Johnson Thomas, MD, FACE

4.9K posts

Johnson Thomas, MD, FACE banner
Johnson Thomas, MD, FACE

Johnson Thomas, MD, FACE

@JohnsonThomasMD

Endocrinologist. Passionate about thyroid nodules, cancer & AI. Creator of Endo Tools iphone app,TUMScore, AIBx. Tweets ≠ medical advice. Retweets ≠ endorsement

Springfield, Missouri เข้าร่วม Nisan 2012
2.9K กำลังติดตาม2.4K ผู้ติดตาม
Johnson Thomas, MD, FACE
Johnson Thomas, MD, FACE@JohnsonThomasMD·
I'm honored to share that I've been selected as the inaugural Editor-in-Chief of AACE Endocrine AI, a new publication from the American Association of Clinical Endocrinology. This is a reflection of the incredible momentum building at the intersection of AI and endocrinology — and the vision AACE has shown in creating a dedicated space for this work. I'm looking forward to collaborating with a talented team of sub-editors and contributors to curate meaningful, clinically relevant content that helps our community navigate AI's role in patient care. #MedTwitter #EndoTwitter #AI More to come as we build this together. @TheAACE @TheEndoSociety @EndocrineToday @AmThyroidAssn aaceendocrine.ai/articles/2026/…
English
3
6
17
1.3K
Johnson Thomas, MD, FACE รีทวีตแล้ว
Razelle Kurzrock, MD
Razelle Kurzrock, MD@Dr_R_Kurzrock·
Interesting. We have 3 papers back from reviewers. The reviews have gotten much longer/more thorough recently— and the reviews of the 3 different papers ask the same questions. Likely, these near identical reviews were generated by AI Should we have AI generate the answers?
English
14
9
82
15.1K
Martin Varsavsky
Martin Varsavsky@martinvars·
We are all watching the giant AI models in the cloud. But underneath there is a revolution happening: AGI on your laptop, in your pocket, in your robot, no internet needed. The future isn't only models in which you have one big brain owned by X, Google, Anthropic or OpenAI; it's also a billion private, sovereign brains that you personally own.
English
9
11
92
13K
Johnson Thomas, MD, FACE รีทวีตแล้ว
Epic
Epic@HeyEpic·
What if you could talk directly to your patient's chart? Meet Ask Art. At @TGHCares, clinicians are asking questions about their patients' records and getting cited answers, anchored in the patient's data.
English
3
9
44
18.1K
Forbes
Forbes@Forbes·
The next era in healthcare innovation is edge intelligence: running AI at the point of care, where data is generated and acted upon in real time. This shift promises faster diagnoses, better patient outcomes and more efficient hospital operations. Sponsored by @DellTech forbes.com/sites/delltech…
English
10
14
67
32.4K
Johnson Thomas, MD, FACE รีทวีตแล้ว
Anthropic
Anthropic@AnthropicAI·
New Anthropic research: Emotion concepts and their function in a large language model. All LLMs sometimes act like they have emotions. But why? We found internal representations of emotion concepts that can drive Claude’s behavior, sometimes in surprising ways.
English
986
2.6K
17.4K
3.5M
Johnson Thomas, MD, FACE
Johnson Thomas, MD, FACE@JohnsonThomasMD·
Great for on device inference. Hope they can compress a good speech to text model also, then we could fit our on device medical models under 1 GB
PrismML@PrismML

Today, we are emerging from stealth and launching PrismML, an AI lab with Caltech origins that is centered on building the most concentrated form of intelligence. At PrismML, we believe that the next major leaps in AI will be driven by order-of-magnitude improvements in intelligence density, not just sheer parameter count. Our first proof point is the 1-bit Bonsai 8B, a 1-bit weight model that fits into 1.15 GBs of memory and delivers over 10x the intelligence density of its full-precision counterparts. It is 14x smaller, 8x faster, and 5x more energy efficient on edge hardware while remaining competitive with other models in its parameter-class. We are open-sourcing the model under Apache 2.0 license, along with Bonsai 4B and 1.7B models. When advanced models become small, fast, and efficient enough to run locally, the design space for AI changes immediately. We believe in a future of on-device agents, real-time robotics, offline intelligence and entirely new products that were previously impossible. We are excited to share our vision with you and keep working in the future to push the frontier of intelligence to the edge.

English
1
3
19
3K
Johnson Thomas, MD, FACE รีทวีตแล้ว
euan ashley
euan ashley@euanashley·
New AI paper from us this week. When my student first showed me his initial findings, I really didn’t know what to make of them. I felt that this was an interesting but curious loophole phenomenon that would shortly be closed. I was very wrong. arxiv.org/abs/2603.21687
euan ashley tweet media
English
59
262
1.5K
466.3K
Johnson Thomas, MD, FACE รีทวีตแล้ว
Andrej Karpathy
Andrej Karpathy@karpathy·
- Drafted a blog post - Used an LLM to meticulously improve the argument over 4 hours. - Wow, feeling great, it’s so convincing! - Fun idea let’s ask it to argue the opposite. - LLM demolishes the entire argument and convinces me that the opposite is in fact true. - lol The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.
English
1.7K
2.4K
31.1K
3.3M
Johnson Thomas, MD, FACE รีทวีตแล้ว
Craig M. Wax D.O.
Craig M. Wax D.O.@drcraigwax·
“What matters is this: a medical student used a large language model to generate a differential diagnosis last night. The output was correct. The student learned nothing. And no one in the institution knows whether that student can reason independently when the tool is unavailable, wrong, or hallucinating. That is the problem.” @MikeMooreDO @Substack open.substack.com/pub/mikemoored…
English
29
52
394
33.5K
Johnson Thomas, MD, FACE
Johnson Thomas, MD, FACE@JohnsonThomasMD·
@satyanadella @hoifungpoon @LuizaJarovsky @elonmusk @sama @DarioAmodei @sundarpichai @demishassabis @tim_cook @JeffBezos @finkd x.com/johnsonthomasm…
Johnson Thomas, MD, FACE@JohnsonThomasMD

What if a single H&E slide could reveal 21 protein markers without ever touching an antibody? Built a 3D interactive viewer around Microsoft's GigaTIME model — upload an H&E tile, run inference, and explore DAPI, CD3, CD8, PD-L1, Ki67 and 16 more channels as a rotatable point cloud The tumor microenvironment, decoded from routine staining 🧬 #DigitalPathology #AI #Oncology

QME
0
0
0
1.1K
Johnson Thomas, MD, FACE
Johnson Thomas, MD, FACE@JohnsonThomasMD·
clinicalmentis.com ClinicalMentis is a comprehensive clinical companion app that harnesses Gemini's native multimodality and advanced reasoning to address these challenges through five integrated features. Auth, database backend - everything created through Google AI studio. 1. Grounded Medical Q&A — Search-grounded answers from trusted sources (PubMed, CDC, Mayo Clinic) with citations, eliminating hallucinations where accuracy is non-negotiable. 2.MemCard / Flash card creator to remember key points from the answers generated by Gemini 3. Socratic Educator — A teaching chatbot that guides clinicians through differential diagnoses rather than giving direct answers, building diagnostic reasoning skills for physicians, residents, and students. 4. Multilingual Medical Scribe — Raw audio streams directly to Gemini 2.5 Flash, capturing code-switching conversations (e.g., Hindi-English) that traditional speech-to-text APIs miss. Generates structure
English
0
0
0
75
Logan Kilpatrick
Logan Kilpatrick@OfficialLoganK·
we are designing something special for Google IO and we want you to be part of it reply with your AI Studio app, along with a 1-sentence story on how and why you vibe coded it
English
193
46
974
67.7K

ค้นพบ

@ajumathew_ @TomVargheseJr @TheAACE @TheEndoSociety @EndocrineToday @AmThyroidAssn @vtangpricha @martinvars