We’re all slowly realising that global state was never evil all along…

On death of a loved one: if you’re lucky, it will illuminate how truly considerate people can be. It can bring people together in pursuit of reducing shared suffering by even the tiniest degree. It encourages us all to cultivate and cherish human connections in the here and now.

The key piece of the puzzle is IQ is a scale that is specifically manufactured to be normed around a mean of 100 (with median usually close to 100) while also exhibiting bell-shaped form up to a couple of standard deviations either side of the mean. Crucially this involves monotonic but non-linear rescaling of test score distributions that are often already asymmetric or fat-tailed (this is what @ThePrimeagen is alluding to here). Scores are non-linearly warped to create roughly symmetrical mass about the mean while preserving rank. This geometry is by design, not a discovery! Therefore everything we’re tempted to infer from IQ models at the tails (±3 SDs and beyond) becomes increasingly unreliable, and left and right tails are differentially unreliable due to non-linearity and floor/ceiling effects. Yet, entire careers have been made splashing around in these tails like they are truth preserving. The lesson is to not take everything in the IQ literature totally seriously, while also respecting its predictive power up to roughly ±2-3 SDs from 100. I welcome @nntaleb to utterly eviscerate me if I got anything wrong here.

what’s up with toothpaste nowadays turning raspberry ripple when u spit

If you lived forever, you'd look back fondly on every era. Don't let yearning tinge your nostalgia. Now is the past you'll long for.

Studying higher mathematics and comparing GPT 5 vs 4o, I'm averse to 4o's sycophancy, but I find myself engaging more with the material by writing about my understanding with 4o. It's the difference between a smart, terse PA and a co-researcher. This distinction is underexplored.

Does anyone know if we get penalised to a different degree for utterly silly mistakes like this? 😬 @_MathAcademy_

I'm learning integration by substitution in MF III. I'm unsure of myself as I work through it, but I'm consistently getting the right answer. Usually if I'm unsure, I'll make mistakes and be punished harshly and course correct (I love this about MA). Anyone had this? Any ideas?

After a lackluster week, I really had to put in the work today to make it back into Ruby. First time ever doing two quizzes in one day! 😮

MA quizzes used to spook me every time they showed up. I’ve psyop’d myself to love them: 1) it’s the only way to earn XP while knowing exactly how long it will take, and 2) do them now; you’ll be excited to fix your failures in the reviews tomorrow.

I have a working familiarity with calculus but holes in other parts of my math foundations. It's been so exciting and impressive going through MF II and experiencing how delicately MA sneaks up on calculus, from average rates of change, to limits at points, to limits at infinity.

Let's go @gabemays! That last minute Sunday sesh (another one after logging on @xtreeks) was clutch! 🫡

Shout out to everyone at @EpicGames who made it so simple to make custom C++ IKRig solvers that integrate nicely into the editor out of the box. Amazing work all. CC: @kiaran_ritchie

What we've been working on. A while in the making! The @GreatApeGames team absolutely smashed it with this trailer! 🙌