The Truth About GTD Software Tools
- Thread starter Dave Edwards
- Start date
I would argue what GTD deserves is to never have this level of automation, because it loses the human at the middle of it -- the human for whom this entire process exists to help in the first place.
To go back to the original topic, I believe, Tom, that you didn't quite get what he was saying in the video. As others have tried to clarify, GTD is not something that can be codified the same for every individual, so a specific piece of software can't exist for it. Not fully. At most, they'll be a list manager that prioritizes organizing things in ways GTD suggests organizing them, but that's not GTD software, that's just GTD-aligned list management.
By "holistic" he means the definition of the word: it's a multifaceted system, of which software can only ever be a part and not the whole. Why? Because GTD relies on you, the person, to critically analyze your lists, your projects, your calendar, and make decisions based on what you see and what you feel. Software just *cannot* do that.
So no, GTD cannot be wholly implemented on paper, because like software, paper is just data management. List management, reference material management, calendar management. But paper, like software, also cannot make decisions for you.
However, this doesn't mean it's unimplementable. It just means one tool cannot do 100% of the GTD process for you. Just like a robot could choose the ideal food for you to eat at every meal, when to eat it based on your calories burned and body type, and even shove it down your throat, it cannot digest it for you. *You* have to do that.
GTD needs the human (and yes, the human, not an LLM or AI or whatever other "live my life for me" tech that's being invented is out there now or in the future) interacting with the data and making decisions at the center of it to work. And software cannot do that for you.
"But you can write software to analyze your lists and projects and calendar and have it prompt you what to do next!" Sure, yeah, you can. But only with the data you put into it. And, realistically, how many of us here keep absolutely perfect systems, all the time? How many of us realistically capture every tiny point of data that contributes to making a decision? I think I can confidently say zero, and say anyone saying they do is lying to themselves. We are all human and make mistakes.
But also -- not all data is worth the effort it takes to input into the system. There are some things you'll never forget, or at least shouldn't, that are just wastes of time to plug into software just so it can do the human element of the work for you. Let's do a small example -- say you have this fancy software that does, theoretically, make all of your decisions on what to work on next. And it tells you to call John Smith to get input for a project. What number do you call? His cell phone? His desk phone? Is he in a meeting? If he is, then it's clearly not the best next action for you to tackle as he's unavailable. But in order to know that, it needs access to his calendar. Okay, say it has that. His calendar needs to be accurate. Okay, say that it is. But maybe you just passed him by the coffee pot and he was having a quick chat with someone 2 minutes ago. No one puts that on their calendar, but the tool can't know that. You know there's a good chance he's not at his phone to answer it. So now the software needs, what facial recognition software and security camera access? You see how crazy this gets, very quickly.
You, as a human going over your lists, can instinctively just parse this and pick another task to do, or quickly decide that off-topic convo at the coffee pot could be interrupted with work questions, hop up and do it. But if you rely on a GTD tool to do this decision-making process for you, it's going to stall out all the time during these specific things that crop up, constantly. Building a tool that can handle the decision-making process can only go so far because it can only ever know so much.
Your tool cannot live your life for you, so it cannot make your critical, in-the-moment decisions for you, either. Only you can -- and I would argue, should -- do that. GTD is a holistic system based on a human being in the middle of it. And software just can't ever replace you, because if it did -- why would it need GTD in the first place?
Hope this makes sense.
Tom_Hagen
Registered
So, back to the original question: please list the elements that can't be codified in GTD for individual users. I keep reading generalities and saying it can't be implemented.
The argument that GTD requires individual commitment and therefore can't be implemented is an argument that a word processor can't be written— because, after all, it's humans who generate content. Absurd.
It's not true that software can't make decisions. It can, and it does. This is precisely the advantage of software over paper: that paper is static by nature, while software can have built-in mechanisms that allow for the design of information flows according to the user's preferences.
I completely fail to understand the argument that we can't program GTD because who among us maintains a perfect system? So what? Word or Writer couldn't have been created with that assumption. Do we write perfect texts from the start?
The phone number example is also flawed. Firstly, I see that most people here expect a GTD program with AI elements, but that's not the point. Secondly, if you don't have a current phone number, it's not a software problem, it's not a notebook or a piece of paper problem. It just means you don't have the necessary information. It has nothing to do with the program. But if we're going to fantasize, at least the program could display the number or confirm its absence and ask if it's current. And then suggest the next action, something like: Get John Smith's number. But these are details, and even if they weren't, I don't think it's necessary to call it a successful GTD implementation.
The fundamental mistake: Yours, and many others before you, have some strange assumption that a GTD program should think for us, make every decision, and live for us. I don't think Allan meant that, and I certainly don't.
Sorry, but this is another message along the lines of: it's impossible because... human... because it's complicated... because who knows what else.
I believe at this point you're either intentionally misreading what people are saying and approaching this conversation in bad faith, or are unfamiliar enough with what GTD actually is that a re-read of the book is in order before continuing the discussion. Everyone in this chat has been abundantly clear, I think, in answering your question. And my entire comment was listing specific ways in which software cannot handle specific stages of the GTD method. I'm not sure how to write it more clearly for you, but I'll give it one more effort.
Alright. I truly believe we have to go back to the absolute basics of GTD to get the point we've all been trying to make to you across.
Let's revisit The 5 Stages of GTD:
STEP 1
CAPTURE
Collect what has your attention
Write, record, or gather any and everything that has your attention into a collection tool.
Some capturing can be automated by software through detection of key words, phrasing, recognition of questions, etc., but a lot of it simply requires human input. It requires you to type something in an inbox somewhere, to write on a slip of paper, to dictate to your phone or device. This is just reality. Unless you suggest that 100% of these situations can be automated somehow? Remember, we're living in reality, so things like personal privacy, recording laws, etc. exist, as well as situations where a human does as humans do and doesn't have whatever thing that automatically captures stuff for them on their person at a moment where capturing is necessary.
STEP 2
CLARIFY
Process what it means
Is it actionable? If so, decide the next action and project (if more than one action is required). If not, decide if it is trash, reference, or something to put on hold.
Theoretically, software intelligent enough could detect a good chunk of what, on a capture list, contains an actionable item, but again, we live in reality. Do you really, truly think it's realistic that software is going to know every microscopic detail a human would use to make an accurate decision on clarifying whether an item is actionable, and if so if it's a project, or if not, if it's trash or reference material and to then organize it properly? This is why I gave the phone number example. Which, by the way, your response shows you misunderstood. I never suggested you didn't have a phone number for the person. Kindly read that section of my previous comment again to understand it. The example was to show how many small bits of information are needed to appropriately decide a next action, and how easy it is for the software to not have access to that information. Again, in reality.
STEP 3
ORGANIZE
Put it where it belongs
Park reminders of your categorized content in appropriate places.
Theoretically, software could do a fair chunk of this. However, there is no tool on the market that comes even close to being able to accomplish it, so going back to the video and David's comments, asking him to endorse something that doesn't exist is nonsensical. Right now, this step needs to be done by a human. You can automate chunks of the process, such as putting the filing behind keyboard shortcuts or using a variety of automated tools to flag certain things, but deciding where it should go, hitting the keyboard shortcut for the proper filing location, or constantly updating the triggers to be detected for automated filing requires a person. But again, there is no tool on the planet that advertises being able to do anything even remotely close to this for everything, so we've left reality at this point.
STEP 4
REFLECT
Review frequently
Update and review all pertinent system contents to regain control and focus.
This step is the one software can do the most of, I think. This can have rules clearly written for software to follow through almost the entire process. Software could read every project, and look inside each one, and detect when something was last engaged with (but again, engagement requires a human), to suggest things that need another look. But the big benefit of the REFLECT stage is FOR the person utilizing GTD -- it reminds your brain, specifically, of your commitments, so you know what the heck you committed to doing and puts them in your memory. If there's no human involved in this process, you might as well just skip the step entirely.
STEP 5
ENGAGE
Simply do
Use your trusted system to make action decisions with confidence and clarity.
This step 100% requires a human to do. Period. The entire reason GTD exists is to make this step successful for the human responsible for doing it. If software were to decide what to do and do it for you, you should be fired, no? You're not necessary. This theoretical software can just have your job.
Action by a human is a core part of the definition of GTD, which fundamentally means GTD cannot be entirely automated via software. GTD only exists to enable action by a human, and said action is part of the GTD process.
Perhaps the point isn’t that there are elements that can’t be codified, but rather would the investment in developing such a system bring any particular benefit?
Clearly people have got a lot of value from GTD, and used a variety of analogue and digital systems to realise that. I obviously can’t speak for David Allen, but I expect he would only progress this if it was a viable product, and I’m not (and assume he isn’t) convinced it is.
I expect you might dismiss this as you have other similar comments, but as I’ve engaged with GTD over a long period, I also feel that intention and action are really at the heart of the system, more than the sum of various decisions/steps.
I’d be interested to know why you’re not developing this app?
Tom_Hagen
Registered
I think I know Allen's GTD by heart.
This is the source of the misunderstanding. In my opinion, it's not me who doesn't understand the problem. Why do you keep focusing on automation? If you expect some device to constantly observe you, listen, and even read your thoughts through implants in your brain, so that all your needs and ideas can be dumped into your Inbox, then the answer is: "no, something like that can't be created today."
However, I repeat for the umpteenth time: no one has abandoned writing a word processor just because a human has to enter text into it. I wrote this to you recently, but you ignore my comments and wander into some AI-style automation.
I won't comment on the rest because the same problem recurs: automation, doing everything for you, including decision-making. I really don't understand this focus on that.
A GTD application should implement what's in Allen's book, namely support for structures and processes. If I remember Allen's presentation about his ideal GTD program, it included a dashboard with summary information, etc.
Software is a support, not GTD itself. Just as a word processor is not text. It's obvious that as a human you have to make decisions; it's obvious that as a human you know how much energy you have, where you are, and how much time you have. Software can support you, but it can't replace you. And here, everyone is focused on doing the work for you.
I don't think that's what Allen had in mind.
Tom_Hagen
Registered
Okay, I can agree with that. But then you should say, "It's not worth it," not "It can't be done."
To be honest, I wrote this application for my own needs. Writing it for a wider audience, however, is a completely different challenge, as it involves greater responsibility for data storage, backup, application development, bug fixes, etc., etc. And of course, I don't have time for that.
But: yes, I wrote this application, so I'm writing from a practical perspective, not from a theoretical one.
Let me put it bluntly: implementing GTD is one of the simplest projects (I'm also writing from experience). There's really nothing complicated here.
Writing an efficient compiler, database server, or antivirus heuristics might be complicated, but this is a simple, if not trivial, project.
But who is saying that? I didn't pick up where you got this from in the original video? At 1:45 David actually says the technology probably does exist, but not the market.
I understood David’s ‘holistic’ point differently than you did - I don’t think he is saying that the system can’t be created because it is holistic, but rather his vision of a holistic GTD app would cover too many other existing apps that people already use and would likely prefer. I think this is a very practical observation - I have used apps in the past that have tried to do too much and lacked in certain areas. If I have a preference for a particular list management app and a calendar app, it would be a tall order for someone to develop one app that beat both - even if they did, the chances it would then also suit someone else is slim.
You said in your original reply “claiming there's no market for a given product without trying to sell it is pure guesswork” - do you really believe that the best way to approach a new product is to fully develop and deploy it and see what happens, rather than make a call that it might not do well?
I don’t think David is claiming that it is impossible, as you said - but I would also be surprised if anyone ever does, so potentially the same end result…
Tom_Hagen
Registered
I don't buy it. Why does Allen need PowerPoint in GTD? Why does Allen even need all the programs integrated with GTD? If Allen writes about multiple inboxes (digital and physical) in his book, especially the latter, he must assume there will be supporting tools for the GTD application itself. It's a bit like a pen and a calendar. I really don't buy that logic.
I'd also like to point out that Allen mentions designing something that no one has implemented.
The dashboard he mentions is trivial, and I had something similar in my own application.
Indeed, Allen mentions the market. But let's note that his reservations aren't so much about the technology as about people who "still have their ideas in their heads." This has nothing to do with the software. Moreover, there are paid GTD-like solutions on the market that are doing very well.
Also, note that Allen mentions two attempts – one took place in 1995. Perhaps I should relate this to the history of AI.
If you trace the history of AI, it experienced what are known as two winters – due to hardware limitations. Today, as memory has become cheaper and processor performance has increased – as you can see, we have a flood of these solutions.
Sorry, but if your reference point for this not being possible is 1995...
fooddude
Registered
Yup.
I think at the end of the day, David Allen has designed GTD as a system, and the book is a practical guide in how to implement that system, regardless of specific tools.
The fact that you have developed a personal system that incorporates some of the ideas David mentions is great, but I don't think anyone is doubting that or are surprised - I expect pretty much everyone who participates in a forum on GTD has had some success in developing such a system, although they may look very different.
I think that the disconnect here is that the video isn't talking to someone's ability to create a system based on GTD - I'd say that's the entire point of the book, and it has been extremely successful. What they are discussing is David's consideration on the prospect of developing and rubber-stamping what would need to be THE official GTD app - if I put myself in David's shoes, I don't think I'd want to progress it either.
fooddude
Registered
And how complex or simple do you make it? Just the diversity of implementation of those on the forum here is significant. I think a GTD specific tool would hurt the brand rather than help it. And you know if people want convenience, as soon as a feature is not convenient, they will stop using app.
A word processor is just a singular piece of software. It is not a holistic method for tackling and following through on commitments. Your comparison is apples to oranges in this discussion.
The point many of us are trying to make here is that half of GTD, because it is a *system*, is habits and behavior. Things a person has to do. And software *cannot* do that. GTD isn't the Dewey Decimal System. It's not just a method for making lists. It requires behavior, and software can't do that.
But even excluding all of this, he's very clear in the video. You can't have one GTD tool do everything because in reality, you can't have one tool solve all of your problems. Notion comes kind of close, but then when you need to collaborate, you need everyone else to use Notion. You need to move all of your stuff off other platforms and into Notion. Everyone *else* needs to move all of their stuff too. You really think people are going to move off Office (sorry, CoPilot 365 or whatever it's called now) or Google to migrate en masse to one product? And then Notion has to stay in business. Frankly it's just bad policy to use one platform for everything; when it sinks, it takes everything down with it.
And it's just not reality that this will ever happen. Other people will always use other tools, other things will always come up, and having one tool be able to tie every possible situation together is just impossible to implement. Even when things like Zapier exist they can't handle all situations (or even most).
By human nature, I suppose that when dealing with repetitive flows, repetitive tasks, be GTD or whatever in life, humans will seek for improvements, automations… But that’s probably my lean 6 sigma bias, when I see defect production along a flow, I look at the flow and modify it to eradicate defects. As simple as that.
I watched David's video. It really isn't about a GTD app.
It's about designing an excellent human-computer interface for people who are using the GTD productivity methodology.
For example, on a Windows computer David's "GTD App" might actually be a replacement for the Windows Shell or at least customized virtual desktops where a GTD adherent would follow the GTD methodology.
It is interesting that David's ideal interface is keyboard rather than touchscreen or mouse oriented. Much more efficient. You don't see a virtuoso musician clicking on piano key images on a screen.
Of course, I'd love to see a video of David standing in his huge Minority Report style holographic office, grabbing objects from a giant in basket, yelling "IS IT ACTIONABLE?" and hurling it into the monster flaming incinerator trash bin.
It's about designing an excellent human-computer interface for people who are using the GTD productivity methodology.
For example, on a Windows computer David's "GTD App" might actually be a replacement for the Windows Shell or at least customized virtual desktops where a GTD adherent would follow the GTD methodology.
It is interesting that David's ideal interface is keyboard rather than touchscreen or mouse oriented. Much more efficient. You don't see a virtuoso musician clicking on piano key images on a screen.
Of course, I'd love to see a video of David standing in his huge Minority Report style holographic office, grabbing objects from a giant in basket, yelling "IS IT ACTIONABLE?" and hurling it into the monster flaming incinerator trash bin.
Tom_Hagen
Registered
I really don't understand this approach, where software should do everything for the user. Weeding the flower beds too?
And I don't understand why everyone around us should use the same GTD tool. So, none of us use GTD because there are people we work with who don't?
The arguments are getting more and more bizarre.
Tom_Hagen
Registered
I also think he placed a lot of emphasis on the interface, hence the dashboard. However, it would be good to know which element of such an interface is "impossible."
TruthWK
Registered
So I've seen David talk about this before and watched the linked video just now. It seems in this video, he is saying that the minimum viable product wouldn't have a market, and that having an app that could integrate all the apps we use would be required to get a complete system to eliminate any friction. He also mentioned trying it in the 90s and again in 2013. He also mentions the difference between list managers and a GTD system, that handles workflows with supporting triggers. I think FacileThings and OmniFocus are probably the closest to what David Allen describes. FacileThings, especially, tries to integrate the triggers that he mentions and have areas that mimic the workflow process not just giving you a place to organize things according to GTD.
The statement to me seems like a mix of perfectionism by requiring a complete system and wanting a market for a minimum viable product. I would think if you just came up with something that is a measure better than FacileThings or OmniFocus with a decent UI, you could get some buy in for an MVP. Especially now in the age of vibe coding, although in fairness, this statement was made before LLMs came to prominence I believe, at least originally.
Going from other videos i've seen where David talks in more detail about an ideal app, his main point always seemed to be incorporating the workflow piece not just the organization piece. He says in this video multiple times that the tech exists to implement the workflow piece, and has for awhile. He just said most list managers and other software used for GTD does not choose to implement the workflow piece. What I observe is that they try to be method agnostic or at least give too much flexibility. If they don't, a portion of their user base will revolt. Maybe the GTD community is too small in David's mind to demand a very strict GTD implementation that can include the workflows in a friction free way. Once again, I would say that FacileThings is probably the most strict implementation of GTD with the workflows included to a much larger extent than other apps. It does fail some in the UI as it is a very small team.
The other part relates to different people having different workflows in their mind. Here, you would just need to get a broad enough sample of different styles to give options and a low friction onboarding and switching process that would make it easy to try a few to see what gels with the user.
The statement to me seems like a mix of perfectionism by requiring a complete system and wanting a market for a minimum viable product. I would think if you just came up with something that is a measure better than FacileThings or OmniFocus with a decent UI, you could get some buy in for an MVP. Especially now in the age of vibe coding, although in fairness, this statement was made before LLMs came to prominence I believe, at least originally.
Going from other videos i've seen where David talks in more detail about an ideal app, his main point always seemed to be incorporating the workflow piece not just the organization piece. He says in this video multiple times that the tech exists to implement the workflow piece, and has for awhile. He just said most list managers and other software used for GTD does not choose to implement the workflow piece. What I observe is that they try to be method agnostic or at least give too much flexibility. If they don't, a portion of their user base will revolt. Maybe the GTD community is too small in David's mind to demand a very strict GTD implementation that can include the workflows in a friction free way. Once again, I would say that FacileThings is probably the most strict implementation of GTD with the workflows included to a much larger extent than other apps. It does fail some in the UI as it is a very small team.
The other part relates to different people having different workflows in their mind. Here, you would just need to get a broad enough sample of different styles to give options and a low friction onboarding and switching process that would make it easy to try a few to see what gels with the user.
GTDgeekette
Registered
The owner of the YouTube David Allen interview video clip posted above invited email inquiries that he could relay to DA himself, at the end of the video: daveedwards@outlook.com Perhaps you you could ask DA for more detail on this? Or simply submit the question to him directly through his business email (though it will likely get a better response through an intermediary like this guy Dave Edwards, who has a professional connection to DA).