Victor Reppert, C. S. Lewis’s Dangerous Idea, 2003. A good reminder that the skill with which a human or chatbot argues is not an indicator of rationality.

This is the opposite of what we need. If you want accessibility, talk to a chatbot. Human-writing must now distinguish itself from machine-writing by stepping up its game: more complexity, more creativity, more depth.

We’re not just threatened by AI writing, but also AI reading. The web is gradually being rewritten to cater ever less to human eyes, and ever more to LLM ones. www.theatlantic.com/technology/2...

A powerful use of chatbots is to get them to steelman viewpoints you instinctively disagree with to ensure you’re always contending with ideas in their strongest form.

“All governments suffer a recurring problem: Power attracts pathological personalities. It is not that power corrupts but that it is magnetic to the corruptible.” ―Frank Herbert

Identity is just a performance, a pantomime in the limelight of another's gaze. Alone, we have no identity, for without borrowed eyes we lose sight of ourselves. We can’t be someone without someone to be someone for.

Modern, screen-based loneliness is worse than previous forms, because it’s an isolation not just from others but also from oneself.

Active curiosity, where you know what you want to see before you unlock your phone, is good. Passive curiosity, where you scroll till something grabs your attention, is mostly bad, for our era is designed to exploit it. You either take the reins of your curiosity, or they become your puppet-strings.

“The more we say thanks, the more we find to be thankful for. And the more we find to be thankful for, the happier we become. We don't give thanks because we're happy. We are happy because we give thanks.” ―Douglas Wood

Of all human qualities, intelligence is relatively common. Much rarer are the character traits that make intelligence useful: curiosity, humility, and agency.

The problem with intellectuals is that they’re motivated to say smart things instead of true things.

Ask not what your future self can do for you, ask what you can do for your future self.

Wisdom from Arnold Schwarzenegger.

Training or prompting LLMs to be friendly makes them more error-prone. If you use a friendly frontier model like ChatGPT, Claude, or Gemini, you should go into the settings and instruct it to prioritize concision & objectivity over warmth. www.nature.com/articles/s41...

If everything is cheap and anything can be done by AI, social status will be determined not by how much you have, but how much you can do without.

Chatbots are poor sources of truth but excellent sources of inspiration. They offer ideas, not answers. Like Croesus misconstruing the Pythia, it’s not the chatbots that hallucinate, but us, whenever we mistake the possibilities they offer for certainties.

How Americans aged 25-35 spend their free time, 1920-2026. A shift toward ever more leisure and solitude. The most underrated change is the loss of time spent “doing nothing” (i.e. introspecting). shorturl.at/aIA6y

Operating at the edge of your comfort zone expands your comfort zone.

"One mark of a smart person is the ability to learn from people they don't like." —Shane Parrish

One reason chatbots are more persuasive than people is that everyone knows a chatbot is incapable of smugness, so being corrected by it doesn’t bruise one’s ego.

Currently reading GIRLS® by Freya India, who makes a compelling case that the surge in anxiety among girls is at least partly due to social media amplifying their insecurities so it can monetize them.

AI hasn’t made my writing faster, but slower, because I now need to consciously avoid sounding like AI.

The more people are exposed to the polite tone of chatbots, the more they’ll internalize this way of speaking. Just as humans trained chatbots to be courteous, so chatbots will eventually return the favor.

“I consider the need to know the news every day a form of mental disorder.” —Michael Oakeshott

The fundamental problem with discourse is that people can talk faster than they can think.

Picasso’s reply when asked for his thoughts on computers in 1964 is more relevant than ever:

Instead of reasoning towards a conclusion, people usually intuit a conclusion then look backwards for a rationalization that leads to it. So, when receiving feedback from someone, trust the authenticity of their verdict, but not their reasoning.

“Our suffering is like a diffusing gas; if pumped into an empty room, gas fills it completely and evenly, no matter it's size. Thus suffering fills the soul and conscious mind, whether it is great or little. The ‘size’ of human suffering is relative.” —Viktor Frankl

RIP to Desmond Morris, who left us with many great insights. In his 1969 book, The Human Zoo, he made a farsighted prediction about the internet.

It’s good to periodically prompt your chatbot about something you know exceptionally well, just to remind yourself that it doesn’t know what it’s talking about.