START ↗

Think you can beat the machine?

A little Pong. A better question.

Continue to Think →
Explore Human + machine ↓

THINKING

The more powerful the machine,
the more human we must become.

The machine can give us answers. We still have to decide which questions matter, what should exist and what happens next.

This is the belief behind our work: as technology gives us more capability, human judgement, imagination and care deserve more attention.

An established Ideas Shed robot illustration

A THOUGHT TO TAKE WITH YOU

Stay curious.

Certainty gets things done. Curiosity discovers what might be worth doing.

New opportunities often begin when somebody notices the assumption everyone else has stopped seeing. We want curiosity and creativity to be valued as highly as certainty.

A THOUGHT TO TAKE WITH YOU

Ask better questions.

Before producing answers faster, make sure the question is worthwhile.

AI makes immediate answers abundant. Framing the real problem, testing assumptions and staying curious become more valuable. It helps to examine the premise before falling in love with the answer.

A THOUGHT TO TAKE WITH YOU

Use human judgement.

Not everything that can be built should be built.

Judgement decides what matters, what is responsible, what is useful and what stands a credible chance of working. Human understanding keeps the consequences in view, including for people who did not choose the technology.

A THOUGHT TO TAKE WITH YOU

Rethink what is possible.

Powerful technology expands what small teams can do.

We can research, design, build, test and improve at a speed that changes the economics of making things. That opens new possibilities for work and enterprise, while asking difficult questions about whose skills, opportunities and livelihoods change.

A THOUGHT TO TAKE WITH YOU

Learn from reality.

An idea creates value when it meets the world.

People use it, ignore it, misunderstand it or improve it. Each outcome teaches us something. Imagination gives us alternatives; evidence helps us decide what to do next.

An imagined illustration of Albert Einstein smiling beside the Pink Cow.

ROOM FOR IMAGINATION

WHAT
IF?

Creativity connects what seemed unrelated and suggests a route we had not considered. Our playful experiments give that instinct somewhere to go.

Explore Creativity

INSIDE THE COW SHED

We build our own tools.
Then we show our workings.

We use these systems ourselves, learn from the awkward bits and carry those lessons into the things we build for other people.

Not a proprietary process. Here are the rules. Here are the tools. Here is what we are learning.

Use the thinking if it helps you too.

Build it → Use it → Find the awkward bits → Improve it
Pinky, our hand-drawn pink cow, with his quaint cup of tea

THE SHED’S WORKBENCH

Tools from inside the Shed.

01 / 07Living rules

Cow Rules

The evolving rules and philosophy for building useful, trustworthy software with AI.

Real builds create mistakes, friction and lessons worth remembering. Cow Rules stops the next build relearning them the hard way.

Good rules are earned in use. They stay small, sharp and open to change.

Open the public tool ↗
02 / 07Control room

Cow Console

The shared record of projects, stages, evidence, builds, activity and next actions across the Shed.

Many ideas create noise. The Console makes the current truth and the next useful move easier to see.

Momentum improves when evidence and next actions live together.

Open the public tool ↗
03 / 07Shared connections

CowBridge

Reusable, controlled bridges that help Ideas Shed sites, software and Cows talk to other systems and to each other.

Useful connections should be built once, permissions should stay honest, and the estate should not become pink spaghetti.

Do not build a bridge when there is already a perfectly good gate.

Open the public tool ↗
04 / 07Evidence trail

CowTrace

An experiment in preserving useful AI context with the evidence and provenance behind it.

An AI should not be the only place that remembers, and a human should be able to ask why it believes something.

Context can travel without disguising interpretation as fact.

Open the public tool ↗
05 / 07Second-brain experiment

Cowbrain

A philosophy-first experiment in memory, context, provenance and changing human judgement.

Remembering a person is not the same as defining them. Useful AI memory must leave room for people to change.

Human meaning and machine memory need negotiation, not blind storage.

Open the public tool ↗
06 / 07Useful starting point

Cow One

A focused public starting point for putting the Cow tools and their thinking to work.

A growing family of tools still needs one clear place to begin.

A useful front door matters as much as the machinery behind it.

Open the public tool ↗
07 / 07Field notes · evolving

Cow Trek

A developing way to track what AI can genuinely do now, what is changing and what the Shed should build with it.

Capability moves quickly. Decisions need evidence from the current frontier, not last season's hype.

Separate what works now from what is promising and what is not dependable yet.

Public page not yet safely linked.

Swipe to explore all seven tools