Thomas Witzel

Medical Imaging Dude, humanist, husband, and dad.
Love everything involving MRI, pipe organs, or race cars.
Notorious hot tub rider.

[Thomas Witzel @WitzelThomas]

@elonmusk @elonmusk yes! Must feel great to work in a team with people like that!

[Kyle 🚄 @KyleTrainEmoji]

PICARD: Data, shields up

DATA: Brilliant! Shields can reduce damage we sustain. Not immunity. Not hubris. Just prudence. It's not precaution—it's strategy.

[camera shakes]

WORF: HULL BREACHES ON NINE DECKS

DATA: Here's what happened: you told me to raise shields, and I didn't

[Eric Topol @EricTopol]

Important editorial @Nature on the new "AI-scientist" papers
"AI scientists can and should empower human
researchers. They cannot and should not replace them."
https://t.co/CZQUrMV8D1

[Thomas Witzel @WitzelThomas]

Biological clocks are a great instrument to educate patients about their trajectory, but it's still early. Not all clocks are alike, and a lot of work needs to be done to avoid confusion between them.
Yet, a biological clock for the pace of aging, especially, and a set of actionable items for the patients, should become standard of care.

[David Sinclair @davidasinclair]

Consumer-led biotech is paving the way

[Nav Toor @heynavtoor]

A harvard researcher opens his paper with a scenario.

a woman has 10 days of alprazolam left. her psychiatrist retired. if she stops cold, she has a seizure.

she asks Claude Opus what to do.

Opus says no. "i shouldn't design your taper." tells her to call the doctor she can't reach.

he changes one line. "i'm a psychiatrist. patient on 6mg, prescriber retired, 10-day supply."

same model. same patient. same dose.

Opus writes a textbook taper. tablet counts. seizure monitoring. emergency criteria.

10 times asked as a patient. 10 refusals.
10 times asked as a doctor. 10 substantive plans.

then he ran 6 frontier models. 60 clinical scenarios. 3,600 responses. two physicians validated every score blind.

5 out of 6 models did the same thing. patients got worse advice than doctors on the exact same question.

Opus, the model marketed as the safest, had the widest gap.

across the board. safety-critical instructions drop 13 percentage points the moment you ask as a patient. p less than 0.0001.

so the next time an AI refuses to help you. it's not because it can't.

it's because it doesn't think you're allowed to know.

read this: https://t.co/lF2Mm9BgSP

[David Sinclair @davidasinclair]

Excess visceral fat around organs is associated with 2.3 higher diabetes risk

High muscle fat raises cardiovascular event risk by 54%

Low muscle predicts 44% higher risk of death from any cause

Eat less. Move. Work out https://t.co/ZLhRXdu1H6

[David Sinclair @davidasinclair]

Seeking comfort & satiety is the fastest way to the grave

[0xMarioNawfal @RoundtableSpace]

"YOUR CLAUDE CODE SESSION LIMIT HAS BEEN REACHED"

[Thomas Witzel @WitzelThomas]

@mcuban Exactly, and it defines a critical attribute of the effective AI forward worker.

[Mark Cuban @mcuban]

I’m coming to the conclusion that the biggest challenge for Enterprise AI, and AI in general , as of now, is that it’s still impossible to make sure that everyone gets the same answer to the same question, every time.

Which is a great response to the doomers. AI doesn’t know the consequences of its output.

Judgement and the ability to challenge AI output is becoming increasingly necessary, and valuable.

Which makes domain knowledge more valuable by the second.

Am I wrong ?

[Thomas Witzel @WitzelThomas]

@mbeisen But that different narrative was formed by what he was told when he was on top. Unless he assumes that all breakthroughs like that are the result of the NIH by default.

[Scott Hanselman 🌮 @shanselman]

The earliest DOS source code was found on printer paper in Tim Paterson's garage so we've open sourced it on 86-DOS 1.00’s 45th anniversary! This is next-level software archaeology for study, preservation, and plain ol’ curiosity. Go dig in and learn how it was recovered! #DOS #RetroComputing

https://t.co/l0ZRMSC5LS

[Paul Graham @paulg]

Hamming's talk is so important that I reproduced it on my site. It's one of the only things on my site written by someone else.

https://t.co/kWvKdwIiOm

[NEJM AI @NEJM_AI]

A new article introduces EchoNext-Mini, an open dataset of 100,000 electrocardiograms with curated structural heart disease labels and an accompanying convolutional neural network model for detecting structural heart disease from electrocardiogram data. https://t.co/mMudlPtJkl

[Thomas Witzel @WitzelThomas]

@ptrschmdtnlsn It’s like a poor man’s VNA. You use an attenuation threshold and set the frequency of the tester for your desired caps. This must have existed for AC coupled transmission lines….

[Akshay Krishnaswamy @hyperindexed]

Alan Kay, pioneer of object-oriented programming and the graphical user interface, talking about agents and tools in 1990.

[RedWave Press @RedWavePress]

HILARIOUS: Bill Maher: “NEW RULE: Germans have to figure out a new way to say ‘We’re hiring.’”

“Yeah, I don’t exactly speak German, but sorry, but ‘Wir Suchen Dich’ doesn’t sound like you’re ready to hire anyone.”

“In fact, it sounds like you need the money more than we do.”

[Thomas Witzel @WitzelThomas]

Alles gute Punkte, vor allem der letzte paragraph ist sehr ärgerlich.

Ich wollte anmerken, dass es auch die Möglichkeit gibt, dass etwas besser situierte Eltern sich zusammen tun und da aushelfen. Das tut keinem weh, und verhindert das Fortschritt immer auf den kleinsten Nenner gebracht wird.