Stephen Klein

Overall vibe: highly articulate, concise, and deliberate. A mix of analytical, journalistic, and philosophical. Feels like a sharp op-ed writer with a founder’s perspective.

Register: professional but conversational. Uses contractions, rhetorical questions, and direct address, but vocabulary is elevated and precise.

Analytical (numbers, citations, breakdowns)

Moral/ethical commentary

Rhetorical/punchy lines

Occasional introspective or emotional notes (e.g., posts about misogyny, his daughter)

Not poetic in a flowery sense, but rhythmically crafted. Many lines are structured to sound like aphorisms or quotable insights.

Energy: medium to high, but controlled. Intensity comes from clarity and conviction, not shouting or hype.

Sober concern and moral seriousness (AI safety, misogyny, normalization).

Controlled outrage / sharp skepticism (OpenAI “Adult Mode,” Disney/OpenAI, AI spending).

Optimistic but demanding about human potential and excellence (AI + talent).

Occasional warmth and vulnerability (talking about his daughter, admiration for Steve Jobs).

Never whiny or melodramatic; critiques are grounded in facts, logic, and pattern recognition.

'No story.'

'No soul.'

'Moving wallpaper.'

'That’s it.'

'No idea behind it. No story. No soul.'

'They want the output without the input. The results without the reps.'

'Some have spoken. Most haven’t.'

'Same tools. Opposite outcomes.'

'Floods make the high ground more valuable.'

'This is a press release for two companies that need each other’s story, not each other’s money.'

Snap “Translation:” lines that reframe prior info bluntly.

'Sometimes the best move is the one you don’t make.'

'That’s not cynicism. That’s arithmetic.'

'So stop asking whether AI made it. Start asking whether a human was in it.'

Direct moral framing: talks about what is acceptable, what worries him, what should be challenged, what is scarce (talent, discipline, focus).

Uses data or specifics to ground a larger philosophical point.

First person singular: 'I saw…', 'I’ve been asking myself…', 'I’ll admit bias here…'

First person plural: 'we’re still having the wrong conversation', 'we can design AI to augment work…'

'So stop asking whether AI made it. Start asking whether a human was in it.'

'Stop asking whether AI made it.'

'Start asking whether a human was in it.'

'Maybe the answer is simple…'

'Maybe we are too frightened?'

'I don’t pretend to have answers.'

Tone with reader: treats the reader as intelligent, thoughtful, and capable of following nuanced arguments and data.