People don't get smarter when the stakes are higher. And the same brain that governs your food choices, and your media consumption habits, and your work ethic, also governs the large choices that you make; where to move, who to spend time with, what to spend your time doing. These two facts together should give any self-described rationalist tremendous cause for concern. 

It seems that many people here treat akrasia with the same sort of disdain one one might treat keeping their room clean- the "I could do it if I wanted to, but I have better things to do." However, if you look at the stated goal of rationality; "doing the thing that is the right thing to do" and the definition of akrasia "not doing the thing you know is right" then red flags may begin to pop up. 

The realization I have had that has generated the most worry in me is this- In general, the big, important choices we make have open ended results. I may decide that moving to a new city is what is best for me, after much consideration and thought, but I can never know with any certainty whether I was right. The decisions we make on a day to day basis, however, often do have clear cut inputs, outputs, and often mostly established answers; deciding what to eat, when to do work, how much sleep to get; and any amount of akrasia suggests that we're getting a lot of those choices wrong. 

Say you're in a math class, for instance, with homework assignments (whose answers you can check) and a final (whose grade is hidden). If you're curious how you did on the final, your best guess might be to look at your homework scores- to see how well you were able to implement the things you learned in small problems, with simple answers, before you had to apply them to the big, open-ended questions. 

In the end, the brain that you have is the same- and the brain that tells you to get ice cream before bed is the same brain that tells you where to live. If you're deciding to move to a different city, and that different city has an ice cream parlor you're particularly fond of- if you're not able to control yourself when you're thinking what to eat, that same part of your brain will still be screaming 'ice cream' when you have to think about the cities. And then your metric for city moving will be, at least in some small part, 'which city has better ice cream.' 

Now, you might say, "Aha! That doesn't apply to me, in the same way that I don't procrastinate close to a deadline! When it really matters, I'm able to think rationally!" To this I say, "Are you kidding?!" It is true, somewhat true, that we put more thought into problems we care more about. But more thought is not better thought, and if the source of those thoughts thinks that ice cream is the thing-to-do, then the source of those thoughts is corrupted. 

Because really, akrasia as stated doesn't exist. You can never do a thing that you don't think is the thing-to-do. We can only ever to what we really think is the thing-to-do. If you're procrastinating on work, you HAVE justified it- either with promises of future productivity, or by balancing your internal wants, or by thinking yourself unable to. If you indulge in ice cream, you do think that eating ice cream is the thing-to-do; perhaps you have justified it to yourself with the idea that it is okay to indulge occasionally, or that you'd rather save your willpower for something else, or that some sugar isn't too bad for you- but you have justified it, and as a whole, the brain-that-controls-the-body thinks it the thing-to-do. 

So what do we mean, then, when we say akrasia? To me it is the screaming in pain of the mind, the last call out from rational thought as it dies unheard, the surrender from intellect to base instinct and emotion. Often enough, or akrasia comes from emotional bottom lining; first (Tired, fatigued, bored) -> "I don't want to work right now", and then "I'll be more effective if I do this later," and then the act of procrastinating itself. If we were thinking rationally, we would ask "When will I be most effective doing this work?" and come to perhaps a different answer- or our original! But once the question is genuinely asked and answered, it is no longer akrasia; it is the guessed thing-to-do. 

However, this path out is incredibly dangerous. If you think to yourself "Well here's my solution- I just have to ask that question whenever I feel like procrastinating" then your chain of thought will look like this: (Tired, fatigued, bored) -> "When will I be most effective doing this work?" -> "I don't want to work right now", and then "I'll be more effective if I do this later, and since I asked the question, it's okay to not work right now," and then the act of procrastinating itself. We haven't solved anything! Rather, there are two real solutions; the question must be asked when there is no prospect of work, or evaluated post hoc. 

To ask in advance, consider when you are not hungry what a good thing to eat is. I do a lot of dance, and afterwards I rarely have anything to do before I go to bed. I know the main problem with eating ice cream that I have is that for an hour or so after I eat it, I struggle to get much work done, and want to take a nap. This is not a problem if my plan is to go to bed an hour or two later. Plus, I really enjoy eating ice cream after dance. So, I conclude (far from my wants) that eating ice cream after dance practice is a thing-to-do. 

Similarly, I might consider at the start of the day what the most effective time to do project work is- when the work is not imminent, and I can think unimpeded. Why this makes the difference? Well, I have a guess or two. These I shall share later.

But say that something really does come up! I'm about to sit down to do my work, when my terminally ill mother calls, and wants to talk. Am I not able to then put my work off? Is this akrasia? Well, in the moment there's no way to know. A terminally ill mother seems a decent excuse, but with your vision clouded by present-ness, all you can do is act on instinct. However, the next day, your vision is clear- you can evaluate how well you did, and if the decision you made in fact was the thing-to-do. And if you make the right choice enough, you may learn to trust yourself in the moment, when it really does seem that you would be more effective later. However, this requires some serious self reflection, and my guess would be that very few of us are nearly so good at guessing as we act on. After all, how often do you think to procrastinate, and then say- "well, actually, I'm pretty bad at guessing when I'm going to be most effective, so I'll just do my work now."