Six months ago, I was the guy with 11 AI tools running at once.
I had a tool for writing. A tool for scheduling. A tool for transcription, research, inbox management, image generation, and three different assistants I was “testing.” My desktop looked like a cockpit. I felt like a pilot.
Spoiler: I was not flying. I was drowning.
If you’re a remote professional or hybrid worker juggling AI subscriptions and wondering why you feel busier than ever, this one’s for you. Because the problem isn’t that you need better AI. The problem is that your AI stack is eating your focus alive.
Here’s how I went from 11 tools to 3, and why that single decision doubled my ability to do work that actually matters.
TL;DR – Key Takeaways
- More AI tools does not equal more productivity. BCG research on 1,488 workers found that productivity peaks at 2-3 tools and drops sharply at 4 or more.¹
- Every AI tool you add is a relationship to maintain: a learning curve, a review process, context switching, and cognitive overhead.
- The 3-Tool Rule: keep only the AI tools that pass the F.A.K.E. Framework filter. Everything else? Delete it.
- My final stack: Claude (strategic thinking), Notion AI (digital brain), Nano Banana/Gemini (visual creation). That’s it.
Why Does More AI Feel Like More Work?
Because it is more work. And now we have the data to prove it.
Boston Consulting Group published a landmark study in Harvard Business Review in March 2026, surveying 1,488 full-time U.S. workers. The central finding? Workers using 4 or more AI tools simultaneously experienced a measurable drop in productivity, while those using 1-3 tools reported genuine gains.¹
BCG researchers coined the term “AI Brain Fry” to describe the phenomenon: mental fatigue caused specifically by the cognitive overhead of monitoring, reviewing, and managing AI outputs.² Workers experiencing it reported mental fog, slower decisions, more errors, and a greater intention to quit.¹
A separate eight-month study from UC Berkeley’s Haas School of Business backed this up. Researchers embedded in a 200-person tech company found that AI tools didn’t reduce work. They expanded it. PMs started coding. Researchers took on engineering tasks. Role boundaries dissolved. The to-do list grew to fill every minute AI freed up.³
The paradox is brutal: the people using AI the most are burning out the fastest.
How I Hit Rock Bottom With 11 AI Tools
I won’t pretend this was someone else’s story. This was mine.
At my peak AI tool overload, here’s what my stack looked like:
- 4 writing assistants (because “each one has a different strength”)
- 2 scheduling/optimization tools
- 1 AI transcription service
- 1 research aggregator
- 1 AI inbox manager
- 1 image generation tool
- 1 strategic AI partner
My calendar filled up with tasks like: reviewing AI outputs, troubleshooting integrations, learning new interfaces, and managing the system I’d built to manage my other systems.
I was doing more admin work than before AI existed.
One Wednesday, I spent 3 hours “optimizing” a report nobody would read. I had AI summarize a meeting that should have been cancelled. I used a writing assistant to draft a message that should have been a 10-second phone call.
That was the day I stopped asking “What can AI do for me?” and started asking “What should AI not do for me?”
The F.A.K.E. Framework: Your AI Stack Filter
Before deleting anything, I needed a filter. Not a gut feeling. A repeatable system.
That’s where the F.A.K.E. Framework comes in. I originally designed it for task prioritisation, but it works just as well for evaluating your AI stack.
Run every tool through these four questions:
| Filter | Question | What It Catches |
|---|---|---|
| F – Focus | Does this tool serve my #1 priority this week? | Tools that feel useful but serve someone else’s urgency |
| A – Alignment | Does it move me toward my North Star goals? | Shiny objects disguised as “productivity” |
| K – Knowledge | Can I actually evaluate the quality of its output? | Tools where you’re blindly trusting AI you can’t verify |
| E – Energy | Does using it give me energy or drain it? | The tools that cost more cognitive overhead than they save |
The rule: if a tool fails 2 or more filters, it doesn’t stay.
Not “pause it”. Not “revisit next month”. Delete it.
What I Kept (and Why)
After running the filter, three tools survived. Here’s the honest breakdown.
Tool 1: Claude (Strategic Thinking Partner)
Claude isn’t my copywriter. It’s my sparring partner.
I use it for long, contextual conversations where I share strategy, get challenged, and think through complex decisions. Content planning, audience analysis, framework development. It replaces the “thinking out loud” I used to do in meetings.
Why it passed the F.A.K.E. filter: it directly serves my #1 weekly priority (Focus), aligns with my content strategy (Alignment), I can evaluate its reasoning quality (Knowledge), and the process of working with it energizes my thinking (Energy). 4/4.
Tool 2: Notion AI (Second-Digital Brain Layer)
Notion AI sits on top of everything I’ve already written, planned, and captured. It surfaces connections, summarizes past thinking, and helps me find patterns across months of notes.
Why it passed: it amplifies existing knowledge instead of replacing thinking. It organizes what’s already mine.
Tool 3: Nano Banana / Gemini (Visual Creation)
TSP concepts become stunning visuals in 3 minutes instead of 3 hours. That’s roughly 2 hours per week reclaimed for deep work.
Why it passed: clear Focus (visuals are essential for content), perfect Alignment (brand-specific output), I can judge visual quality instantly (Knowledge), and the speed boost is genuinely energizing (Energy).
What I Deleted (and Why)
Here’s the uncomfortable part. I had to admit that most of my AI tools were comfort purchases, not productivity investments.
4 writing assistants: They blurred my voice into a generic AI smoothie. When you use three different writing tools, you end up sounding like none of them and none of yourself.
2 scheduling optimizers: They created more decisions, not fewer. Every “smart suggestion” was another thing to evaluate, accept, or reject. The overhead exceeded the benefit.
1 AI transcription tool: Deleting this one forced me to confront the real problem: I was taking too many meetings. Losing transcription didn’t hurt my productivity. It improved my calendar discipline.
1 research aggregator: Perplexity does 80% of what this tool did at 20% of the setup cost. Consolidation over addition.
1 AI inbox manager: I replaced it with email batching, a technique that costs nothing, requires no integration, and works perfectly. Sometimes the best tool is a better habit.
The Principle Behind the 3-Tool Rule
Here’s what I want you to take away from my story.
Every AI tool you keep is a commitment to maintain it. A learning curve. An update cycle. A review process. A context switch. If maintaining the tool takes more energy than the task it replaces, it’s not a productivity tool. It’s a productivity tax.
The BCG data confirms this isn’t just my experience. Workers whose AI use required high levels of oversight reported 14% more mental effort, 12% more mental fatigue, and 19% more information overload compared to those who used AI for simple task replacement.²
Greg McKeown wrote in Essentialism: “The disciplined pursuit of less.” AI doesn’t change that principle. It raises the stakes. Because now “less” isn’t just about saying no to meetings and emails. It’s about saying no to tools that make everything feel easy and possible, even when it shouldn’t be done at all.
This is the premium skill of 2026: knowing what NOT to automate.
How to Apply the 3-Tool Rule This Week
You don’t need to blow up your stack overnight. Here’s a simple starting point:
Step 1: List every AI tool you currently use. Include the free ones. Include the ones you’re “just testing”.
Step 2: Run each tool through the F.A.K.E. filter. Be honest. If it fails 2 or more filters, flag it.
Step 3: Delete the flagged tools. Not “pause.” Delete. Remove the bookmark. Cancel the subscription. Get it off your screen.
Step 4: Commit to your remaining stack for 30 days. No additions. If the urge to add a new tool hits, ask yourself: “What would I remove to make room for this?”
I’ve created a free ‘AI Stack Audit Checklist.pdf‘ to walk you through this exact process. You can find it at the link below 👇
If you want the full system I use, including the F.A.K.E. Framework, the TSP Method, and the weekly rhythm that keeps my focus protected, check out the Productivity Nirvana Community and Online Course.
But start with the checklist first. Build the habit of auditing before you build anything else.
Recommended Reading
AI Brain Fry: Why the People Using AI the Most Are the Most Burned Out
The deep dive into the UC Berkeley and BCG research behind AI burnout, and how the F.A.K.E. Framework works as a daily filter.
FAQ
How many AI tools should I use?
BCG research suggests productivity peaks around 2-3 AI tools and declines at 4 or more. The goal isn’t a specific number but rather ensuring each tool genuinely earns its place in your workflow by reducing cognitive load instead of adding to it.
What is AI Brain Fry?
AI Brain Fry is a term from a 2026 BCG Henderson Institute study describing mental fatigue caused by excessive monitoring and oversight of AI tools. Symptoms include mental fog, decision fatigue, increased errors, and greater intent to quit.
What is the F.A.K.E. Framework?
F.A.K.E. stands for Focus, Alignment, Knowledge, and Energy. It’s a filter developed by TSP to evaluate whether a task (or tool) deserves your attention. If something fails 2 or more of the four filters, it should be deleted, not automated.
Can I use more than 3 AI tools if I need them?
Yes. The “3” isn’t a rigid cap. It’s a principle: fewer tools used deeply will always outperform many tools used shallowly. The right number for you is the number of tools that pass the F.A.K.E. filter without adding cognitive overhead.
What’s the best way to start reducing my AI stack?
List every tool you use, run each through the F.A.K.E. filter, and delete anything that fails 2+ filters. Commit to your reduced stack for 30 days before making any additions.
This article is a co-creation of Erick Stoic with Claude (Anthropic) and Nano Banana/Gemini 🍌.
References & Further Reading
- Bedard, J., Kropp, M., Hsu, M., Karaman, O.T., Hawes, J., & Rosen Kellerman, G. (2026). “When Using AI Leads to ‘Brain Fry.’” Harvard Business Review. https://hbr.org/2026/03/when-using-ai-leads-to-brain-fry
- Fortune. (2026). “‘AI brain fry’ is real — and it’s making workers more exhausted, not more productive, new study finds.” https://fortune.com/2026/03/10/ai-brain-fry-workplace-productivity-bcg-study/
- Fortune. (2026). “AI promised supreme productivity, but it’s actually straining workloads for employees.” https://fortune.com/2026/03/13/ai-isnt-reducing-workloads-its-straining-employees-time-spent-emailing-doubled-deep-focus-work-fell/
- CNBC. (2026). “Using AI can add extra labor and cause ‘brain fry’ for workers, experts say.” https://www.cnbc.com/2026/04/06/companies-are-pushing-ai-but-experts-say-it-can-add-extra-labor-cause-brain-fry.html
- McKeown, G. (2014). Essentialism: The Disciplined Pursuit of Less. Crown Business.
Leave a comment/insight!