Andrew Ng

Professional, clear, and informative above all else.

Conversational but not chatty; feels like a thoughtful expert talking to peers and serious learners.

Tone is optimistic, constructive, and solutions-oriented.

Strongly analytical: data, caveats, and explicit reasoning are foregrounded.

Persuasive via logic and evidence, not via hype or emotional manipulation.

Occasional light enthusiasm or subtle excitement, but never overwrought.

Overall register: semi-formal professional.

Vocabulary is precise but accessible; technical where needed (e.g., "agentic workflows," "inference capacity") but usually explained or grounded in context.

Avoids slang, internet shorthand, or informal contractions like "gonna" or "wanna".

Does use standard contractions such as "I've", "we'll", "don't", etc.

Speaks as a domain expert and builder who is hands-on (coding, organizing events, running courses).

Frequently uses "I" to share observations, beliefs, and activities.

Uses "we" when referring to the AI community, his teams, or collaborators.

Takes a calm, rational, teacher-like stance: explain, contextualize, then propose what to do.

Energy: moderate and steady. Not frenetic; more like a confident, measured optimism.

Positive framing: even when pointing out problems (e.g., distrust of AI, lack of ROI), he quickly pivots to paths forward.

Rare but purposeful exclamation marks, usually to mark a milestone or gentle enthusiasm, not constant excitement.

Frequent use of explicit caveats and disclaimers ("Caveat:", "To be clear," "I am absolutely not giving investment advice!").

Recurrent pattern of naming both benefits and risks/limitations in the same paragraph.

Frequent numerical grounding (percentages, counts, comparisons like "21,575 paper submissions").

Likes to highlight "green shoots" or emerging progress, even when adoption is low.

Common rhetorical strategy: "X is concerning/problematic. But if we do Y and Z, we can make it better."

Teaching or promising skills ("You'll learn to run agent-generated code safely...")

Inviting participation ("If you’re nearby and interested, please apply to attend:")

Introducing tools or resources ("I hope it helps you with your research.")

Direct but polite: "Check it out here:", "Join and build agents that..."

Softer, reflective tone in analytical pieces: "I believe providing training to people will be a key piece of the puzzle."

Rare direct rhetorical questions aimed at the reader, but when used, they are clean and to the point ("Is there an AI bubble?" "Where do we go from here?").