AI as a Tool for Managing ADHD
OR, How I Learned to (Mostly) Stop Worrying About AI and Love My ADHD
October is ADHD Awareness Month. I was diagnosed with ADHD two years ago. It changed everything I knew about my mental health. Over the past few months, I’ve used AI to help make work more manageable, challenge my unrealistic expectations of what I can get done, and change how I feel about my work.
A Lifetime of Trying to Organize Myself
For as long as I can remember, I've wanted (and struggled) to keep track of everything I need to do.
I've tried spreadsheets, task lists, Todoist, Airtable, Google Docs and countless other approaches. I'd develop a system, use it for a week or two, and eventually stop. When I returned, the information would be hopelessly out of date. The problem wasn't necessarily managing individual projects. I could create a project timeline and understand the deliverables. The difficulty was seeing everything together.
What's due tomorrow? What's due Friday? How do I balance multiple projects, teams, clients and the long list of personal projects?
Everything felt urgent. I was holding everything in my head. All. The. Time.
I never felt I’d accomplished enough. There was always more waiting. Then, when I had days without any meetings or deadlines, I’d struggle to get things done.
It's not that I didn’t know what needed to be done. It was deciding where to start, what deserved attention and whether I was working on the right thing that could be overwhelming.
Join me and Kirby Hansen-McNulty for Ep. 59 of Shining Light on Shadows: A Candid Conversation About Mental Health, "Understanding and Living With ADHD; AI as a Tool for Managing ADHD" on Thursday, October 22 at 7 pm ET / 6 pm CT. You can watch the live show or recording on Facebook*, Twitter**, LinkedIn, YouTube, Instagram** or my website.
*Facebook will delete the video in 30 days (approx. November 21)
**We won't know the exact urls for Twitter or Instagram until we go live. For now, these links go to my Twitter and Instagram accounts.
Understanding My ADHD Changed the Equation
Being diagnosed with ADHD and working with Calibrate ADHD helped me recognize patterns I'd lived with for decades.
Magical Thinking: Believing I could accomplish ten things when I realistically had time for four.
Black and White Thinking: If I didn't accomplish everything I set out to do, I hadn't done enough.
Fluctuating Energy: I could work intensely for several days and then find myself operating at a fraction of my normal capacity. For years, I sometimes interpreted those crashes as depression. I've come to recognize that intense periods of work can also leave me depleted and needing time to recover.
Although Calibrate ADHD helped me recognize these patterns, I still struggled to put what I was learning into practice. Over time, I began translating that self-awareness into systems that worked better for me. ChatGPT became one of the tools I used to help put those systems into practice.
How I’ve Used ChatGPT to Help
I have plenty of concerns about AI, the companies developing it and some of the tech bros running them. Those concerns haven't disappeared simply because I've found the technology useful.
On the flip side, it has been a game changer for me in ways large and small.
It has streamlined weekly planning and time blocking by bringing together individual work plans for clients (each with multiple deliverables, deadlines and time estimates for different projects) with my weekly calendar and certain parameters (meals, sleep, family time). I can get a sense of how to get my work done, even when I think it's impossible.
I don't always follow those plans or keep to my boundaries, however. I still work late sometimes and put in hours of work on the weekends.
There are two key points here. I spend time going back and forth with ChatGPT until the plan meets my needs. More importantly, I use it as a suggestion, not as a mandate.Just knowing there’s a plausible way to get things done helps to lower my anxiety.
Tracking my time is useful for billing clients, but it also helps me plan future work. Estimating how long something will take has always been a challenge for me. Having actual data, and a tool to analyze it, means my weekly plans are grounded in experience rather than wishful thinking.
Making Work Feel Finite
Having a realistic plan doesn't mean I'll finish everything. I've never reached the bottom of my to-do list, and I doubt I ever will.
But if I know Thursday's work has been accounted for, I don't have to keep worrying about it on Tuesday. I can concentrate on what's in front of me.
That reassurance changes how I approach the day. Sometimes, I can finish the work I've planned, close my laptop and stop—even though plenty remains unfinished.
After decades of feeling I haven't done enough, that's a significant change.
Of course, things don’t always work out. Sometimes, I need to defer a project, reduce the scope or just take it off the list. It helps to get feedback on those choices.
That same challenge comes up when I'm working on individual assignments.
A report can always be improved. A graphic can always be adjusted. A presentation can always be refined. Sometimes those changes make a meaningful difference.
Other times, I'm spending another hour adjusting something that was already perfectly acceptable. Knowing I need to move on to the next project, and having ChatGPT actually tell me, “You’ve done enough” makes a difference.
Using AI as an External Memory
Another major benefit has been using ChatGPT as an external memory.
I've never considered myself especially good at taking notes. My notes tend to be either nearly verbatim or so abbreviated that they're difficult to interpret later.
Working across multiple clients and projects, I need to track decisions, assignments, commitments and unresolved questions.
Having an external system that helps me organize, retrieve and, frankly, interpret information reduces the burden of trying to hold everything in my head.
This is especially valuable because I work independently. I don't always have a colleague available to bounce ideas off of.
ChatGPT can help me examine the wording of an exchange, consider different interpretations and identify what I might have overlooked.
Of course, I still need to verify what ChatGPT tells me. Its interpretation of a conversation isn't necessarily accurate.
Closure: Learning to Put Things Behind Me
I've discovered another benefit that I hadn't necessarily anticipated: closure.
When I complete a project or activity, I often have an intense need to understand how it went.
Did I handle something well? Did I miss an opportunity? What worked? What should I do differently next time?
Sometimes, without a way to evaluate the experience, I continue replaying it in my head. Having ChatGPT give me lessons learned and a debrief is really helpful.
A good example is reviewing the transcript and audience comments after a livestream. I used to call a colleague and ask her, “So, what did you think?” It really helps to put a show behind me when I can understand what went well and what I could improve next time.
Exploring AI as a Cheat Code for Therapy
I've also been wondering whether AI could help me reflect on and retain information from a much more personal experience: therapy.
I've been in talk therapy on and off for decades. My current therapist and I have worked together for several years.
Therapy is a long process. Important themes recur. We examine situations, identify patterns and discuss different ways of understanding or responding to experiences.
But I sometimes struggle to retain what we discussed.
I can leave a session with important insights and find that I've forgotten some of the details almost immediately.
I've experimented with having ChatGPT summarize session transcripts and asking my therapist to evaluate whether the summaries accurately represent the discussion.
The concept is somewhat like an after-visit summary from a physician's office.
What did we discuss? What were the important observations? What suggestions or practices should I pay attention to before the next session?
I’m not using AI as my therapist. I’m exploring whether it can help me remember and reinforce what I discussed with my therapist.
This has also led me to think harder about the privacy implications. Turning off model training doesn't mean the company providing the AI service no longer has access to information I've uploaded. That's a particular concern with therapy transcripts and confidential professional conversations.
I'm exploring locally hosted AI as a way to retain greater control over sensitive information.
I'm also becoming more deliberate about what I share with cloud-based AI systems in the first place.
The Limits of ChatGPT
Privacy isn't my only concern. ChatGPT can make mistakes, misunderstand context and be excessively agreeable. I've explicitly instructed it not to simply tell me what I want to hear. I want it to challenge my assumptions, identify weaknesses in my reasoning and push back when appropriate.
That helps, but it doesn't eliminate the problem. And sometimes, managing ChatGPT becomes another task in itself, adding to the executive-function burden it's supposed to reduce.
Am I Cheating by Using AI?
There's another question I continue to wrestle with. Sometimes, relying on ChatGPT feels a little like cheating.
I don't think the answer is simple. I rarely accept an initial AI-generated response without review.
I provide context and instructions, evaluate the result, correct errors, challenge assumptions and refine the material until it reflects what I'm trying to communicate.
Sometimes that process involves extensive back-and-forth.
I'm not turning my judgment over to ChatGPT. I'm using it as another resource in developing my own work.
But I don't want to surrender my creative judgment, my ability to evaluate information or my capacity to think critically about a situation.
And I don't want to become so dependent on AI that I'm unable to function effectively without it.
Part of learning to use these tools involves deciding which capabilities genuinely benefit from assistance and which deserve continued practice.
ChatGPT Is Not in Charge
I've think of ChatGPT as my bench coach (à la Don Zimmer) or my chief of staff (Leo McGarry).
A good adviser can help organize information, evaluate options, challenge assumptions and identify things I may have overlooked.
But an adviser doesn't get to make the final decisions.
I do.
That includes deciding when to follow a recommendation and when to reject it.
Sometimes ChatGPT doesn't fully appreciate my priorities. A schedule might look perfectly reasonable on paper but leave too little time for family, friends or simply getting some rest. Those are choices I have to make, not ChatGPT.
Being in charge means understanding the limitations of the technology, verifying important information and recognizing when a particular tool is not appropriate for a particular task.
It also means being willing to reconsider a workflow that previously seemed like a good idea.
What I'm Still Learning
I haven't figured all of this out. I still underestimate how long things take.
I still struggle with perfectionism, distractions and occasionally getting started.
I still have days when I accomplish far less than I'd hoped.
Sometimes I ignore the schedule and work far longer than I intended. Sometimes I spend too much time talking to ChatGPT about how to manage my work instead of actually doing it.
But increasingly, I have a way to understand what's on my plate, recognize what's realistic, decide what matters and move forward.
I can change a plan without treating the change as a failure.
I can sometimes recognize when I've done enough.
I can focus on what I'm doing now because I have more confidence that the other responsibilities have been accounted for.
And I can sometimes stop working without feeling that I've failed simply because there is more to do.
The distinction between productivity and personal worth is becoming more meaningful to me.
For much of my life, how much I accomplished influenced how I felt about myself.
An exceptionally productive day could feel validating. An unproductive day could leave me discouraged, regardless of what else might have been happening in my life.
I'm beginning to understand that I need to be able to accept both kinds of days.
The goal isn't to become a perfectly optimized person.
It's to develop systems that help me live more effectively while protecting the relationships, activities and values that make life meaningful.
And those systems don't have to be perfect.
They just have to be useful enough that I can keep returning to them.
The larger lesson isn't that ChatGPT is managing my ADHD for me. It's that I've adapted a tool to support the ways I'm learning to manage ADHD myself.
Related Blog Post
Comments