Shannon showed that any communication channel has a maximum capacity (measured in bits per second) and that it’s possible to transmit data reliably at rates approaching this limit using appropriate encoding schemes.
The paper deliberately sidesteps the question of meaning, focusing only on the statistical properties of messages. This was both its greatest strength (making the math tractable) and its greatest limitation (leaving semantics for future generations to figure out).
Links:
]]>Reading this tweet requires you to sign up for our constant surveillance of what you see, when you see it, how many seconds you keep gazing at it, who you talk with, and what kinds of things can sell you better ads.
That content is datagated.
]]>See also: indieweb
]]>For example, lack of money leads to unhappiness, but more money beyond a certain point doesn’t lead to more happiness—it just reduces unhappiness.
]]>Challenges include privacy concerns and usability issues, but the approach opens new possibilities for accessibility and inclusivity in design.
]]>See also: posse
]]>We already live in information saturation. Personalized AI may be necessary for a healthier relationship with information.
]]>Their signature question is “what job is the customer trying to get done?”, instead of “who is our customer?”. Another way of phrasing it is “What is our product getting hired for?”
]]>The principle connects information theory to thermodynamics: information is physical, and manipulating it has thermodynamic consequences.
Links:
]]>Inflation is a classic example: reporting “5% inflation” hides the vector of price changes across categories weighted by their importance to different people. My inflation is not your inflation, depending on whether you spend more on food, fuel, or housing.
Star ratings suffer the same problem. A 3.5 star average could represent uniform mediocrity or violent disagreement: half loving it, half hating it. The number destroys the distribution it summarizes.
]]>Malleable Systems promote software that is as adaptable as it is functional.
]]>Instead of posting directly to Twitter, Instagram, or Medium, you first publish on your own domain, then automatically (or manually) cross-post to silos. Your site becomes the canonical source; the platforms become distribution channels.
Benefits:
The inverse pattern is PESOS (Publish Elsewhere, Syndicate to Own Site), where you post to platforms first and aggregate copies back home. Less ideal but pragmatic when platforms don’t support outbound syndication.
]]>Links:
]]>Instead, they mostly serve other entities’ interests.
]]>This gap exposes the limits of A Mathematical Theory of Communication: Shannon deliberately sidestepped meaning because it was “irrelevant to the engineering problem.” But for humans, meaning is the entire point. Data without context is noise; the same bit pattern can be life-changing or worthless depending on the observer.
We lack formal vocabulary to distinguish meaningful information from mere data. This might be why we struggle to define consciousness, creativity, and understanding in mathematical terms.
]]>Surveillance capitalism treats human experience as free raw material to be extracted, processed into behavioral data, and sold as “prediction products” in behavioral futures markets. Companies like Google and Facebook pioneered this model: your attention, your clicks, your location, your relationships—all harvested and refined into predictions about what you’ll do next, what you’ll buy, who you’ll vote for.
While industrial capitalism exploited nature, surveillance capitalism exploits human nature itself.
The key insight is that this isn’t just advertising—it’s a new logic of accumulation. The goal isn’t merely to know you, but to shape you. The more precisely your behavior can be predicted, the more valuable that prediction becomes. The endpoint is not just surveillance but modification: nudging, prompting, herding human behavior at scale.
What makes it especially insidious is how it operates outside traditional capitalist critique. These companies still maximize profit and pursue growth—they just do it by claiming ownership over the intimate patterns of human life, with remarkably little legal constraint.
See also: Datagated, Personal Computing
]]>Taoists view “naming” and “counting” the ten thousand things as a departure from the undifferentiated Tao. While practical for navigation, this categorization creates artificial distinctions that obscure the underlying unity of existence.
The concept warns against mistaking our mental models (categories, numbers, names) for reality itself.
Links:
]]>You cannot know if someone has seen your passwords or private keys. You cannot verify that data has truly been deleted; only that it’s been made harder to find. The bits may persist on a disk until overwritten; copies may exist on backup tapes, in memory dumps, in someone else’s notes.
This is a fundamental limit on privacy and data ownership: encryption can make data unreadable, but cannot prove that no copy exists elsewhere.
]]>You can’t know whether you are the only entity that knows a secret. You can’t know if it is written anywhere else in the universe. Maybe it’s in a different language. Maybe it was encoded in the DNA of an Escherichia coli.
You will never know for sure if anyone else knows your private keys.
See also: universal_unsearchability
]]>Wigner considered this “unreasonable” because there’s no obvious reason why abstract mathematical structures invented by humans should correspond so well to the laws governing the universe. He called it “a wonderful gift which we neither understand nor deserve.”
The essay raises deep philosophical questions about the relationship between mathematics, physics, and the nature of reality.
Links:
]]>