Sharpening the Axe — the Weekly AI Learning Routine Your Company Needs to Install
Every Friday, the entire Fhinck team stops working and enters the classroom to learn a new AI technique. Without this rhythm, falling behind happens within 6 months.
Sharpening the Axe — the Weekly AI Learning Routine Your Company Needs to Install
Lincoln is said to have declared: "Give me 6 hours to chop down a tree and I will spend the first 4 sharpening the axe." In 2026, in AI, this has become an institutional survival rule — not a motivational suggestion.
"Every Friday Fhinck stops and enters the classroom to learn a new AI technique. Unfortunately, the pace of AI change is very fast, and if we go without studying weekly, we fall behind very quickly. (Sharpening the axe: from time to time we need to stop working to sharpen the axe, because efficiency/productivity begins to drop — with a dull axe you chop less wood in the same time.)"
— Paulo Castello, February 2026
The Ritual That Defines Fhinck's Culture
At Fhinck, every Friday is Sharpening the Axe. The entire company (all 6 Fhinckers) stops working and enters the classroom. Fixed theme: one new AI technique.
It is not a formal corporate training. It is not an event. It is a mandatory weekly routine, part of the cultural contract of working there.
The justification is not motivational. It is mathematical.
The Mathematics That Make Sharpening the Axe Mandatory
The pace of AI market evolution in 2024-2026:
- New models: approximately 1 relevant launch per month (Claude, GPT, Gemini, Llama, etc.)
- New frameworks: about 2-3 per quarter (LangGraph, AutoGen, CrewAI, MCP-based)
- Emerging standards: MCP (Nov/2024), NemoClaw (Mar/2026), and whatever comes next
- Revised best practices: continuously
A team that stops studying for 1 month is 1 quarter behind when it resumes. For 3 months: 1 year behind.
In any other field (finance, marketing, sales), a 1-year lag is manageable — the industry also evolved slowly. In AI, a 1-year lag means exiting the game.
Conclusion: continuous learning is not an "optional people investment." It is survival infrastructure.
Why One-Time Training Does Not Work
The common C-level reaction: "ok, I'll hire an annual AI training for the team."
It does not work. For two reasons:
1. Training Content Ages in Weeks
An AI course recorded in January 2026 is already outdated by March. The current standard: courses live 90 days and then become obsolete.
Sharpening the Axe generates content in real time — always based on what is coming out this week.
2. Without Cadence, No Habit Forms
An 8-hour training on a single day is corporate theater. People leave motivated, return to their routine, and forget within 30 days.
Weekly cadence creates habit — it becomes part of the DNA, not an additional task.
What Happens Inside a Sharpening the Axe Friday
Structure observed at Fhinck (adaptable):
Morning (3 hours)
- Hour 1 — Weekly Updates: each Fhincker shares 1 new AI thing they saw/tested. Collective discussion.
- Hour 2 — Practical Demo: someone presents a hands-on session of a new tool/technique. Live, with shared screen.
- Hour 3 — Collective Workshop: the entire team tests the same tool/concept in parallel. Results are compared.
Afternoon (variable)
- Individual Hands-On: each person applies the learning to a real project
- Sometimes, an external guest (practitioner — not a keynote speaker) for 1 hour
- Sometimes, a keynote watch party (NVIDIA GTC, OpenAI DevDay, Anthropic event)
What Does NOT Happen
- No formal instructor every week (presentation rotates among team members)
- No "course material" produced (content is live)
- No certificates (it is not theater)
- No "mandatory attendance" — those who are not at Fhinck by choice, are not there
How to Install Sharpening the Axe in a Larger Company
For companies with 50, 500, or 5,000 people, an adapted approach:
Phase 1 — Pilot Team (90 Days)
- CEO + 5-10 people from the most critical areas (IT, Product, Innovation, Data)
- 1 dedicated morning per week
- Structure: updates + demo + hands-on
- Metrics: tools/agents created, knowledge applied in decisions
Phase 2 — Leadership Expansion (Next 90 Days)
- All C-levels + directors
- Same format, perhaps biweekly frequency
- Metric: % of the board that personally creates agents/prompts
Phase 3 — Organizational Expansion (Next 6-12 Months)
- Each area designates a responsible person to install its own rhythm
- Organizational calendar reserves the slot
- Resources: corporate AI tool, prompt library, internal community of practice
Phase 4 — Installed Culture (12-24 Months)
- Sharpening the Axe becomes part of the DNA — even if management changes
- New employee onboarding includes the rituals
- Weekly metrics reported to the board
The Most Common Objection (and the Answer)
"We do not have 4 hours per week to stop working."
You do. You are simply prioritizing incorrectly.
The right question is: "Do you prefer to lose 4 hours per week or lose competitive advantage in 12 months?"
A company that does not stop to sharpen the axe has a team cutting wood more and more slowly, with increasing effort, generating decreasing results. The cost of inaction > the cost of the pause.
As Paulo wrote in another post:
"If we go without studying weekly, we fall behind very quickly."
Conclusion
Weekly Sharpening the Axe is not optional in 2026. It is survival infrastructure.
Companies that install it maintain competitiveness. Companies that ignore it become spectators. The difference is not talent — it is weekly discipline.
Fhinck installed this ritual in 2023 and attributes to it a significant share of its capacity to navigate the AI First transition without getting lost in the market's velocity. If you want to understand how to adapt it for your company, schedule a conversation.