Getting Things Done Podcast
Episode 330: GTD and AI
October 1, 2025
Host: John Forrester (A)
Guest: Deb Smith Hemphill (B), Management Productivity Consultant
Episode Overview
This episode takes a deep dive into the intersection between GTD (Getting Things Done) methodology and artificial intelligence (AI). Host John Forrester interviews productivity consultant Deb Smith Hemphill about how the rapid integration of AI is shaping productivity systems, client adoption, and the cultural shifts needed to harness technology as a true productivity partner. Deb discusses practical client examples, key AI concepts, prompt engineering tips, and how GTD principles amplify the value of AI.
Key Discussion Points & Insights
1. Deb’s Background & Introduction to GTD
- Location & Work: Deb is based in the East Bay near San Francisco and leads a management productivity consulting firm that shifted to a remote work model post-Covid. Her team handles training, executive coaching, and special projects—especially AI implementation. [00:45]
- GTD Journey: Deb discovered GTD through David Allen’s book and a San Francisco workshop, striking up a relationship that continues today. GTD has remained her core productivity system, even as tools evolved: “as we've evolved tools over the years and we've gone from paper to electronic to groupware to...AI based, it's still always with the GTD thought process and structural thinking in mind.” [03:44]
2. Types of AI in Productivity Work
Deb outlines four main AI types her clients interact with—each with different productivity potentials:
-
Generative AI:
Writes, summarizes, generates reports. E.g., ChatGPT, Perplexity.
"[It] takes her about 10 minutes. So it's a really useful tool in terms of productivity for her." [06:32] -
Boundaried AI:
Operates with strict data boundaries (e.g., a legal department that can search outwards but protects internal data). -
AGENC/Agentic AI:
"Transformational...designed to take something and make it into something else." For example, responding to customer queries, moving tasks through sales chains autonomously. [07:16] -
Embedded AI:
“Subsurface” AI baked into apps (like Tana), automating tasks behind-the-scenes:
"I don't have to think about it, I don't have to see it, but it's operating there sort of silently, behind the scenes." [08:40]
3. AI as a Collaborative Teammate, Not Just a Tool
-
Deb repeatedly emphasizes AI is most powerful when treated as a “collaborative teammate” that augments human thinking, not something to fear or simply automate busywork:
"We really want our clients to use it as a collaborative teammate...like it's one more person on the team." [06:57] -
Key client exercise: Identify “the tasks on your list that make you say, oh, if I have to do one more of those, I'm gonna scream” and use AI to teach those tasks’ criteria. [10:14]
“I said, so if you had a clone with your brain of brilliance attached...you would give it to the brain, right?...as long as you can teach it the criteria that you use.” [10:52]
-
Meta-cognitive Skill:
"What you're encouraging clients to do is think about how they think and then use AI to train...but to get there, they have to know, oh, I am thinking in a certain way" [12:57, Forrester].
4. Prompting Skills & the New Work Interface
- “Our new work interface is the prompt window.” [15:11]
- Deb recommends foundational courses like Google Essentials for prompt writing.
- Story about switching from filing system mindset to relying on search:
"He said, you're supposed to apply your brilliance to creating the document. The computer is supposed to take it from you, park it someplace safe, and bring it back when you need it." [18:08, Apple Trainer] - Context trumps search:
"Context is what beats search today...giving it really good context…here's the situation, here's what I'm looking for, here's the outcome that I want to generate.” [21:13]
5. Lazy vs. Leveraged AI: The Importance of Contextual Conversation
-
Deb shares a memorable metaphor:
“If you're using AI as a search tool, it's kind of like using a Stradivarius to teach your 5-year-old how to play violin.” [20:50]
-
Lazy AI: The student asks, “Write my report on Abraham Lincoln.” No context—minimal output.
-
Leveraged AI: The student shares context (“I’m an 8th grader”), asks what should be included, then iterates in conversation—a model for effective prompt building and learning. [23:03]
6. Structured Prompting for Better Results
-
Recommend Jeff Woods’ CRIT model from "The AI Driven Leader":
- Context
- Role AI should play
- Interview: Invite AI to ask clarifying questions (3–5)
- Task you want performed
“Literally inviting the large language model to ask you questions...really use it as partner, as opposed to just a tool.” [25:00]
-
Comparison to Directing Actors:
“It’s a whole different thing than just saying, read these lines.” [26:16, Forrester]
-
Higher-value output comes from treating AI as a partner and continuously guiding and iterating.
7. Navigating the AI Productivity Culture Shift
- The major differentiator among clients isn’t the type of AI but their adoption and culture—from cautious tool users to organizations where AI is fully integrated as a trusted teammate [09:48].
- For managers: AI will handle lower-level, reactive, “fire-fighting” tasks, freeing leaders to focus on strategic, advanced work. Those only operating at a superficial level risk being left behind:
"If you only do first-level thinking and you're not capable of second-level thinking, then that is a possibility." [27:00]
8. GTD Methodology’s Role in the Age of AI
- Clarity of desired outcome remains essential—otherwise, “you’ll just get something that’s kind of good. It’s good enough.” [30:22]
- The Natural Planning Model is as vital for briefing AI tools as it is for humans.
- Deb even experimented with prompting AI to answer as “David Allen using the natural planning model”—with excellent results. [31:56]
"Outcome thinking is one of the most valuable things that there is." [30:41]
- Nvidia CEO Jensen Huang’s advice:
- Flatter the AI, e.g., “You are the world's expert.”
- Give clear context and desired outcome.
- Iterate politely for best results.
"When it gives you back stuff, you still have to be nice...If I don't understand what that criteria is...I haven't done the thinking." [32:52]
9. Critical Thinking Over Rote Use
- Early “hallucinations” (fabricated AI answers) occurred because users didn’t provide enough context or verify outputs. Example: the lawyer who lost their license after submitting fake AI-generated cases. [35:24]
- The real risk isn’t fake papers in academia, but fake sources—a call to teach students and professionals critical thinking and appropriate use of AI. [36:39]
"We should be teaching people how to think...Let's give them the tools they need to use the tools they're going to have available to them in the future." [36:47, 38:40]
Memorable Moments & Notable Quotes
-
On AI as a Team Member:
“Not as a tool, not as something to be afraid of, but as a collaborative teammate. Like it's one more person on the team.” (Smith Hemphill, 06:57)
-
On Search vs. File Structure:
"He said, you're supposed to apply your brilliance to creating the document...the computer is supposed to take it from you, park it someplace safe, and bring it back when you need it." (Apple trainer to Deb, 18:08)
-
On Prompts & Context:
“Context is what beats search today...context and the back and forth conversation that you have with it that really gets you the value.” (Smith Hemphill, 21:14, 23:03)
-
On AI Adoption & Higher-Level Work:
“AI is going to either fight or eliminate a lot of those [firefighting problems] and they have to be ready to jump to that second and third level thinking.” (Smith Hemphill, 28:04)
-
On GTD’s Role:
"Outcome thinking is one of the most valuable things that there is. And the natural planning model is a wonderful way to think 360 degrees around something." (Smith Hemphill, 30:41)
-
On Critical Use of AI:
"When we first start with clients, we suggest to them that they think of AI as a really, really scary, smart intern who's been with you for two weeks." (Smith Hemphill, 34:47)
Useful Tips & Takeaways
- Prompt structure is everything: Ground your AI requests in clear context, desired roles, specific questions, and well-defined tasks.
- Treat AI as a teammate: For best results, “train” AI in your thought processes and iterate with feedback.
- Leverage, don’t replace thinking: Use AI to amplify your creativity and strategic output, not as a substitute for original thought.
- Outcome clarity matters: The more specific your goals, the better AI (and you) will perform.
- Education must shift: Teach people how to think critically—and how to use modern, context-driven tools.
Timestamps for Key Segments
- [00:45] – Deb’s background and company overview
- [01:58] – How Deb met David Allen and began GTD
- [03:44] – The evolution of productivity tools in Deb’s practice
- [04:51] – Four types of AI in client work and productivity
- [06:47] – Treating AI as a collaborative teammate; client examples
- [10:14] – Teaching criteria to AI; not “losing thinking”
- [12:57] – Meta-cognition: teaching clients to know how they think
- [15:11] – The “prompt window” as the new work interface
- [18:08] – Search vs. file/folder paradigm change
- [20:50] – Lazy AI vs. Leveraged AI metaphor
- [23:03] – Iterative, conversational prompts for leveraged AI
- [25:00] – The CRIT prompting model
- [27:00] – How AI shifts the manager’s role to higher-level thinking
- [30:22] – Applying GTD’s outcome thinking and planning model to AI
- [32:52] – Flattery, context, and polite iteration for best AI results
- [35:24] – The risks of hallucination, academics, and AI misuse
- [36:47] – Deb’s education philosophy and call for teaching thinking skills
This episode offers both a practical and philosophical roadmap for using AI as a productivity partner, as long as users bring GTD-level clarity, outcome thinking, and iterative partnership to the table. It's an essential listen for GTD enthusiasts and anyone wanting to harness AI without losing their edge as a critical, creative thinker.