Mitch Harris

i think part of the problem is that we should stop teaching ANOVA/ANCOVA and just teach linear regression

I love Super Bowl statistics. (“No team that has failed to convert at least 35% of their 3rd downs has gone on to win” etc.) They are the best introduction to p-hacking that can be found outside major journals.

Smart people are too skeptical of AI and ML because the consultant class has hooked onto these terms to leech money from the credulous But never forget that we’ve solved computer vision and speech recognition in the last 5 years, and close to real time universal translation

Here's a visual proof that (2a+1)²=8Tₐ+1, where Tₐ is a triangle number Tₐ = 1+2+...+a

Seven design principles from Programming Pearls by Jon Bentley: 1. Work on the right problem. 2. Explore the design space of solutions. 3. Look at the data. 4. Use the back of the envelope. 5. Build prototypes. 6. Make trade-offs when you have to. 7. Keep it simple.

Who here realises the ENORMOUS privilege that the use of English as an international science & publishing language conjures on native English speakers in academia?

> just matrix operations Which is really just a fancy way of keeping track of things you count. Net takeaway: Excel is an AI.

https://twitter.com/neuroecology/status/904542555165089792 …

Me, standing outside Geoffrey Hinton’s office, matted hair, bloodshot eyes, shouting at passers-by: IT’S ALL A SCAM. THERE IS NO AI. NEURAL NETS ARE JUST MATRIX OPERATIONS. *sobbing softly as guards approach* they’re just matrix operations.

In my experience, if the paper is math heavy enough, it is typically properly reviewed and the reviewers check the math. If the paper is not primarily theory but have some cool math, reviewers almost never check the math.

Believe me - every single company that looks like they are crushing it is still a giant shitshow internally. That’s normal.

A videogame about making and releasing a videogame on time. Boss battles focus on appeasing entitled brats on Twitter.

I often feel as if I am forgetting something important. Thankfully, naps allow me to sleep through this worry. When I wake up, I usually forget that I may have forgotten something. I am a bear.

Let me just lay it out there. The problem with most hospitals is that they have no good structured way of evaluating/working/piloting w/startups. One person once told me "here's how it works. Big company CEO has dinner with hospital CEO and deal is done"

Lastly, hosp admins should stop blaming startups for not knowing their needs. Instead make it visible. Also, many hosps don't even know their actual needs. Announce a problem you're trying to fix + accept startup pitches to fix it. CMS innovation program does that.