Conscious Tinkering
A method for helping organizations adopt AI without big budgets, technical teams, or multi-year strategy decks. Bottom-up, hands-on. One small bet at a time, until they've changed how your organization works. This is how building turns into leadership. Used on a daily basis by the Ashoka AI Lab.
What it is
AI adoption often gets stuck because management teams believe in the McKinsey playbook: A strategy must be commissioned. A steering committee must pick priorities. A vendor must scope a pilot. Eighteen months later, the strategy is out of date and nothing has shipped. The people who were supposed to use AI never touched it.
Conscious Tinkering inverts this. The people doing the day-to-day work (not committees, not consultants) pick tasks they struggle with and build something to help. Sounds boring, but reliably changes how your organization works over time. Works especially well for nonprofits, foundations, social ventures, and teams inside larger orgs that cannot wait for a grand strategy.
The loop
Three steps. Each pass adds value to at least one team and improves at least one pillar of AI Readiness. Become more and more ambitious over time while staying in control.
Aim
Find a bottleneck on a team where the fix would also strengthen an AI Readiness pillar (see below).
Build
The simplest thing that helps. 40 to 80 hours, no more. Applies to prototypes and new features for existing flagship applications.
Compound
The team gets new value. At least one pillar of AI Readiness is a notch stronger. So the next iteration can be more ambitious.
Conscious Tinkering was featured by
Five pillars of AI readiness
The "conscious" part. With each app, you improve what your organization is capable of. None require a strategy doc to start. At least one gets stronger every time you ship.
Governance
Clear rules for what's safe to ship and who decides.
IT Infrastructure
The plumbing: accounts, hosting, integrations.
Tinkering Capacity
People with permission and skill to build things.
AI Literacy
The ability of teams to use AI tools confidently and responsibly.
Data Management
Having the data you need, in the form you need it in.
Ashoka AI Lab
I developed this method with my partner in crime Daniela Matielo while co-leading the Ashoka AI Lab.
We built simple apps to put AI on Ashoka's agenda. Storytelling Coach is a live audio roleplay partner that gives feedback on how you communicate Ashoka's core ideas in your own words. Fair Fit helps you write job posts that follow Ashoka's Diversity and Includion guidelines. Some are in daily use. Others were retired after a week. Some inspired the flagship applications that are changing how Ashoka works.
Like AIshoka, a chatbot that combines a dozen applications, including access to Fellow data, live news from the field, board books, and running due diligences on potential partners. Or Rapid, a comprehensive AI system that collects updated information from Ashoka Fellows, which allows us to do pretty much everything at Ashoka more efficiently.
Small team. Three years. 30+ prototypes. 4 flagship applications. That's the power of Conscious Tinkering.
Want to apply this in your organization?
I take a small number of engagements per year.
See engagements