Edit: I've made a top-level post for sharing anti-akrasia techniques -- go ahead and share your techniques as well. Let's continue the discussion there.
A very quick outline (I'll post a detailed version later).
Determine what is your current better judgment. This is critical -- I noticed that I hesitate to trick myself into doing anything I don't consider to be relevant to my goal.
Constant asking myself: "is the activity I'm doing at the moment advancing me toward the desired state of reality"? If the answer is "no", know that I'm procrastinating.
80/20 elimination, Tim Ferris / Pareto style (I'm skeptical about the rest of Tim's book, but the Elimination chapter is pure gold).
Parkinson's law (work expands to fill the time allotted). Again, Tim has some advice on it -- basically, it boils down to scheduling most important things (in the 80/20 sense) first, with aggressive deadlines.
PJ Eby's secret meaning of "just do it". He considers the article to be outdated, but its key paragraph worked wonders for me. Basically, "just do it" = "don't do anything else". In its pure form, "not doing anything else" is too macho for me, so I leave a line of retreat for myself -- I permit myself to eat, think about anything (not just the task), walk, have sex, but no surfing unless it's on-topic, no doodling on paper unless it's on-topic, etc.
Self-priming -- I try to expose myself to stimuli related to my current task, and to shield myself from irrelevant stimuli, no matter how pleasant (e.g. I run away from my toddler daughter, because prolonged exposure to cuteness tends to totally ruin my ability to work efficiently :)
Begin. For example, if you need to do some stuff in Excel, just open an empty spreadsheet and type in the table header. Just stupidly staring at this makes you better at spreadsheets and your task -- your mind pulls linked concepts to the fast-access cache, without your consent, and you don't need to do anything.
Mindless repetition of things like "I want to do/make/design/create/code the best X in the world" (X stands for a short-term todo item, not some far-off goal) -- I've been doing this for at least a decade, and it seems to work (don't know how, but perhaps it's related to Cached Selves).
Sorry if I forgot something -- gotta run (will edit later.)
(Obligatory note of caution -- these are tricks, they are not proven by experimental results, I've been using them for just about 3 months, they aren't tested under serious stress / pressure, I have no deep theories to explain why they work, and the evidence I offer is purely anecdotal.)
The subsumption architecture for robotics invented by Rodney Brooks is based on the idea of connecting behavior to perception more directly, with fewer layers of processing and ideally no central processing at all. Its success, e.g. the Roomba, stands as proof that something akin to control theory can be used to generate complex agent-like behavior in the real world. In this post I'll try to give some convincing examples from literature and discuss a possible application to anti-akrasia.
We begin with Braitenberg vehicles. Imagine a dark flat surface with lamps here and there. Further imagine a four-wheeled kart with two light sensors at the front (left and right) and two independent motors connected to the rear wheels. Now connect the left light sensor directly to the right motor and vice versa. The resulting vehicle will seek out lamps and ram them at high speed. If you connect each sensor to the motor on its own side instead, the vehicle will run away from lamps, find a dark spot and rest there. If you use inverted (inhibitory) connectors from light sensors to motors, you get a car that finds lamps, approaches them and stops as if praying to the light.
Fast forward to a real world robot [PDF] built by Brooks and his team. The robot's goal is to navigate office space and gather soda cans. A wheeled base and a jointed hand with two fingers for grabbing. Let's focus on the grabbing task. You'd think the robot's computer should navigate the hand to what's recognized as a soda can and send out a grab instruction to fingers? Wrong. Hand navigation is implemented as totally separate from grabbing. In fact, grabbing is a dumb reflex triggered whenever something crosses an infrared beam between the fingers. The design constraint of separated control paths for different behaviors has given us an unexpected bonus: a human can hand a soda can to the robot which will grab it just fine. If you've ever interacted with toddlers, you know they work much the same way.
A recurrent theme in those designs is coordinating an agent's actions through the state of the world rather than an internal representation - in the words of Brooks, "using the world as its own model". This approach doesn't solve all problems - sometimes you do need to shut up and compute - but it goes surprisingly far, and biological evolution seems to have used it quite a lot: for example a moth spirals into the flame because it's trying to maintain a constant angle to the light direction, which works well for navigation when the light source is the moon.
Surprising insights arise when you start applying those ideas to yourself. I often take the metro from home to work and back. As a result I have two distinct visual recollections of each station along the way, corresponding to two directions of travel. (People who commute by car could relate to the same experience with visual images of the road.) Those visual recollections have formed associations to behavior that bypass the rational brain: if I'm feeling absent-minded, just facing the wrong direction can take me across the city in no time.
Now the takeaway related to akrasia that I've been testing for the last few days with encouraging results. Viewing your brain as a complete computer that you ought to modify from inside is an unnecessarily hard approach. Your brain plus your surroundings is the computer. A one-time act of changing your surroundings, physically going somewhere or rearranging stuff, does influence your behavior a lot - even if it shouldn't. Turn your head, change what you see, and you'll change yourself.