↗ All paths / Find your questionGutSentry · Evolution Atlas

02 / The learning lab

Better—for which goal?

“Better” sounds like a single destination. In research, it begins with a choice about what matters.

Directed evolution involves variation, selection or screening, and comparison. Different variants can have different properties; the criterion used to evaluate them shapes which ones are carried forward. Our simple exercise below focuses on one part of that reasoning: how a goal changes a ranking.

Different environments, different advantagesAI-generated illustration of variation and environment-dependent selection, not a record of experimental evolution.

Choose what you value

Imagine four abstract designs that differ in speed and consistency. Their scores are invented for teaching and are not biological measurements. Move the slider to change the importance of each property.

Learning station 02 / Illustrative scores only

Same options. A different definition of success.

ASpeed 95Consistency 30Equal-weight score: 62.5

BSpeed 76Consistency 72Equal-weight score: 74.0

CSpeed 50Consistency 93Equal-weight score: 71.5

DSpeed 35Consistency 99Equal-weight score: 67.0

At equal weights, B has the highest illustrative score.

Score = speed × its weight + consistency × the remaining weight. This example illustrates evaluation, not mutation, selection rounds, biological prediction or a completed evolution experiment.

What changed—and what did not?

The options stayed the same. Your evaluation changed. A fast option can become less attractive when consistency matters more. In a real project, the criterion should be tied to a meaningful question and justified before interpreting a result.

This is why a single impressive number is not enough. We need to know the baseline, how the measurement was made, the conditions and whether an advantage comes with a trade-off.

Visual guideThe reasoning behind an evolution cycle

1. 01

   Variation

   Begin with versions that differ.

2. 02

   Evaluation

   Use a clearly defined criterion.

3. 03

   Comparison

   Check against a reference under comparable conditions.

4. 04

   Learning

   Use the outcome to inform the next cycle.

A general learning framework, not a record of a completed evolution campaign by this team.

The bigger picture: variation → evaluation → learning

Idea	In plain language	Question to ask
Variation	There is more than one version to compare	What differs between the versions?
Selection or screening	Versions are evaluated against a criterion	Does the criterion match the intended goal?
Comparison	Performance is checked against a reference	Are the conditions and measurements comparable?
Learning	The outcome informs what happens next	What did we learn, including an unexpected or negative result?

Our current work focuses on design and engineering. We have not yet established a completed directed-evolution campaign. The evolution perspective explains what would be needed to make that claim.

A quick reflection

Which statement best supports an improvement claim?

Think about the comparison, the measurement and its conditions. Open the explanation below whenever you are ready.

Read the explanation

A defined comparison under stated conditions is the strongest of these statements. A visual impression cannot substitute for evidence, and a large number has no meaning without its units, baseline and uncertainty.

Take this conversation into a classroom

Ask learners to write their definition of “better” before showing the options. Compare choices across the group, then invite them to explain what they would measure. This exercise is intended for future teaching sessions and has not yet been used in our outreach programme.

Our education programme · Project story · Official directed-evolution learning resources