While AI has become a buzzword that has lost much meaning, there are still many use cases that can help with our ADHD. One of the places that I have been enjoying exploring is how I can outsource some of my executive functions to these applications so that I can save my brain for the important stuff.
In today’s episode, we’ll first examine executive function so that we know what we’re really working with here. Then, we’ll explore a number of tools that we can use—from large language models (LLMs) to tools to help with your calendar, and even smaller tools fit for goblins.
Also, before we get into anything about AI, I do want to make a quick disclaimer about some of the ethics of using these tools and also how they were created. While these tools can be great in assisting with what we create, it’s also important that we’re still doing our own work and not passing off what LLM did as our own work. I have been using some of these tools to help create episodes, and even cite some of the ways I am doing so in this episode. But what I’m not doing is feeding a prompt into an LLM and just using that. That would be lazy, irresponsible and not something that anyone wants. However, just because I’m not doing that doesn’t mean that other people aren’t—this is something to particularly look out for with new books about ADHD that aren’t coming from established authors and creators.
Additionally, there are ethical concerns about how the training data for many of the LLMs was obtained. This is especially a concern in regard to the lack of content used to source this material. This has been seen more in terms of art and how those models are trained without the artist's knowledge, but it also applies to written word. While I still have been using these systems, I think it is important that we encourage more ethical sourcing of training data and only using data with consent.
To get started, we need to ensure that we understand exactly what we’re talking about when we refer to executive functions.
Simply put, these are the mental skills that help us plan, manage ourselves, and respond to situations. They’re the brain's control center, guiding how we handle everything from organizing a task to reacting in the moment. You can think of them as the tools we rely on when we’re not just coasting on autopilot.
Dr. Russell Barkley, a leading expert on ADHD, explains that executive functions are closely tied to self-regulation—the ability to manage our behavior, emotions, and thoughts in pursuit of long-term goals. In fact, Barkley argues that ADHD is fundamentally a disorder of self-regulation rooted deeply in deficits in executive functions.
This is an important distinction we can work from because, unfortunately, there is no consensus on what an executive function actually is. I know; given how ubiquitous the term has become, this seems wild to me.
But if we can see that difficulties with executive function are one and the same as difficulties with self-regulation, we have much firmer ground on which to stand. That said, we can still use some of the language of executive function as we move forward.
We typically draw from three core executive functions: working memory, cognitive flexibility, and inhibitory control. These mental processes help us retain information, adapt to new situations, and resist impulses.
Working memory is our very short-term memory—the stuff we’re actively holding in our minds while we work with it. It’s supposed to help us keep track of information while we’re doing something else, like following a recipe step-by-step. For many of us with ADHD, working memory is one of the hardest things that we deal with. We start cooking that recipe we’re reading, glance down at the first step, and then walk over to grab the next ingredient, and somehow, we have no idea what we are supposed to be looking for even though it was clear in our mind seconds ago. That’s working memory in action—or perhaps more accurately here, inaction.
Cognitive flexibility can mean creative thinking and thinking outside the box, but it also refers to our ability to adapt when things don’t go as planned. Sure, a lot of us with ADHD excel in creativity, but dealing with the unexpected? Well, I know personally that can depend on the severity of the situation—getting a couple of unplanned free hours, well, I guess I’m not doing anything. But if things really go wrong, I might be your man.
Inhibitory control is about resisting temptations and not acting impulsively. It’s what helps us maintain focus on what’s important and block out distractions—like ignoring the ticking of a clock during a test or staying engaged in a conversation instead of focusing on the TV on the other side of the restaurant.
Executive functions are not just about managing tasks at the moment; they’re crucial for long-term success and goal attainment. When I’m trying to work on a new episode and have to resist other distractions, like playing video games, it’s my executive functions that help me resist the urge to indulge, reminding me of my longer-term goals. This self-regulation involves a series of mental actions—like refocusing, reminding myself why we shouldn’t give in, and visualizing our future success. For someone with ADHD, these self-directed actions are often weakened or delayed, making it much harder to stay on track—and this is also going to be key when we get into these tools that can help us with this self-regulation.
Deficits in executive function can have a major impact on our ability to self-regulate. This is why things like managing time and thinking about the future can be so tough. It makes it hard to plan ahead, anticipate what’s coming, and stay on track with our goals.
It’s also important to note that executive function deficits aren’t exclusive to ADHD, and in fact, executive dysfunction isn’t part of the ADHD diagnostic criteria in the DSM-5. However, there’s such a significant overlap that it’s hard to imagine an ADHD diagnosis without at least some executive function challenges. So, while it’s not an official part of the diagnosis, I believe it’s crucial to consider when we’re trying to manage our ADHD.
All that just to get to what we’re talking about in terms of executive function, but I think it’s an important starting point because when we’re looking for tools we want to think about what they are solving for. It’s easy to see the shiny new tool and think, “Yeah, that’s the one. It’s going to solve everything,” but if it’s not really addressing the issues we’re having, well, then that’s not the right tool.
I don’t want to get too much up on my high horse here, but also, before we dive into the idea of AI tools, I do want to mention that AI has become a buzzword that has lost quite a bit of meaning. A lot of stuff that is claiming to be AI-driven now is really just using some basic automation with maybe a sprinkling of smart features. I mean, even Large Language Models, when you interact with the model, it analyzes the input you've provided, calculates the most likely sequence of words that should follow, and generates a coherent response based on those probabilities. It operates by recognizing patterns and correlations from its training data rather than possessing true comprehension or awareness.
And I’m just putting this here to remind everyone that these technologies have many limitations, and it’s important that you’re not blindly following what is produced. None of these “AI” tools are thinking in any capacity, and it can be quite easy for them to lead you off in the wrong direction if you’re not paying attention.
So, with that in mind, I think the first thing we’re going to want to look at is some simple ways that we can use AI to assist with planning. One of the easiest places to start is with one of the popular Large Language Models (LLMs), such as ChatGPT, Claude, LLaMA, or Gemini. There are a ton of additional models out there as well and some of these work better than others for various tasks, but really just go with the one that you’re most comfortable with.
I like using these tools because they allow me to create easier starting points. One of the things that I notice about my ADHD is that I’m better working from something that already exists than from a blank page.
So, for example, I can load up my LLM of choice and ask it, “Can you help me plan a packing list for a 7-day trip to the Oregon Coast,” and it will come up with a pretty decent list for you. But where I find making these models really shine is not from what they initially provide but from asking the model to ask you questions. So I could follow up that initial query with, “can you ask me five questions that would improve the quality of that packing list?” And now we’ve got something that is going to really narrow down what we’re doing—and as a bonus here, I also like asking the model to ask me the questions one at a time so that I don’t get overwhelmed with too many questions right away.
This also works great for planning out your week. Just go with a prompt such as, “I’d like help planning my week. Can you ask me a series of questions that will determine what my most important tasks are and focus on making sure that I don’t get overwhelmed by doing too much each day? Please ask these questions one at a time?”
And boom, after working through the questions, you’ve got a template for how you might want your week to go.
Now, I want to be clear that this is not a foolproof process, but I think this gives us a great starting point where we don’t have to invest a lot of mental effort into getting started. I personally like using this kind of feature with the ChatGPT app on my phone because I can just do it while sitting and eating my breakfast.
The most important part to take away here, though, is the idea of having the LLM ask you questions. We want to view it as an assistant that is helping us, not something that’s making something new. To this end, if you have trouble coming up with your own prompts, you can ask the LLM for assistance with that as well. “I want to write a prompt that will help me organize a project at work. Can you ask me five questions that will help define the scope of my problem and then write a prompt that I can use to structure my project plan?”
And here’s the fun thing: I was struggling to come up with an example of what I wanted to fill in there, so I took that paragraph and started it with the prompt, “Can you help me fill in the blanks here with a few examples:” and the examples it came up with for me were, Brainstorming Ideas for a blog post, organizing a project, crafting a social media strategy, preparing for a presentation or improve a business process. I felt like the project organization one fit best for this example. That’s what I love using these services for: it gives me options. I don’t want them to try and do the actual work for me because, well, for one, I still want what I produce to be my work. But also, these systems don’t know the difference between what’s good and bad, so you can end up with some extremely crappy work. But using them as an assistant to what you're doing can be incredibly beneficial for saving brain power.
The most important thing to remember with these systems is that bad input will give you bad output. The more specific you can make your prompt, the better off you will be. However, that doesn’t let you off the hook from checking those answers and making sure that the model isn’t just hallucinating some answers.
That last point is incredibly salient. Remember that these are imperfect tools, and using them is not a replacement for doing research. Don’t just accept answers at face value; check the work.
Moving beyond LLMs there are a lot of other tools that we can look at to help with planning.
One of the tips I often talk about is how if you can break your tasks down into it’s more component parts, it makes it a lot easier to get started. When we’re looking at the big picture it can be overwhelming to get started, but just a piece of it can be a lot easier to get our mind around. But where a lot of us can get stuck is that breaking down tasks can be difficult, but there are some great AI tools that can help here, like Goblin Tools, which does just that. You enter a task, and it’ll break it down into smaller pieces, and in fact, if that smaller piece still feels too big, you can ask it to further break down that piece.
Additionally, on these fronts, tools like Todoist and Microsoft To-Do have features that go beyond just being digital checklists. With AI features, they can help you prioritize tasks based on deadlines, importance, and even your own productivity patterns.
Trello or Asana can be great for project planning. Their AI features can help you map out the steps needed to reach your goals, set deadlines, and visualize your progress. They can even suggest which tasks to tackle first based on your deadlines and workload.
Tools like Clockwise and Reclaim.ai can automatically schedule your tasks into your calendar, making sure you have dedicated time to work on what's important.
For working memory support, tools like Otter.ai or Scribe are AI note-taking tools that can automatically transcribe meetings, conversations, or verbal instructions.
Additionally, I also use Grammarly to help tighten up my writing—while the spelling and grammar help is great, the ability to make sure that what I’m writing is clear and concise is fantastic. However, I will say it can be a little too aggressive in what it corrects sometimes, and I worry that it is cutting out a bit of my voice from what I’m writing, but that’s why I don’t have to follow all of its suggestions.
As I mentioned earlier, the most important part of using these tools is making sure that they actually address what we need them to. While there is no problem with having some fun and playing around with these tools, they can also become obstacles in their own right.
To this end, I also want to mention that I don’t personally use most of these tools. With my current workflows, I can usually get what works best for me using ChatGPT. But I thought it would be good to mention a number of the ways that we can jump into this idea of using AI to help with our executive function needs.
I also want to reiterate that we need to be cautious while using these tools. While they can sometimes provide good starting points, we shouldn’t expect them to replace doing the work.
This Episode's Top Tips
We can better understand executive function through the lens of self-regulation. There are three core areas that we can look at here: working memory, cognitive flexibility, and inhibitory control.
Tools like LLMs can help us create starting points for our work. By prompting the LLM to ask us questions instead it can help us better think through our problems.
When looking to add a new tool to your repertoire, think about what problem it is going to be solving for rather than just looking for the new shiny.