If you have ever done some Mathematics or Data Science, then last name Hamming should be familiar (Hamming distance). I found this book through Stripe Press, an amazing list of books labeled as Ideas for Progress.

• Education is what, when, and why to do things. Training is how to do it.
• In science if you know what you are doing, you should not be doing it. In engineering if you do not know what you are doing, you should not be doing it.
• Is programming closer to novel writing than it is to classical engineering? Yes.
• Think before you write the program.

The idea that the modern life is too comforting and it is beneficial to go out of comfort zone both mentally and physically: "Embrace Discomfort to Reclaim Your Wild, Happy, Healthy Self".

I especially liked idea of importance of nature in life. In the Japanese movie "Perfect Days" (2023), the main character, who works as a public toilet cleaner in Tokyo, every day spends time in the park, sitting among the trees.

• In today's economy, where people can't detach from work emails, nearly a quarter to half of all employees say they're burned out. Nature may be the best recovery tool for the condition.
• Having plants in your office can increase your productivity.

Another great thought was related to curiosity.

• Creating impressions in your scrapbook. If you are seeing and doing all the same things over and over, your scrapbook looks pretty empty when you take inventory of your life. So we need to do more novel things to start creating more impressions in our scrapbooks, so we don't feel like the years are flying by.

A concept of a "technopoly", which is defined as a society where technology is king and culture seeks its authorization, satisfaction, and direction from technology.

Apparently some tutor named William Farish from Cambridge University in 1792 pioneered to grade students' papers. A little idea that changed the way we live and see things now:

• Yet his idea that a quantitative value should be assigned to human thoughts was a major step toward constructing a mathematical concept of reality. If a number can be given to the quality of thought, then a number can be given to the qualities of mercy, love, hate, beauty, creativity, intelligence, even sanity itself.
• To a man with a pencil, everything looks like a list. To a man with a camera, everything looks like an image. To a man with with a computer, everything looks like data. And to a man with a grade sheet, everything looks like a number.
• A new technology does not add or subtract something. It changes everything.
• New technologies alter the structure of our interests: the things we think about. They alter the character of our symbols: the things we think with. And they alter the nature of community: the arena in which thoughts develop.

A great book on how to design conversational bots.

I found it practical and used it to design the tone of voice and personality of Maria, the first AI agent at Orbio AI.

Entire book is worth reading to understand the ideas behind, but shortly here is the framework to apply:

1. Interaction goals - personality serves the JBTD
2. Level of personification - how close to a person must the system be?
3. Power dynamics - bot acts as who compared to the user?
4. Character traits - max 4 to support interaction goals
5. Tone - what tone will convey the character traits?
6. Key behaviors - describe how bot should behave under key situations in ways that consistent with personality

From my experience building conversational AI agent, the most influential at the end are (1) what is the goal, and (6) key behaviors.

Once you know what objective the conversational AI agent needs to achieve with the user (e.g. run interviews), then the rest are decided pretty quickly. In (6), beyond thinking of possible scenarios, is where the tone and character traits are specified in a practical way that you can add them to LLM system prompt as examples.

Uff, parenting.

I have two kids, both toddlers. Before having them I have read a lot of books on parenting, first around what to do during the first 6 months, then the first year, then the first 3 years. My kids now are older than 3 and I haven't read a book about parenting kids older than 3. So I got this book, recommended by Naval.

Over time I learned that good parenting is extremely cultural, there is no single way of doing it. I have my own opinions and I don't fully agree with the book, but I found this book convincing with some great points that made me rethink again.

Surprisingly, I found much more interesting the last few chapters of the book, where the author explained the core philosophical ideas behind his parenting method. A particular phrase I liked a lot was Replace "find yourself" with "develop your tastes and preferences".

tl;dr

• Knowledge growth is coming from guesswork or conjecture and internal criticism if the guesswork solves or explains.
• Learning is a sovereign act. Everyone, including kids, has a chance to come up with good conjecture, so treat kids seriously.
• Conjecture generation requires freedom, which means no authority and that implies no rules enforced on kids from parents (authority). Let the kids figure out everything.
• Parenting is an act of teaching kids how to be problem solvers and that means letting them practice critical rationalism and knowledge generation themselves via own conjecture.