freeze my computer so that none of my keystrokes or mouse gestures have any effect. Do you agree that that would give the person monitoring me a way to stop me from using my computer to procrastinate that does not rely on shame? (It would work better if he had a way to freeze all the apps on my computer except for the app he is using to communicate with me because then he can give me pep talks or negotiate with me while my machine is in the (mostly) frozen state.)
I do agree. That sounds like a slightly different experiment to the one described in the original post, though - such a system could not fail to work. I wonder if cousin_it percieves the experiment differently to you - do you agree with his description """you become ashamed of procrastinating online. You get several "clean" hours every day, where you either do work or stay away from the computer - no willpower required.""" ? (If you wish to try that in your experiment, Dameware Mini Remote Control ( has a "lock remote keyboard and mouse" feature, and a 30 day free trial).
I am not entirely sure what you want to know. Do you want me to tell you how I think the arrangement Vladimir and I have been using produces the effects it does produce? The short answer is, I do not know. And I doubt I will ever know because why I procrastinate and why this arrangement stops it are probably extremely complicated. If I had sent you a short explanation the day before the start of the experiment of why I expected the experiment between Vladimir and I to succeed, the explanation would definitely not have contained the word "shame". I never felt nearly as ashamed while being watched by V as I felt the last time I procrastinated by watching a show on Hulu in the middle of the day. I have never procrastinated while being watched by V except for "sham procrastination" to verify that V is actually watching. V and I were still getting to know each other when he started watching me, so of course I wanted to make a good impression on him like I would with any other person who was investing time and effort in getting to know me. My desire to make a good impression definitely affected my choices of what to work on and definitely caused me to try to work at a faster pace while he was watching. (I want to stop trying to work faster because I have read things that make me believe it does not actually make one work faster.)
I am not entirely sure what I want to know either, but how it does produce the effects it does produce are part of it. Previously the idea of remote monitoring gave me a negative Orwellian feeling. When I first read the book which I quoted in email to you, I was surprised that the author could be comfortable with such a situation, which may be why I remembered it easily. You two doing something similar stood out and I am interested in why you both chose to do it, how well it works and what the cost of reduced procrastination is to you - learned negative feelings about using a computer at all would be a high cost, learned positive feelings about having company while working could be an unintended benefit and pretty much no cost, for instance. It turned out to be more interesting as you are both two halves of the same experiment and getting similar results out of it, but apparently both doing so via different reasons, and might be considering the experimental purpose differently. I don't really have any reason to dig further - thank you for elaborating on it.
I wasn't suggesting you build the product, it was just an idea. Building a screenshot timer wouldn't make you a self-help guru anyway, no fear. I used to read some PJEby python stuff, surprised and personally glad he turned up here, I think he contributes pretty good content while trying to subliminally advertise his product. Can you unpack this? I assume you're still talking about PJ?
That (subliminally advertising our product) is what we all do... PJEby is just someone who happens to be monetising the product too. :)
I would brag about being more-instrumental-than-thou, except that I generally spend more time obsessing over getting things right than I have spent on actually making money. LessWrong is actually a bit of a mixed bag for my business -- participation here has challenged me in lots of interesting ways, and I do end up with the odd customer or two (no pun intended), but for a long time it also had some very nasty negative effects on my ability to communicate confidently or effectively outside LW... not to mention being an attractive place to waste time.
Apologies for tempting you to spend more time here, but what difficulties have you picked up in communicating outside LW?
Mostly , a severe limitation on what I was able to contemplate saying without hearing an internal chorus of critical nitpicking, dismissive, and fully-general counterarguments. To put it another way, participation in LW greatly strengthened my mental muscles of "defense against learning anything new"... and caused me to waste considerable amounts of time trying to devise workarounds for the defenses. For communication outside LW, this is a waste of time, since most people's anti-learning defenses are not quite at that "more intelligence makes you stupid" level. (Also, non-LW persons are somewhat less likely to have brains that give them status reward pings for coming up with clever arguments for not learning anything new.) Of course, for communication on LW, it's also a waste of time. The main difference is that there are plenty of ideas I don't need to communicate here: the loss if I don't say them here is LW's, not mine. Outside LW, though, it's a different story, since I need to be able to teach those ideas to my customers, even if some of them happen to also be LW participants! So, time wasted on trying to make those communications LW-friendly is a loss for me and my customers -- and even more so if the time wasted doesn't result in any actual creative output. (Which is what usually happens when you engage self-criticism during creative output.) LW overemphasizes epistemic rationality at the expense of instrumental rationality. But the usefulness of epistemic rationality in helping people actually change is quite limited, since the function of "truth" is only useful if it convinces someone to take effective instrumental actions. (Worse still, if only a person's logical mind is convinced of the "truth", this may not have any effect on their behavior except to increase their frustration with their own behavior!) So, if you only need to use a technology (not invent or improve on it), then you don't need a "true" theory, just a good enough intuition pump. I've
Disclaimer: I am willing to play such tricks on myself in order to get results I want (e.g. - loosely related but too tired to think of a better example - allowing myself to think 'sour grapes'). I'm about to attack that part of us both. Also, I appreciate your honesty. What you're saying seems to amount to: if you want to make a sale (convince someone to act/believe), don't confuse them with the whole truth. I think this is the right way for you to operate if you want only to increase the amount of satisfied customers you have (I presume they'll be satisfied as long as they can feel that you're making them feel/think/act differently than they would have had you not sold them on new practices). But you then make it harder for those who are truth-hygienic to listen to you. Of course, you haven't confessed to anything that wasn't already apparent in your attempts to help and/or convince people here (while sometimes, like all of us, protecting your feeling of being clever and wise, sometimes in an unwise way).
Whaddya mean, "used to be"? ;-) I'm still well-known for far greater evils in the Python world (cough, setuptools, cough) than being a self-help guru, thank you very much.
Apologies. My phrasing was very poor. Being a heavy user of Python, of course I know about setuptools.
2sfb14y does something similar - it tracks program names and window titles, and which have the focus, and you can group them into different categories and assign desirability weightings to each category. You can use it personally, or as a group and then see reports on how productive you are day by day against your own productivity measures. No screenshots, though.
I'm using rescuetime as an eye of sauron. Because it logs everything i do on the computer, at the end of the day i can always check my progress and how much and where my time goes. I have set goals: 6h of work, 1h of learning a day. And all the activities basically go into 3 large groups: work, fun, learn. It works very very well. At the end of the week it feels nice to check out how much work you have or haven't done. Here is how one of the better days looks like:
I've also been experimenting with Rescuetime, with mixed results. Pros: It makes me more aware of the amount of time I spend procrastinating, and the popup it throws at me after 5 minuates of inactivity can get me back on task if my procrastination takes the form of non-computer stuff. Cons: I don't feel like I'm procrastinating less, and my efficiency graphs definitely confirm that impression.
This would be fairly trivial to implement in many different programming languages. In fact, to lessen bandwidth requirements, you could expand this idea. Set up your software to take the screenshot and automatically upload it somewhere and share the link via instant message with your watching buddy every X minutes. I'd vary the amount of time some random amount so that you didn't start gaming your own system.

Solo, I've had pretty good results with aggressive leechblock settings. My habitual timesinks are only accessible for a half-hour block each day.

Do you not have other browsers besides Firefox installed? My brain know that it can get around Leechblock by using Safari or Chrome. In fact, my brain knows it can disable Leechblock by going Tools > Add-ons and clicking on the appropriate "Disable" button. Or if Leechblock lacks a "Disable" button, by uninstalling Leechblock and installing it again later.
Here's a recipe for a bigger trivial inconvenience. If you're running Windows, you can edit the WINDOWS/SYSTEM32/DRIVERS/ETC/hosts file to include some lines like this: This way, the domains you listed will force-resolve to (a.k.a. localhost). Once done, this will introduce a pretty big trivial inconvenience between you and your favorite timesinks -- hosts is a system file, so you'll have to deal with OS warnings and prompts every time you want to edit it.
For which versions of Windows does this happen? I'm running Windows XP and I also use my hosts file to block time-wasting websites, but I don't ever get any warnings when I edit my hosts file in Notepad. This makes the hosts file not so inconvenient, because all you have to do is (Windows key) + R > type in "cmd" and hit enter, and in command line type: "edit c:\windows\system32\drivers\etc\hosts" and you're done. This can be done in about 15 seconds, unfortunately, and having the line "#" in my current hosts file is a testament to that fact ("#" comments everything out after it on that line).
Uh, I guess that my advice won't work for programmers and admins then. I'm using Windows 7 under a non-admin / restricted account with UAC set to its default setting, level 3, and I don't use command-line to edit. I do it the normal way :) -- I double-click it, it asks me for an admin password, then to chose a program to open it, I chose Notepad, edit it, it won't let me save, I save to Desktop, then close Notepad and copy the file over, it asks me to copy or replace etc. etc. BTW I just tried "edit c:\windows\system32\drivers\etc\hosts" in Win7 under both admin and non-admin accounts (had to replace 'edit' with 'notepad'). It indeed opened the file but it won't let me save the edits afterwards -- probably due to Windows File Protection.
I've actually got mine pinned on the taskbar. ;) Useful for web development... I redirect to my linux VM.
Oh, I've been doing that for years. To increase the size of the obstacle it represents, I once modified the command I am in the habit of using to open files for editing so that any request to open /etc/hosts got changed to a request to open the file containing my to-do list. But the next time it wanted to go to one of the sites on my blocklist, it took my brain about 5 seconds to realize that if it requested the opening of /:/etc/hosts (the /: being a quoting convention in Emacs) rather than /etc/hosts, it could get around my obstacle. Some small software changes have proved effective obstacles for months, but the above was not one of them. I will continue to maintain a blocklist in /etc/hosts because it causes my brain to observe my policy on work and procrastination more than it would without it, but the reason I am involved in remote-monitoring experiments is because it is a more powerful anti-akrasia technique.
Trivial inconveniences have plenty of power.
When I am off balance (e.g., sicker than usual or have suffered any sort of significant setback) trivial inconveniences are not enough to prevent my brain from procrastinating. If my brain has to really think to get around the obstacle, just the act of its getting into the problem-solving state and staying in it for half a minute has always been enough AFAICR for it to realize that it does not actually want to solve the problem. In fact, it usually stops trying to get around an obstacle to procrastination after just five or ten seconds of being in the state necessary to solve an open-ended computer problem. But if my brain can figure out how to get around the obstacle fast enough, the obstacle's days are numbered.
I explicitly uninstalled my other browsers, in point of fact. Reinstalling them is enough trouble that it's no worth it. I know that I've known about the disable-the-addon trick, but I definitely forgot about it. It'll be interesting to see if you've just sabotaged me with the reminder.
Good news! The latest version of Leechblock allows you to disable the disable and uninstall buttons!
Ouch. Sorry!

Very cool idea!

Also, don't underestimate the power of working in a library to cut distractions and add a tiny but perceptible social incentive to not slack off.

Sharing a list of running processes and DNS lookups would be more privacy-sensitive. (I have no idea how to implement the latter, but the former could at least be done on Linux, and possibly Windows, using ksysguard) You might not want to share your screen with a random stranger, but would you share process names and DNS lookups? How about open sockets?

Process names should be safe to share, since all users on a system can see them and a system is supposed to be secure for a user even with other malicious users. No sane software puts keys or passwords into stdargs where it can be seen through a ps or something. DNS lookups should also be secure in a sense since no one in their right mind would write server software where the port/address is the key, but privacy might also be a problem. (Then again, if you are browsing while doing this, maybe you deserve any consequences.)
FYI, I think on Windows you can only see processes for your user unless your user is an Admin. Windows Vista/7 makes not running as Admin more common.
Yes, well, Windows doesn't have a very good history when it comes to supporting multiple users - not nearly as good as Unix systems, anyway. So draconian measures are to be expected.
You need to be careful with what "no sane software" would do, and what real software actually does. I found a database server password today because it was in plain text in the command history.
Good ideas.

I should mention that I've been trying a vaguely similar experiment. I've been using Google Wave to document how I've been spending all my time, and I've given access to these waves to a few friends. Google Wave lets them see these logs as I type them, and add their own comments, in realtime. I was expecting this experiment to have a similar effect as the screen monitoring experiment. I was expecting that if I knew that all the time I've been wasting was documented, and that people could see this, then that would motivate me to waste less time. But it... (read more)

Catch me in gtalk and we'll arrange it.
There's a chat room going that has had around 6 LWers on it for most of the day. I don't know if screen share is an option, video share is.

This sounds really cool! I would definitely like to do this with someone the next time I am in an entrepreneurial setting and/or working from home.

Alternatively, you can get involved now to help this practice grow. Specifically, if you help me with my akrasia now (by monitoring me), I will keep track of the number of hours you helped me, and agree to repay you an equal number of monitoring hours in the future.
Maybe! I'm on a Mac right now and not willing to run Windows; not sure to what extent you can accommodate that. Also, to be clear, I'm not willing to participate in a monitoring exercise while I'm at work, commuting, sleeping, or exercising -- you might find that it takes more effort to figure out when I can monitor you than it would to just find another partner.
I could use your help even if only for a few hours a week. I am very bad at resisting the temptation of going to Hulu Dot Com and mindlessly watching TV shows. How many hours per week will you be working? I am on a Mac right now, too. Nerdy details follow. I can certainly guide you in setting up a view onto my screen or "desktop", but do not yet know how to make that connection "view-only": the app that I suggest you use (which comes with your Mac and is called "Screen Sharing") is intended to be used to operate a computer remotely, so when your "Screen Sharing" app is in the foreground, not only can you see my screen, but also the keystrokes or mouse clicks you make are transmitted to my computer rather than applied to your computer. What that means in practice is that you will occasionally forget that your clicks and keystrokes are being transmitted to my computer, and fleeting harmless hilarious chaos will sometimes ensue at my end.
Mac OS screen sharing is compatible with VNC, so the viewer can use any VNC client with a view-only mode (e.g. Chicken of the VNC (I think) or TightVNC from MacPorts). (Screen Sharing's superior proprietary compression will be lost, though.)
FWIW, we couldn't get OS X screen sharing to work with RealVNC on Windows, but installing Vine Server on OS X fixed all our problems.
Sadly, that is no longer true of the version of OS X released after that was written. Specifically, neither Lion's built-in "screen sharing" server nor Vine Server on Lion works with any of the Windows VNC clients we tried. When someone used Windows remotely to monitor my OS X Lion desktop in 2012, it was through "captures" (still images) of my screen saved to a shared folder of Dropbox. Specifically, I wrote some code to fork and exec "/usr/sbin/screencapture -C" every 3 minutes.

I should add that both Vladimir and I understand that most people do not want to participate in remote monitoring because of the loss of privacy. Also, if I ever do write software to make remote monitoring more efficient, I will take great pains so that the software is not used to monitor people without their consent. That would probably mean assigning copyright in the software to an individual or a group with a reputation for humaneness and not releasing it under and open-source license.

I don't think you need to be so concerned about your open source anti-procrastination software being adapted to nefarious uses. That horse has been out of the stable for a long time.

To clarify: you want to oppose people using your software to illegally spy on people by making it illegal for them to use it? Shouldn't they be indifferent to such limitations? Isn't spying software a commodity anyway, with your software not optimized for the task?
I should have written in grandparent that I realize that some applications of the technique will tend to cause more harm than good, and left it at that. I just did not want to come across as clueless or uncaring about the harms. It is not in general illegal in the U.S. for employers to monitor their employers, and your comment made me realize that I should see if I can apply monitoring software marketed to employers before writing my own software.
Almost certainly. I know that such software is popular on things like oDesk, etc.
It occurs to me that you could get a lot of the lost privacy back by blurring the screen enough to make text illegible. As long as the other person can still recognize known procrastination activities in their blurred form (which should be pretty easy - sites have distinctive color schemes), it still works, but there's less risk of accidentally revealing a private conversation or other secret. I wouldn't worry too much about your software being useful for spying on people without their consent. The standard countermeasure is to put a notice in a little window or on the system tray while it's active. People who want to spy on others' computers already have lots of software to choose from, both repurposed and purpose-built, so you'd have to go out of your way adding concealment features to make it competitive for that purpose. (That said, do be conscious of security: require a password to view, and use a vetted third party library to encrypt the login and the images.)

Your anti-akrasia system is driven by shame? In designing an ideal solution, this wouldn't be one of the must-have features I would add into it.

Any questions are welcome.

Have you considered understanding why you procrastinate and then eliminating that reason?

To quote EY from another context:

"""Explanations are supposed to make you less confused. If you feel like you don't understand something, this indicates a problem - either with you, or your teacher - but at any rate a problem; and you should move to resolve the problem. ""... (read more)


Your anti-akrasia system is driven by shame? In designing an ideal solution, this wouldn't be one of the must-have features I would add into it.

Any questions are welcome.

Have you considered understanding why you procrastinate and then eliminating that reason?

To quote EY from another context:

"""Explanations are supposed to make you less confused. If you feel like you don't understand something, this indicates a problem - either with you, or your teacher - but at any rate a problem; and you should move to resolve the problem. ""... (read more)

Ignore this.

(Test of the RSS feed of this post.)