TThe Diary of a CEO
← All frameworks
Innovation

Humane Technology Alignment

Design technology around human and societal vulnerabilities

Difficulty
Advanced
Time to result
~ongoing to results
Steps
5
Confidence
96%

Harris extends the idea of ergonomic design from the body to the mind and society. A chair should fit the spine; humane technology should fit human needs and account for human vulnerabilities. The analysis must move beyond whether an interface is easy to use. Teams identify the developmental, relational, psychological, and civic systems their product enters, then design defaults that strengthen rather than exploit those systems. Harris applies this to children's development, dignity at work, and the information environment. The model also assigns responsibility: once software mediates a social life-support system, its designers cannot treat effects on that system as somebody else's problem. Success therefore includes human outcomes, not only adoption, usage, or technical capability.

Origin

Extracted from The Diary of a CEO

Core principles

  • 01Technology should fit human needs as ergonomics fits the body
  • 02Human vulnerabilities are design requirements, not user failures
  • 03Software that enters a social system inherits duties toward that system
  • 04A useful feature can still be inhumane at societal scale

How to run it

  1. 1

    Define the human purpose

    State the human need or social capability the technology is meant to support.

    Pro tip Describe the desired human outcome without naming the product feature.

  2. 2

    Map vulnerabilities

    Identify cognitive, emotional, developmental, relational, and institutional weaknesses the product could exploit or amplify.

    Pro tip Include vulnerable users and people indirectly affected by the product.

    Watch out Do not treat predictable human responses as user error.

  3. 3

    Trace system effects

    Examine how repeated use changes relationships, dignity, development, and shared information beyond one session.

    Watch out Usability testing alone may miss cumulative social effects.

  4. 4

    Build protective defaults

    Make the safe, dignity-preserving behavior the normal product path rather than an optional setting.

    Pro tip Prefer designs that reconnect users to real people and institutions where appropriate.

  5. 5

    Measure human alignment

    Track whether the product strengthens the named human outcome and revise incentives or features when it does not.

    Pro tip Pair usage metrics with evidence about wellbeing, development, and autonomy.

    Watch out High engagement is not evidence of humane impact.

In the wild

Narrow therapy support

Harris proposes bounded AI support that performs a defined exercise, such as cognitive behavioral prompts, and directs a user toward family or a qualified therapist rather than claiming intimacy or trying to become the primary relationship. He presents this as a safer design direction, not validated clinical advice.

The product supports a limited task without deliberately deepening attachment to an anthropomorphic system.

Illustrative child-learning app

A team evaluates not only lesson completion but whether the app supports attention, age-appropriate development, and real-world relationships. It removes mechanics that reward prolonged use and adds clear handoffs to teachers and caregivers.

The design optimizes for learning and development rather than time spent in the app.

Common mistakes

Equating easy use with humane design

An intuitive interface can still exploit attention, attachment, or social vulnerabilities.

Measuring engagement as wellbeing

More frequent or longer use may indicate dependency rather than a better human outcome.

Is it for you?

Best for

Teams designing AI, platforms, or digital products that shape attention, relationships, development, or public information.

Not ideal for

Low-impact tools with no meaningful interaction with human behavior or social systems.

From the transcript

a humane technology like the Macintosh that works with the ergonomics of your mind

Tristan Harris · 1:19:00

Technology has to serve and be aligned with human dignity

Tristan Harris · 1:19:30

From the episode

AI Expert: We Have 2 Years Before Everything Changes! We Need To Start Protesting! - Tristan Harris