Do not start with fundamentals. This is an awful approach to learning. Start with so-called "advanced" topics and ask questions until every term/concept is understood. This is the correct, rigorous, scientific way to learn, because the advanced topics are embedded in larger,.

RT @sean_a_mcclure: You should do a lot of reading to understand math. You cannot understand a thing by staring at it. You have to look at….

There will always be someone at the right place and time to be labeled by history as the originator or genius. Given the inherent complexity of society, the sequence of events/insights we view in hindsight of the discovery are statistically guaranteed to have occurred in the.

*Luck* is why we must listen to “leading figures” in a field who have no idea what they’re talking about.

RT @sean_a_mcclure: @pmddomingos 1. Suggesting that what emerges in an LLM is still “statistics” is pure pseudoscience. 2. There is no suc….

“How long is that going to take?”. Fool: “about X amount of time”. Smart: “I don’t know”. Wise: “I don’t care”.

Even the smart lucky ones are usually still too dumb to realize they’re lucky.

There’s a good chance you’re not going to be upset about it later. When 𝙣𝙤𝙬 looks to rob you of a peaceful mind, exist in your 𝒍𝙖𝒕𝙚𝒓.

The worst way to test an AI’s intelligence is to use textbook word problems. AI should be expected to struggle to reduce its inherent complexity down to the narrowly-defined and extremely unrealistic scenarios of such challenges. A thing isn’t “struggling with reasoning” when.

Paperbacks for giveaway arrived. Shipping tomorrow.

More people slowly realizing my point about a life’s pattern not being something you engineer. If you notice intermittent behavior in healthy individuals (e.g. bursts of intensity with longer rest times) the pattern you noticed is correct, BUT the choice to design your life by.

Often, the only way you lose is if you play the game to begin with.

The notion of “role” in science does not extend past the point of normalcy. If someone has a deficiency, then a single isolated ingredient can fix them. But it is absolutely untrue that a single ingredient (e.g. supplement) will improve someone’s health beyond a normal.

Memorize 52 cards. You can attached the number image to your locations, and the suit to 1 of 4 facets of those locations (4 orthogonal directions). But this introduces a two-step recall process. Facets become cognitively overloaded. Instead, make the full information of a.

People who focus on scholastic or competitive endeavors optimize for speed, accuracy, mental/physical stamina, strategy. These do not extend to the real world. These all use a set of constraints, goals, and feedback loops that are highly unnatural. Simply, they don't survive.

People will observe a dynamic system, take note of its mercurial nature, and get frustrated that they can’t land on a consistent recipe to get the outcomes they want. Just because the system appears erratic doesn’t mean you have no control. The point is to continually change.

Don’t “teach”, just tell people to show up and watch you do what you do.

People who take longer to say they understand something are usually the ones asking deeper questions, reconciling more pieces in their mind. There are others who *seem* to grasp something immediately, but upon further questioning demonstrate scant comprehension. There is a.

Evidence has become the diluted knockoff of science. Anything can be “evidence” as long it has been published. Publishing is an absolute disaster in current science. We all know the usual replication crisis, deep bias, p-hacking, file drawer problem, etc. But the fundamental.

The competition version of a thing is always the dumb version of a thing.

In order for an academic test question to have a single correct answer, the real world must be parameterized (constrained) so severely that the scenario has virtually nothing to do with the natural world. Reducing the infinite and messy richness of existence into simplistic and.