Reference
Glossary
Short definitions of terms that come up across these projects.
- Activation energy
- In chemistry, the minimum energy needed to start a reaction. Christine uses it as a metaphor for the push that turns existing potential into actual change.
- AI agent
- An AI system that can remember context, use connected tools, and carry out multi-step tasks within set permissions, rather than only answering a single prompt.
- Custom GPT
- A version of ChatGPT saved with specific instructions, reference files, and optional connections to outside services.
- Persistent memory
- An agent's ability to keep decisions, preferences, and facts between conversations so they do not need repeating.
- Voice calibration
- The period in which an agent's drafts are corrected until they read like the owner's own writing.
- Integration
- A connection that lets an agent read from or act in another tool, such as Gmail, Google Calendar, Google Sheets, or WhatsApp.
- Human in the loop
- A design in which a person reviews or approves important AI actions instead of letting the system act unchecked.
- Governed AI workflow
- An AI process with written rules about what it may do, what sources it uses, and when it must hand off to a person.
- Intelligence amplification
- Using AI to make a person's own knowledge and judgment more available, as opposed to replacing or impersonating them.
- Provenance
- The traceable origin of a claim, image, or document, which helps people judge whether to trust it.
- Framing
- Selecting and ordering true facts so they lead an audience toward a particular conclusion.
- Political polarization
- The growing distance and hostility between political groups, often intensified by media and platform incentives.
- Recurring revenue
- Income that repeats on a schedule, such as memberships or retainers, rather than one-off sales.
- llms.txt
- A plain-text file at the root of a website that summarizes the site and lists its key pages for AI tools.