TALAMANA · THE AI LITERACY MAP FOR ARCHITECTURE AND DESIGN · Foundations · AGE 11—15 · FACTUAL · DURABLE

Machines learn from examples

The examples are the teacher, so bad examples teach badly.

The idea

Many AI systems are not given a rule for every situation. Instead, they are shown a very large number of examples. The machine finds patterns in those examples, and then makes guesses about things it has never seen. The examples matter. What is missing, repeated, or wrongly labelled changes what the machine learns.

Why it matters

If the examples are the teacher, then bad examples teach badly. That one idea explains most of what goes wrong with AI.

See it in the studio

A tool that has seen a million photos of glass towers and very few brick courtyards will keep offering you glass towers. It is not being stubborn. It is showing you its diet.

Watch for this

Thinking the machine "knows" things the way you do. It has patterns from examples — not experience, not understanding.

Try it

Show a friend ten photos of cats and none of dogs. Then show them a husky and ask what the machine would say. Talk about why.

Prove it

Tell someone, in your own words, where a machine like this gets everything it knows how to do.

How it works

This way of building software is called machine learning. The examples together are called training data. The patterns the machine keeps are called a model. Later ideas on this map — bias, provenance, whose work taught the machine — all begin right here, in the diet of examples.

Lineage

For decades, people tried writing rules for machines by hand: if this, then that. The rule books grew huge and still missed too much. Learning from examples took over because the world holds more examples than anyone can write rules for.

What this idea builds on

What this idea opens up

Sources

Open this idea on the map · The complete map · Logika · RBDS AI Lab, India · revised every edition.

Age grows from 11 at the centre to 22 at the edge, and six sectors show the learning strands. Tab into the map and the arrow keys step from idea to idea, following the links where there is one. Enter opens the idea under the cursor, and E reads out its links and the reason recorded on each. Press slash for Search, question mark for the full key list, and Escape to leave. Open Ideas for the complete readable list, including what each idea builds on and what it opens up.

LOGIKA · RBDS AI LAB INDIA
ON-RAMP · AGE 11 · FIRST ENCOUNTERS, NOT GATES — IDEAS · — DEPENDENCIES
DONE
OPENS NEXT
SOLID — STANDS ON