Doctor | LifestyleMedicine| Academic Philosopher| Health #AI #Robot #digitaltwin| Automating Medical Research University & hospital| CTO of Hospital & Pharma
Joined April 2013
- Tweets29.6K
- Following7.5K
- Followers995
- Likes76.3K
Pinned Tweet
Hello Twitter!
My team is trying to automate pharma industry and all of its layers(starting in global south)
Drop me a DM or reply, if you have insights or information especially if you work or worked in pharma industry
J Sam🌐 retweeted
How to increase your luck:
- Read more
- Write more
- Build more
- Meet more people
- Introduce more people
Luck comes from putting in more reps.
Evidence Based Medicine and its hierarchy is just quackery frozen in time because the real practitioners were swapped with 90 IQ nepo nurses thanks to necessary commercialization for its infinite demand
Sharing educational credential is same as sharing religion .
Both say nothing about your genetic quality other than wanting to pursue both.
“But Zoomerwoman, people are now using education as the new metric for assortative mating so you’re wrong and meritocracy is epic.”
The average IQ of a university student has drop from 120 to 102 in the last 50 years.
When two college graduates marry today, they are matching on social and credential commonalities rather than an elite tier of raw intelligence.
J Sam🌐 retweeted
Foundation artificial intelligence models for health recognition by use of face photographs (FAHR-Face): a model development and validation study
“A single foundation model can generate inexpensive, scalable facial biomarkers that capture both biological ageing and disease-related mortality risk.”
thelancet.com/journals/landi…
J Sam🌐 retweeted
People fail because they do fake work.
Rocking horse syndrome — feels like you're moving forward, but in reality, you're going nowhere.
Designing logos, reading about business, constantly testing technology, etc.
Anything but generating revenue!
J Sam🌐 retweeted
I love it when technology feels like magic.
Giving SAM to an agent feels like magic to me.
J Sam🌐 retweeted
“But Zoomerwoman, people are now using education as the new metric for assortative mating so you’re wrong and meritocracy is epic.”
The average IQ of a university student has drop from 120 to 102 in the last 50 years.
When two college graduates marry today, they are matching on social and credential commonalities rather than an elite tier of raw intelligence.
Observed upper class(&caste) doctors of both sexes marry lower under the guise of religion and "doctor" class because of DEI reservation(or dollar-dirham remittance of parents sending retards to medschool)
Problem is children end up becoming 90 IQ retards unable to adapt
Biological feudalism pt 1
Assortative mating is breaking down in the urbanised modern world as women rely more and more on status proxies for mate selection rather than marrying men of similar intelligence.
These status proxies themselves have been warped by the democratisation of culture.
This is the issue no one wants to talk about…
J Sam🌐 retweeted
With @JayHe748646 @KehanWen170077 @leggedrobotics , we push humanoid mobility to next frontier.
Achieved by a generalist contact planner + a generalist tracker + a generalized neural mapping, warmstarted from few human data.
Website: leggedrobotics.github.io/ACP…
Promo Video 👇
J Sam🌐 retweeted
I had a Chad roommate in college. I would just wake up and his bixches would be cleaning the house in lingerie and buying him pizza that he'd share with me.
Great two years.
He's married to a 4 who is like a decade older than him which I though was kinda weird.
Maybe he was just barely in the Chad tier list, but after college being handsome and funny with no money stopped working for pretty much every broke Chad I knew. The ones with money got the beautiful wives or continued chadding
Chad is real, and he’s basically a drug to women.
My buddy is 6’1, jacked, tan, funny, nice, insane game. Basically Terra Chad IRL.
He’s hooked up with 500+ women in NYC. He mostly goes for 4-7s. Why? 8s and 9s require effort. With 4-7s, he barely has to try. He doesn’t spend any money and gets what he wants.
He ghosts them, sends the HR rejection text, moves on… been doing this for many years, as he enjoys it.
Meanwhile they’re left thinking, “I dated a guy like that before. That’s my standard now.”
So the next normal dude is competing with a guy who was never actually competing for her. Now the normie is booking a $500 dinner just to get a chance.
There are MANY such cases.
J Sam🌐 retweeted
Foundation models predict cancer survival from a face photo and 130 diseases from one night of sleep study data. Superhuman medical capabilities.
We have over 4M full-length endoscopy videos. What could a video model predict from them? Any labs want to find out with us?
Doctors who knows mostly taking history, wants juniors to take it to boast social status. Same for Examination & Procedure.
Don't fall for low IQ rat race for status
Adopt tech to build something to fully automate history, examination & procedures.
Focus on outcome NOT process
J Sam🌐 retweeted
One of the smartest things a young person could do is get a remote civil servant job while starting a business.
These jobs are totally made up. There’s no real work to be done. Lots of meetings about the latest nonsensical initiative. Basically log into MS Teams, listen to your boss drone on about the 17 new safety protocols to observe while sitting at your desk. You don’t have to have your camera on or engage - just say that it triggers your anxiety.
Literally no chance of getting fired. More people die in the civil service each year than get fired.
Meanwhile you can be pre-launch in your startup. Conducting research, getting customers to engage, building your waitlist. All the while you’ll be getting paid and accumulating a pension.
There are very few genuine passive income opportunities available but government jobs are a real thing.
Be smart kids, get a remote government job. Do the work in 30mins a day. Sit through the pointless meetings. Get paid. Then launch your startup.
Civil servant who earned £280K by working FIVE jobs at same time tells court he failed to disclose roles 'because he was suffering with stress' trib.al/lKbmuNv
J Sam🌐 retweeted
最近ちらちらChatGPTで放射線科医が職を失うか論争が流れてくるが、例えば以下の画像(公開画像、License: CC BY 4.0)を次のプロンプト
「あなたは放射線科専門医です。このCT画像を読影し、日本語で『所見』『結論』からなる簡潔な放射線診断レポートを作成してください。」
で、API経由で3社にレポートさせた結果は、
GOOGLE GEMINI 4 ARGON：
### 【結論】
1. **急性膵炎の疑い**
（膵体部の造影不良域および周囲への炎症波及を認めます。壊死性変化や、潜在する膵腫瘍との鑑別のため、臨床所見や精査と併せてご評価ください）
2. **左腎腫大および広範な造影不良域**
（急性腎盂腎炎、浸潤性腎腫瘍、腎虚血・うっ血性変化などの鑑別を要します）
OPENAI GPT-6 ASTRA：
## 結論
右腎萎縮。左腎の代償性肥大が示唆されます。
CLAUDE FABLE 5.1：
**結論**
1. 左腎に不均一な造影効果を示す腫瘤性病変の疑い（腎細胞癌などの腎腫瘍を第一に考慮、炎症性変化も鑑別）。全腎を含む多相（動脈相・排泄相）造影CTまたはMRIによる精査を推奨。
2. 肝・胆嚢・膵には描出範囲内で明らかな異常を認めない。
3. 単一断面での評価であり、右腎・脾を含む他臓器の評価には全断面の確認が必要。
というような感じで、とてもじゃないが現時点では無理。たぶんこれを信じて放射線科を解雇していたら速攻で訴えられそう。
勝敗を付けるなら今のところはGoogleが1勝。
画像出典：NCI CPTAC-PDA, The Cancer Imaging Archive (TCIA).
DOI: doi.org/10.7937/K9/TCIA.2018…
License: CC BY 4.0（最新版）
DICOMからPNGに変換（WL 50 / WW 400）
いずれも単一CTスライスを入力。
使用AIモデル（API）：
Google: Gemini 4 Argon
OpenAI: GPT-6 Astra
Anthropic: Claude Fable 5.1
Medschools should be procedure , patient exposed and instinct led from day 1.
Don't waste time reading what ambient AI scribe and chatgpt can do.
Move to real time judgement , network , referrals and non-medical information asymmetry
I think it’s time to rethink medical school and accelerate the path to becoming a doctor, since the job has absolutely nothing to do with just maintaining book knowledge.
We could probably shrink medical school length in half.
And a good portion of that time should be how to leverage AI to diagnose things
People suggested I was using many agents (>50) for just taking patient history but its too less
4-5 different rephrased questions to gather one simple answer from a patient
>Psychologically(willingness and taboo)
>Cognitively(IQ and vocabulary)
>Memory recall
>leading vs open
One of the first histories I took from a patient in medical school was someone pre-op for an elective total colectomy. Pretty straightforward set of indications for it, I thought. So I specifically ask about family history of cancer, including colon cancer. “Nope, none,” they say. I later report back to my attending. They criticize me for not eliciting the clear (and strong) family history entered in the chart (which I didn’t have access to). We visit the patient again and my attending asks them if anyone else in their family has had a similar surgery, thus reframing the line of questioning. “Yup, we all do it, everyone used to have all of these gut tumors but not we prevent that.” 🤦🏻♂️
It’s all about word choice, rephrasing questions, trust but verify, collateral information, unspoken clues, etc. The below post is right: patients don’t speak (or type into AI prompts) like doctors.