Notebook or My Personal Learning Network
14.4K views | +0 today
 
Scooped by Gilbert C FAURE
onto Notebook or My Personal Learning Network
March 10, 2:44 PM
Scoop.it!

5 FREINS DE L’IA EN SANTÉ | Didier AMBROISE

5 FREINS DE L’IA EN SANTÉ | Didier AMBROISE | Notebook or My Personal Learning Network | Scoop.it
Il y a un an, on publiait chez Doshas Consulting notre livre blanc sur lโ€™IA en santรฉ.

ย 
Notre conviction : sans donnรฉes fiables, normalisรฉes, interopรฉrables โ€” lโ€™IA en santรฉ est un chรขteau de cartes.
ย 
Un an et 3 รฉtudes internationales plus tard, jโ€™ai mis ร  jour notre diagnostic.
ย 
5 freins. 5 verdicts. Un constat.
ย 
๐Ÿ”ง ๐—ง๐—ฒ๐—ฐ๐—ต๐—ป๐—ถ๐—พ๐˜‚๐—ฒ โ†’ La tech existe (lโ€™IA multimodale surpasse dรฉjร  les pathologistes sur certains cancers). Mais 88,6% des solutions restent du sur-mesure jamais dรฉployรฉ.
๐™‘๐™š๐™ง๐™™๐™ž๐™˜๐™ฉ : ๐™ก๐™š ๐™›๐™ง๐™š๐™ž๐™ฃ ๐™ฃโ€™๐™š๐™จ๐™ฉ ๐™ฅ๐™ก๐™ช๐™จ ๐™ฉ๐™š๐™˜๐™๐™ฃ๐™ž๐™ฆ๐™ช๐™š. ๐™„๐™ก ๐™š๐™จ๐™ฉ ๐™ค๐™ง๐™œ๐™–๐™ฃ๐™ž๐™จ๐™–๐™ฉ๐™ž๐™ค๐™ฃ๐™ฃ๐™š๐™ก.

โš–๏ธ ๐‰๐ฎ๐ซ๐ข๐๐ข๐ช๐ฎ๐ž โ†’ Lโ€™AI Act est en vigueur. Mais 70% des applis cliniques IA sont ร  TRL 4 ou moins (stade labo, jamais testรฉes en conditions rรฉelles). Aucune ร  TRL 7+ (dรฉployรฉe en routine clinique).
๐™‘๐™š๐™ง๐™™๐™ž๐™˜๐™ฉ : ๐™ค๐™ฃ ๐™– ๐™ก๐™š๐™จ ๐™ฉ๐™š๐™ญ๐™ฉ๐™š๐™จ, ๐™ฅ๐™–๐™จ ๐™ก๐™š๐™จ ๐™ค๐™ช๐™ฉ๐™ž๐™ก๐™จ.

๐Ÿง‘โ€โš•๏ธ ๐‡๐ฎ๐ฆ๐š๐ข๐ง โ†’ 18,2% seulement des รฉtudes en deep learning abordent lโ€™explicabilitรฉ.
๐™‘๐™š๐™ง๐™™๐™ž๐™˜๐™ฉ : ๐™ก๐™– ๐™—๐™ค๐™žฬ‚๐™ฉ๐™š ๐™ฃ๐™ค๐™ž๐™ง๐™š ๐™ง๐™š๐™จ๐™ฉ๐™š ๐™ฃ๐™ค๐™ž๐™ง๐™š. ๐™€๐™ฉ ๐™š๐™ก๐™ก๐™š ๐™œ๐™ง๐™ค๐™จ๐™จ๐™ž๐™ฉ.

๐Ÿ’ฐ ๐ƒ๐จ๐ง๐ง๐žฬ๐ž๐ฌ โ†’ 83% des รฉtudes viennent de pays riches. 0% de pays ร  bas revenus.
๐™‘๐™š๐™ง๐™™๐™ž๐™˜๐™ฉ : ๐™ก๐™š๐™จ ๐™™๐™ค๐™ฃ๐™ฃ๐™šฬ๐™š๐™จ ๐™™๐™š ๐™ฆ๐™ช๐™–๐™ก๐™ž๐™ฉ๐™šฬ ๐™ง๐™š๐™จ๐™ฉ๐™š๐™ฃ๐™ฉ ๐™ช๐™ฃ ๐™ฅ๐™ง๐™ž๐™ซ๐™ž๐™ก๐™šฬ€๐™œ๐™š.

๐Ÿงญ ๐„๐ญ๐ก๐ข๐ช๐ฎ๐ž ๐จ๐ฉ๐žฬ๐ซ๐š๐ญ๐ข๐จ๐ง๐ง๐ž๐ฅ๐ฅ๐ž โ†’ Les principes de lโ€™OMS sont citรฉs. Mais la durabilitรฉ ? 6% des รฉtudes. Lโ€™opรฉrationnalisation ? Quasi nulle.
๐™‘๐™š๐™ง๐™™๐™ž๐™˜๐™ฉ : ๐™ฃ๐™ค๐™ช๐™ซ๐™š๐™–๐™ช ๐™›๐™ง๐™š๐™ž๐™ฃ, ๐™ฅ๐™š๐™ช๐™ฉ-๐™šฬ‚๐™ฉ๐™ง๐™š ๐™ก๐™š ๐™ฅ๐™ก๐™ช๐™จ ๐™™๐™–๐™ฃ๐™œ๐™š๐™ง๐™š๐™ช๐™ญ.

๐’๐œ๐จ๐ซ๐ž ๐ ๐ฅ๐จ๐›๐š๐ฅ : ๐Ÿ ๐ฉ๐ซ๐จ๐ ๐ซ๐žฬ€๐ฌ ๐ง๐ž๐ญ ๐ฌ๐ฎ๐ซ ๐Ÿ“.
๐‘ณ๐’† ๐’”๐’†๐’“๐’‘๐’†๐’๐’• ๐’”๐’† ๐’Ž๐’๐’“๐’… ๐’†๐’๐’„๐’๐’“๐’† ๐’๐’‚ ๐’’๐’–๐’†๐’–๐’†.
ย 
๐Ÿ‘‰ Dรฉtail de chaque frein dans le carrousel ci-dessous.
No comment yet.
Notebook or My Personal Learning Network
a personal notebook since summer 2013, a virtual scrapbook
Your new post is loading...
Your new post is loading...
Scooped by Gilbert C FAURE
October 13, 2013 8:40 AM
Scoop.it!

This notebook..

is a personal Notebook

ย 

Thanks John Dudley for the following tweet

"If you like interesting snippets on all sorts of subjects relevant to academia, information, the world, highly recommended is @grip54 's collection:"

ย 

La curation de contenus, la mรฉmoire partagรฉe d'une veille scientifique et sociรฉtale

Gilbert C FAURE's insight:

... designed to collect posts and informations I found and want to keep available but not relevant to the other topics I am curating on Scoop.it (on behalf of ASSIM):

ย 

the most sucessful being

Immunology, teaching and learning immunology

http://www.scoop.it/t/immunology

and

From flow cytometry to cytomics

http://www.scoop.it/t/from-flow-cytometry-to-cytomics

Immunology and Biotherapies, a page of resources for the DIUย 

ย http://www.scoop.it/t/immunology-and-biotherapies

ย 

followed by

Nancy, Lorraine

ย http://www.scoop.it/t/nancy-lorraine

I am based at Universitรฉ Lorraine in Nancy

Wuhan, Hubei,

ย http://www.scoop.it/t/wuhan

because we have a long standing collaboration through a french speaking medical training program between Facultรฉ de Mรฉdecine de Nancy and WuDA, Wuhan university medical school and Zhongnan Hospital

ย ย 

CME-CPD,

ย http://www.scoop.it/t/cme-cpd

because I am at EACCME in Brussels, representative of the medical biopathology and laboratory medicine UEMS section

ย 

Mucosal Immunity,

ย http://www.scoop.it/t/mucosal-immunity

because it was one of our main research interest some years agoย 

ย 

It is a kind of electronic scrapbook with many ideas shared by others.

It focuses more and more on new ways of Teaching and Learning: e-, m-, a-, b-, h-, c-, d, ld-, s-, p-, w-, pb-, ll- ....

Thanks to all

No comment yet.
Scooped by Gilbert C FAURE
April 10, 3:57 AM
Scoop.it!

Argentique Photo, Rachat d'appareils photos Argentique

Argentique Photo, Rachat d'appareils photos Argentique | Notebook or My Personal Learning Network | Scoop.it
Vendez-nous votre matรฉriel et appareils photos argentique. Nous rachetons toutes marques.
No comment yet.
Scooped by Gilbert C FAURE
April 9, 6:21 AM
Scoop.it!

Across 114 US internal medicine clerkship programs, not one included structured AI teaching, even as students were already turning to ChatGPT more often than their own professors. 1๏ธโƒฃ A national… ...

Across 114 US internal medicine clerkship programs, not one included structured AI teaching, even as students were already turning to ChatGPT more often than their own professors. 1๏ธโƒฃ A national… ... | Notebook or My Personal Learning Network | Scoop.it
Across 114 US internal medicine clerkship programs, not one included structured AI teaching, even as students were already turning to ChatGPT more often than their own professors.

1๏ธโƒฃ A national survey of 114 internal medicine clerkship directors found zero programs with structured AI teaching in their curriculum.

2๏ธโƒฃ Nearly 60% of clerkship directors felt AI should be formally incorporated into the medical school curriculum.

3๏ธโƒฃ Top teaching priorities: using AI to support clinical reasoning (58%) and interpreting AI output (54%).

4๏ธโƒฃ 47% felt students should be taught the ethical challenges of AI in clinical practice, including data privacy and algorithmic bias.

5๏ธโƒฃ Meanwhile, 44% of students were already using ChatGPT at least weekly.
6๏ธโƒฃ 45% of students were more likely to ask ChatGPT a question than their professor or attending physician.

7๏ธโƒฃ Three-quarters of clerkship directors don't use AI for any administrative tasks, including student scheduling or feedback analysis.

8๏ธโƒฃ Faculty knowledge was the top barrier to teaching AI (84%), followed by clerkship directors' own knowledge (75%).

9๏ธโƒฃ Nearly half of clerkship directors had received no AI training at all; only 1 in 114 described it as "very adequate."

๐Ÿ”Ÿ The authors call for institutional investment in faculty development now, before student use outpaces what educators can safely guide.

โœ๐Ÿป Navin L. Kumar, Casey McQuade, Eliana Bonifacino, Irene Alexandraki, Kathryn K. Hufmeyer, Michael Kisielewski, Cindy Lai, Elexis McBee, Prashant Patel, @Nora Y. Osman. Artificial Intelligence in the Internal Medicine Clerkship: Results of a National Survey. JGIM Journal of General Internal Medicine. 2026. DOI: 10.1007/s11606-026-10354-1
No comment yet.
Scooped by Gilbert C FAURE
April 9, 6:15 AM
Scoop.it!

Le professeur Didier Raoult vient de se faire rétracter 2 études, donc il passe à 50 études rétractées à son actif. Nous pouvons donc féliciter ce grand "chercheur" français, élite de la nation sel...

Le professeur Didier Raoult vient de se faire rétracter 2 études, donc il passe à 50 études rétractées à son actif. Nous pouvons donc féliciter ce grand "chercheur" français, élite de la nation sel... | Notebook or My Personal Learning Network | Scoop.it
Le professeur Didier Raoult vient de se faire rรฉtracter 2 รฉtudes, donc il passe ร  50 รฉtudes rรฉtractรฉes ร  son actif. Nous pouvons donc fรฉliciter ce grand "chercheur" franรงais, รฉlite de la nation selon ses propres propos, et dont la plus fameuse รฉtude de 2020 sur l'hydroxychloroquine รฉtait d'une mรฉdiocritรฉ telle que pas mรชme un รฉtudiant de 2e annรฉe en mรฉdecine n'aurait osรฉ la prรฉsenter ร  son professeur, qui passe presque dans le top 10 mondial des pires "chercheurs".

Une pensรฉe รฉmue aux mรฉdias franรงais (et suisses) qui l'ont invitรฉ sur les plateaux TV sans contradicteur pour raconter absolument n'importe quoi sur le covid, ainsi qu'aux institutions franรงaises qui ont fermรฉ les yeux pendant des annรฉes (tout comme avec l'histoire du Mediator, et de Christian Perronne qui a รฉtรฉ blanchi par la chambre disciplinaire de l'ordre mรฉdecins, on commence ร  avoir l'habitude dans ce pays de ne pas toucher aux "grands").

Lien vers le Retraction Watch Leaderboard :
https://lnkd.in/eZZ-c_-f | 16 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 8, 10:48 AM
Scoop.it!

This study offers a striking demonstration of how easily fabricated health information can infiltrate the systems millions of people rely on for medical guidance. A fictional disease, seeded throug...

This study offers a striking demonstration of how easily fabricated health information can infiltrate the systems millions of people rely on for medical guidance. A fictional disease, seeded throug... | Notebook or My Personal Learning Network | Scoop.it
This study offers a striking demonstration of how easily fabricated health information can infiltrate the systems millions of people rely on for medical guidance. A fictional disease, seeded through two obviously bogus preprints, was repeated as fact across major AI platforms within weeks, and eventually cited in peer-reviewed literature. The researchers themselves were alarmed at how well it worked. https://lnkd.in/er24NDG2

Anyone familiar with how LLM developers collect content and sell their services as information intermediaries would not be surprised. Our information environments are more shaped by technology, design choices, and commercial interests than most people think.

This goes well beyond AI safety or academic integrity. We need to stop framing health misinformation as primarily a communication problem, solvable through better messaging or more trusted spokespeople. Marketers are already using LLMs to generate health content optimized for AI overviews. Developers are embedding AI into health apps without standardized evaluation. Researchers are using LLMs to shortcut analysis, sometimes without reading the sources those tools cite. At every node in this chain, the mere availability of an AI tool has become sufficient justification for its use, with almost no conversation about what skills, governance, or values should accompany that.

What would actually help: mandatory pre-deployment evaluation of health AI for misinformation susceptibility. Transparency requirements for AI-generated health content. Real health and algorithmic literacy in education. Independent, non-commercially driven health knowledge infrastructure. Stronger norms around AI use in scientific publishing. And public health voices in the rooms where these systems are designed, not just called in afterward to clean up.

Public health cannot message its way out of an information swamp. If this continues to be treated as a communication challenge rather than a structural one, we will keep arriving at the shore with a megaphone while the water keeps rising.
No comment yet.
Scooped by Gilbert C FAURE
April 7, 10:47 AM
Scoop.it!

How to start using Claude Code in 30 minutes (even if you’ve never written a single line of code) → 0–5 min Install Claude Code. Sign in. → 5–10 min Create your context folder. Add an about-me.md...

How to start using Claude Code in 30 minutes (even if you’ve never written a single line of code) → 0–5 min Install Claude Code. Sign in. → 5–10 min Create your context folder. Add an about-me.md... | Notebook or My Personal Learning Network | Scoop.it
How to start using Claude Code in 30 minutes
(even if youโ€™ve never written a single line of code)

โ†’ 0โ€“5 min
Install Claude Code. Sign in.

โ†’ 5โ€“10 min
Create your context folder.
Add an about-me.md.

This is the step 90% of people skip and itโ€™s the most important.

โ†’ 10โ€“15 min
Start your first conversation.
Use Opus 4.6+.
Let it ask questions before it builds.

โ†’ 15โ€“20 min
Open the live preview.
Give clear, simple feedback:
โ€œMake the headline bigger.โ€
โ€œUse an off-white background.โ€

โ†’ 20โ€“30 min
Give it a real project.
A landing page.
Something youโ€™ve been putting off for months.

โธป

Pro tip:
Turn on โ€œBypass permissions.โ€
Claude Code works uninterrupted no constant approvals.

โธป

Most people overcomplicate this.

You donโ€™t need to โ€œlearn to code.โ€
You need to learn how to collaborate.

Thatโ€™s the shift.


โ™ป๏ธ Repost to help your network
๐Ÿ” Follow Paul Storm for no-BS AI insights, viral tools, and growth hacks! | 88 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 7, 4:52 AM
Scoop.it!

Public Turns to AI Chatbots for Health Info Amid Access, Affordability Concerns | KFF posted on the topic

Public Turns to AI Chatbots for Health Info Amid Access, Affordability Concerns | KFF posted on the topic | Notebook or My Personal Learning Network | Scoop.it
Our latest Tracking Poll on Health Information and Trustโ€ฏfinds that one-third of the public report using AI chatbots for health information and advice in the past year, similar to the share who have relied on social media for health.

One in five adults who use AI for health cite difficulties accessing or affording health care as reasons they turned to these chatbots, including larger shares of younger and lower income users.

The latest edition of The Monitor explores these findings and more.
No comment yet.
Scooped by Gilbert C FAURE
April 6, 11:33 AM
Scoop.it!

Drug Administration Routes Impact Treatment Outcomes | Dr. Aarunee Krishna posted on the topic

Drug Administration Routes Impact Treatment Outcomes | Dr. Aarunee Krishna posted on the topic | Notebook or My Personal Learning Network | Scoop.it
Most people think a drug works simply because of what is inside it.
In reality, how it enters your body can be just as important.

This is called the route of drug administrationโ€”oral, intravenous, intramuscular, subcutaneous, inhalational, topicalโ€”and each route changes how a drug behaves inside you.

Why does this matter for the general population?

Because the same drug can act very differently depending on how it is taken:
โ€ข A tablet may take 30โ€“60 minutes to act, while an injection can work within seconds
โ€ข Some drugs are destroyed in the stomach and must never be taken orally
โ€ข Incorrect use (like crushing sustained-release tablets) can lead to toxicity
โ€ข Inhalers, if used incorrectly, may deliver almost no benefit despite regular use

In simple terms: right drug + wrong route = wrong outcome

As an MD trainee in Medical Pharmacology, this is not just academic knowledge for meโ€”it is a responsibility.

We often see:
โ€ข Antibiotic misuse due to wrong administration practices
โ€ข Poor control of chronic diseases because of improper drug use
โ€ข Adverse drug reactions that could have been prevented with basic awareness

Educating people about how to take medicines correctly can:
โ€ข Improve treatment outcomes
โ€ข Reduce side effects
โ€ข Prevent drug resistance
โ€ข Empower patients to participate in their own care

Pharmacology is not just about drugsโ€”it is about optimizing how those drugs interact with human biology.

And sometimes, the smallest detailโ€”like the route of administrationโ€”makes the biggest difference.

#Medicine #Pharmacology #PatientEducation #RationalUseOfMedicines #Healthcare #MedicalEducation
No comment yet.
Scooped by Gilbert C FAURE
April 6, 11:25 AM
Scoop.it!

Claude 4.5 est devenu une vraie suite de travail. Et si vous êtes dirigeant, voici l’essentiel à retenir : โœฆ 1. Choisissez le bon modèle Opus 4.5 → stratégie, raisonnement complexe, sujets à fo...

Claude 4.5 est devenu une vraie suite de travail. Et si vous êtes dirigeant, voici l’essentiel à retenir : โœฆ 1. Choisissez le bon modèle Opus 4.5 → stratégie, raisonnement complexe, sujets à fo... | Notebook or My Personal Learning Network | Scoop.it
Claude 4.5 est devenu une vraie suite de travail.

Et si vous รชtes dirigeant, voici lโ€™essentiel ร  retenir :

โœฆ 1. Choisissez le bon modรจle

Opus 4.5 โ†’ stratรฉgie, raisonnement complexe, sujets ร  fort enjeu
Sonnet 4.5 โ†’ rรฉdaction, synthรจse, usage business quotidien
Haiku 4.5 โ†’ tรขches simples, vitesse, gros volumes

Ne demandez pas la mรชme chose ร  tous les modรจles.

Cโ€™est comme utiliser le mรชme vรฉhicule pour livrer un colisโ€ฆ ou traverser un dรฉsert.

โœฆ 2. Claude ne sert pas quโ€™ร  รฉcrire

Il peut aussi :

โ†’ chercher sur le web
โ†’ analyser des fichiers
โ†’ travailler avec Google Drive
โ†’ gรฉrer des Projects
โ†’ coder avec Claude Code
โ†’ produire des livrables avec Artifacts

Le sujet nโ€™est plus โ€œfais-moi un texteโ€.

Le sujet est : fais avancer mon travail.

โœฆ 3. Les fonctions les plus utiles pour un dirigeant

Artifacts โ†’ produire un plan dโ€™action, une SOP, une proposition, un tableau

Web Search โ†’ veille, prรฉparation de rendez-vous, analyse concurrentielle

Analyse de fichiers โ†’ comprendre vos ventes, leads, devis, marges

Projects โ†’ crรฉer une mรฉmoire de travail par sujet : marketing, commercial,
recrutement, direction

Claude Code โ†’ prototyper un outil interne, accรฉlรฉrer un projet digital, crรฉer un
MVP

Tรฉlรฉversement de fichiers โ†’ donner PDF, contrats, comptes rendus, captures
dโ€™รฉcran pour quโ€™il analyse votre rรฉalitรฉ

โœฆ 4. Les meilleurs cas dโ€™usage concrets

โ†’ prรฉparer un rendez-vous commercial en 3 minutes
โ†’ transformer une rรฉunion en dรฉcisions + tรขches + prioritรฉs
โ†’ analyser un fichier de ventes ou de leads
โ†’ crรฉer une proposition commerciale plus vite
โ†’ structurer un plan 90 jours
โ†’ transformer des documents dispersรฉs en systรจme clair

โœฆ 5. La vraie mรฉthode

Le dรฉbutant dit :

โ€œRรฉsume-moi รงa.โ€

Lโ€™utilisateur avancรฉ dit :

โ†’ voici mon objectif
โ†’ voici mes fichiers
โ†’ voici mes contraintes
โ†’ pose-moi les questions manquantes
โ†’ propose un plan
โ†’ exรฉcute
โ†’ amรฉliore la V1

En clair : il ne โ€œpromptโ€ pas.

Il manage Claude comme un collaborateur.

Cโ€™est รงa, la vraie mise ร  jour.

---------------------------------

PS : jโ€™organise une masterclass le 10 avril ร  20H pour vous aider ร  faire votre mise ร  jour Claude et apprendre ร  lโ€™utiliser pour vous libรฉrer de lโ€™opรฉrationnel, rรฉduire votre surcharge mentale et accรฉlรฉrer la croissance de votre entreprise.

๐Ÿ‘‰ Inscription ici :

https://lnkd.in/eN6mBd9B | 12 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 3, 11:33 AM
Scoop.it!

#phd #research | Faheem Ullah | 16 comments

#phd #research | Faheem Ullah | 16 comments | Notebook or My Personal Learning Network | Scoop.it
PhD Students - How to check if your research idea is actually new?

First, let's understand why novelty is important for research

Here is what reviewers will look for in your research

1๏ธโƒฃ ๐๐จ๐ฏ๐ž๐ฅ๐ญ๐ฒ โ†’ Is it new?
2๏ธโƒฃ ๐’๐ข๐ ๐ง๐ข๐Ÿ๐ข๐œ๐š๐ง๐œ๐ž โ†’ Is it important for anyone?
3๏ธโƒฃ ๐Œ๐ž๐ญ๐ก๐จ๐๐จ๐ฅ๐จ๐ ๐ฒ โ†’ Is it conducted the right way?
4๏ธโƒฃ ๐•๐ž๐ซ๐ข๐Ÿ๐ข๐œ๐š๐ญ๐ข๐จ๐ง โ†’ Can other researchers verify it?
5๏ธโƒฃ ๐๐ซ๐ž๐ฌ๐ž๐ง๐ญ๐š๐ญ๐ข๐จ๐ง โ†’ Is it presented in the right way?

You see novelty comes on the top of this list.

To confirm novelty, meet ๐๐š๐ญ๐’๐ง๐š๐ฉ ๐„๐ฎ๐ซ๐ž๐ค๐š.

Eureka thinks like an IP expert.

Here is how it works.

1. Go to https://lnkd.in/dqiq55cM
2. Describe your research idea in 20-30 words
3. Eureka scans 200M+ patents to compare your idea
4. It shows you a side-by-side table of your idea vs existing ones
5. Export the entire novelty report to share with others

๐–๐ก๐ฒ ๐ฌ๐ก๐จ๐ฎ๐ฅ๐ ๐ฒ๐จ๐ฎ ๐ญ๐ซ๐ฒ ๐ข๐ญ?

โœ“ Confirms the novelty of your research idea
โœ“ Gives you confidence in your research direction
โœ“ Change research idea if it's not novel
โœ“ After confirmation, dive deep into your research

๐ŸŽ—๏ธ Try Eureka for FREE: https://lnkd.in/dqiq55cM

โ„๏ธ Anything you'd like to add?

#phd #research
| 16 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 2, 1:21 PM
Scoop.it!

Le TOP 10 | Elliacare | 19 comments

Le TOP 10 | Elliacare | 19 comments | Notebook or My Personal Learning Network | Scoop.it
L'IA s'est installรฉe dans les outils mรฉdicaux du quotidien. Pas par dรฉcision collective. Par glissement progressif.

On vous prรฉsente le Top 10 de ses usages en santรฉ ๐Ÿ“Š

On a voulu faire le point. Pas sur les promesses, sur ce qui existe rรฉellement. Sur ce qui est dรฉployรฉ, ce qui est encore en cours, et ce qui reste ร  construire.

Ce carrousel recense les 10 usages de l'IA les plus documentรฉs pour les soignants en 2025-2026.
L'IA n'a pas attendu d'รชtre invitรฉe. Elle s'est glissรฉe dans les logiciels de prescription, les DPI, les outils de documentation, les systรจmes d'alerte.
Par touches. Par intรฉgrations successives. Et sans que la formation des soignants n'anticipe ce deplacement

Ce que le panorama montre, c'est une rรฉalitรฉ ร  deux vitesses.

D'un cรดtรฉ, des outils qui tiennent leurs promesses.
โ–ช๏ธLe scribing IA divise par deux le temps de documentation
โ–ช๏ธLa veille bibliographique qui prenait une heure se fait dorenavant en 15 minutes. A condition de bien savoir la rรฉaliser avec lโ€™IA.
โ–ช๏ธL'aide ร  la dรฉcision clinique, quand elle est couplรฉe au jugement mรฉdical, produit de meilleurs diagnostics que l'un ou l'autre seul.

Le paradoxe est net. Les outils fonctionnent. Les gains sont mesurables
Et pourtant, 45 % des soignants n'informent jamais leurs patients qu'ils utilisent l'IA dans leur prise en charge.
Alors que 84 % des Franรงais souhaitent l'etre.

Ce n'est pas de la mauvaise volontรฉ. C'est l'absence de cadre.
On automatise des taches, mais on ne forme pas ร  ce que ca engage, puis on dรฉploie des outils.
Mais on ne prepare pas le soignant ร  รฉvaluer ce qu'il dรฉlรจgue, ร  identifier lร  ou l'outil รฉchoue, ร  maintenir son jugement clinique souverain face ร  une recommandation algorithmique.

La formation n'est pas un accessoire du dรฉploiement. C'est sa condition de validitรฉ.

C'est exactement ce qu'Elliacare adresse. Pas l'enthousiasme pour les outils. La compรฉtence pour les utiliser, et pour savoir quand ne pas s'y fier.

๐Ÿ‘‰ Swipez le carrousel pour voir les 10 usages, leurs niveaux de maturitรฉ et les chiffres qui les รฉtayent.

Sur ces 10 usages, lesquels faites-vous dรฉjร , et lesquels vous manquent encore ?

#IAenSantรฉ #Elliacare #MรฉdecineAugmentรฉe #FormationIA | 19 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 2, 1:13 PM
Scoop.it!

Steady rise: female scientists & engineers reach 7.9 mln | Eurostat

Steady rise: female scientists & engineers reach 7.9 mln | Eurostat | Notebook or My Personal Learning Network | Scoop.it
The number of women working as scientists and engineers in the EU reached 7.9 million in 2024, representing 40.5% of the scientists and engineersโ€™ workforce across all economic activities. ๐Ÿง‘โ€๐Ÿ”ฌ๐Ÿ”ฌ

Across EU regions, highest shares in:

๐Ÿ‡ช๐Ÿ‡ธ Canarias (58.8%)
๐Ÿ‡ต๐Ÿ‡น Regiรฃo Autรณnoma dos Aรงores (57.3%)
๐Ÿ‡ต๐Ÿ‡น Madeira (56.4%)

Lowest in:

๐Ÿ‡ญ๐Ÿ‡บ Kรถzรฉp-Magyarorszรกg (30.0%)
๐Ÿ‡ซ๐Ÿ‡ฎ Manner-Suomi (30.7%)
๐Ÿ‡ฎ๐Ÿ‡น Sud (31.1%)

โ„น๏ธ Please note that the map includes available regional data from EU countries, EFTA and candidate countries. The ranking in the caption of the post is based on data from EU countries only.

Learn more ๐Ÿ‘‰ https://lnkd.in/eHQqWP_g
No comment yet.
Scooped by Gilbert C FAURE
April 1, 10:15 AM
Scoop.it!

https://french.visitbeijing.com.cn/article/47DYL69KFDB

Lao...

No comment yet.
Scooped by Gilbert C FAURE
April 10, 7:38 AM
Scoop.it!

Après les hallucinations, les IA de pointe médicales souffrent d’un nouveau travers, l’effet Mirage: elles construisent des diagnostics à partir de données qui ne leur ont jamais été fournies

Après les hallucinations, les IA de pointe médicales souffrent d’un nouveau travers, l’effet Mirage: elles construisent des diagnostics à partir de données qui ne leur ont jamais été fournies | Notebook or My Personal Learning Network | Scoop.it
Des chercheurs de l'Universitรฉ de Stanford ont dรฉcouvert (et nommรฉ) l'effet Mirage des IA. Ces derniรจres, de GPT-5 d'OpenAI ร  Claude Opus 4.5 d'Anthropic en passant par Gemini 3 Pro de Google, arrivent ainsi ร  livrer un diagnostic mรฉdical avec assurance ร  partir d'รฉlรฉments visuels qu'elle
No comment yet.
Scooped by Gilbert C FAURE
April 9, 6:22 AM
Scoop.it!

Une cyberattaque vient de révéler l'écosystème ultra-secret qui alimente ChatGPT et Claude en données d'entraînement. Le groupe TeamPCP a compromis Mercor, l'un des sous-traitants clés d'OpenAI… |...

Une cyberattaque vient de révéler l'écosystème ultra-secret qui alimente ChatGPT et Claude en données d'entraînement. Le groupe TeamPCP a compromis Mercor, l'un des sous-traitants clés d'OpenAI… |... | Notebook or My Personal Learning Network | Scoop.it
Une cyberattaque vient de rรฉvรฉler l'รฉcosystรจme ultra-secret qui alimente ChatGPT et Claude en donnรฉes d'entraรฎnement.

Le groupe TeamPCP a compromis Mercor, l'un des sous-traitants clรฉs d'OpenAI, Anthropic et Meta. Rรฉsultat : 200 Go de bases de donnรฉes et 3 To d'informations sensibles exposรฉs. Meta a immรฉdiatement suspendu ses projets, laissant des centaines de contractuels sans travail.
Ce qui se dessine derriรจre cette faille, c'est toute une industrie invisible. Mercor, comme Surge ou Scale AI, emploie des milliers de personnes pour crรฉer des jeux de donnรฉes propriรฉtaires gardรฉs jalousement secrets. Ces entreprises sont les vรฉritables fabriques de l'intelligence artificielle moderne.
Mais l'incident rรฉvรจle surtout la fragilitรฉ de cette chaรฎne d'approvisionnement critique. Une seule vulnรฉrabilitรฉ dans un outil tiers โ€” ici LiteLLM โ€” peut paralyser la production de donnรฉes pour les modรจles les plus avancรฉs au monde.
L'asymรฉtrie de pouvoir saute aux yeux : les gรฉants de l'IA externalisent leurs risques vers des sous-traitants prรฉcaires, qui deviennent les maillons faibles de toute la chaรฎne. Quand รงa casse, ce sont eux qui paient le prix fort.
Cette architecture en silo crรฉe des vulnรฉrabilitรฉs systรฉmiques. Que se passerait-il si des coopรฉratives de donnรฉes remplaรงaient ces intermรฉdiaires prรฉcaires ? Si les travailleurs de l'annotation possรฉdaient collectivement leurs outils et leurs donnรฉes ?
Comment repenser cette chaรฎne de valeur pour qu'elle ne repose plus sur la prรฉcaritรฉ organisรฉe ? | 13 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 9, 6:19 AM
Scoop.it!

The ChatGPT Symptom Spiral | Sophie Nadeau ๐Ÿ‡จ๐Ÿ‡ฆ

The ChatGPT Symptom Spiral | Sophie Nadeau ๐Ÿ‡จ๐Ÿ‡ฆ | Notebook or My Personal Learning Network | Scoop.it
Our Canadian Medical Association Health & Media Tracking Survey found that people are 5x more likely to experience harm when they use AI for health advice which is a huge issue when nearly half of those surveyed use it. This piece in The Atlantic does a great job explaining how that happens and why we should all be concerned. https://lnkd.in/eBkfQtVC

Read our research here: https://lnkd.in/g-EsyexG
No comment yet.
Scooped by Gilbert C FAURE
April 9, 4:44 AM
Scoop.it!

Patients are not waiting for permission: the rise of the AI-empowered patient | Hedieh Mehrtash, PhD MPH

Patients are not waiting for permission: the rise of the AI-empowered patient | Hedieh Mehrtash, PhD MPH | Notebook or My Personal Learning Network | Scoop.it
๐Ÿคณ Patients are already using AI to make health decisions, but we haven't started measuring it yet.

As someone who has spent the past decade developing and validating PREMs and PROMs, this gap is hard to ignore.

๐Ÿฅ Our measurement frameworks currently measure a patient interacting with a health worker, inside a facility setting. Today, patients are interpreting symptoms with AI, exploring diagnoses before appointments, and potentially making decisions between or even instead of clinical encounters. We need to start making the shift to measure experience and outcomes beyond just care beginning and ending in the health facility.

The recent Lancet Primary Care paper by Riggare et al calls this shift "triadic care" : the evolving relationship between health worker, patient, and AI. We need to start responding as a research community because right now, a growing part of the patient experience exists outside our field of vision.

โš–๏ธ Patients are navigating systems that may be inaccessible, stigmatizing, or dismissive of their needs. AI chats may offer new forms of support, but it also carries risks of bias, misinformation, and new forms of exclusion. Measurement must not only capture these experiences, but ensure they are visible for accountability.

How do we begin to question the experience we're trying to measure including whether patients are genuinely supported to make autonomous, informed decisions when it unfolds entirely outside the system?

We need to begin a new era of capturing experiences: Did you arrive with AI-generated information? Did it shape your understanding or your anxiety? Did it support your ability to make an informed choice or did it make that harder?

Critical AI health literacy on whether patients can evaluate and use AI-generated health information is a necessary foundation. But literacy alone doesn't tell us what people actually experienced, decided, or felt. That's what PREMs are for.

This is the next frontier in moving the needle forward on person-centered care ๐Ÿš€.

๐ŸŒ It's time to further explore what we measure, and why.

https://lnkd.in/eM3ZpZps

(PREM: Patient Reported Experience Meaures; PROM: Patient Reported Outcome Measures)
No comment yet.
Scooped by Gilbert C FAURE
April 7, 1:45 PM
Scoop.it!

Le podcast natif - Flore Di Sciullo - Marie-Eva Lesaunier - Arnaud Mercier - 1re édition | Lgdj.fr

Le podcast natif - Flore Di Sciullo - Marie-Eva Lesaunier - Arnaud Mercier - 1re édition | Lgdj.fr | Notebook or My Personal Learning Network | Scoop.it
Retrouvez Le podcast natif de Flore Di Sciullo, Marie-Eva Lesaunier, Arnaud Mercier, sur la librairie juridique Lgdj.fr - Livraison en 24 heures pour les livres en stock & Frais de port ร  partir de 0.01 euro !
No comment yet.
Scooped by Gilbert C FAURE
April 7, 5:12 AM
Scoop.it!

Panorama des services d’IA générative –

Panorama des services d’IA générative – | Notebook or My Personal Learning Network | Scoop.it
Lโ€™intelligence artificielle gรฉnรฉrative nโ€™est pas un simple outil. Cโ€™est une technologie ร  usage gรฉnรฉral, comparable ร  Internet ou ร  lโ€™รฉlectricitรฉ, qui impacte lโ€™ensemble des mรฉtiers et des secteurs . Mais cette puissance sโ€™accompagne dโ€™une complexitรฉ croissante : une multiplication d'outils et services avec des positionnements flous et des promesses parfois difficiles ร  comparer. Notre panoramaโ€ฆ
No comment yet.
Scooped by Gilbert C FAURE
April 7, 4:43 AM
Scoop.it!

17 000 notaires. 9 milliards d'euros de chiffre d'affaires. Et 50% de leurs tâches quotidiennes sont automatisables aujourd'hui. ๐Ÿ“‰ La Chambre des Notaires de Paris a lancé VictorIA, un projet d'I...

17 000 notaires. 9 milliards d'euros de chiffre d'affaires. Et 50% de leurs tâches quotidiennes sont automatisables aujourd'hui. ๐Ÿ“‰ La Chambre des Notaires de Paris a lancé VictorIA, un projet d'I... | Notebook or My Personal Learning Network | Scoop.it
17 000 notaires. 9 milliards d'euros de chiffre d'affaires. Et 50% de leurs tรขches quotidiennes sont automatisables aujourd'hui. ๐Ÿ“‰

La Chambre des Notaires de Paris a lancรฉ VictorIA, un projet d'IA pour automatiser la vรฉrification documentaire. Quai des Notaires a levรฉ 7 millions d'euros et facilitรฉ 110 000 actes. Legapass a dรฉployรฉ VigiNot dans la moitiรฉ des รฉtudes de France. Et depuis janvier 2025, la loi impose la dรฉmatรฉrialisation des actes par dรฉfaut.

Le revenu moyen par notaire a chutรฉ de 30% depuis 2019. Et l'IA arrive pour automatiser ce qui prenait des heures.

Ce qui est en train de mourir dans le notariat :
๐Ÿ‘‰ La vรฉrification documentaire manuelle. Un acte de vente nรฉcessite des dizaines de piรจces. Un agent IA les identifie, les classe et les vรฉrifie en quelques minutes. Le clerc qui y passait une journรฉe voit son rรดle fondre.
๐Ÿ‘‰ La formalitรฉ comme prestation. 300 000 formalitรฉs automatisรฉes par Quai des Notaires. La rรฉdaction d'actes standards, les dรฉclarations, les enregistrements : l'IA fait รงa plus vite, sans oubli, 24h/24.
๐Ÿ‘‰ Les frais opaques. 82% de moins de litiges sur les actes signรฉs รฉlectroniquement. Le processus se digitalise, se simplifie. Le client se demande pourquoi il paie ce qu'il paie. 47% des Franรงais jugent encore les notaires indispensables. Mais pour combien de temps ?

Le test de survie du notariat :
1๏ธโƒฃ Devenir le conseil patrimonial de confiance. Le notaire qui survit n'est plus celui qui tamponne. C'est celui qui conseille une stratรฉgie successorale, optimise une transmission d'entreprise, structure un montage familial. L'architecte du patrimoine.
2๏ธโƒฃ Automatiser l'usine, humaniser la relation. VictorIA et Quai des Notaires montrent la voie. Toute la vรฉrification et la formalitรฉ aux agents. Le temps gagnรฉ, c'est du temps de conseil ร  forte valeur ajoutรฉe.
3๏ธโƒฃ Jouer la transparence sur les honoraires. Le modรจle "frais de notaire" que personne ne comprend est un hรฉritage. Le notaire qui explique, dรฉtaille et justifie la valeur de son conseil gagne la confiance d'une gรฉnรฉration qui compare tout en ligne.

Ma conviction de vieux briscard : le notaire ne disparaรฎtra pas. C'est le garant de la sรฉcuritรฉ juridique. Mais l'รฉtude de 15 clercs qui traite du papier, oui. Le notariat de 2030, c'est 5 experts patrimoniaux puissance IA et un systรจme qui fait le reste.

๐Ÿš€ Dirigeants : votre patrimoine mรฉrite un architecte, pas un tampon. ๐Ÿ‘‰ https://lnkd.in/e6k46944
๐ŸŽ“ Consultants : chaque secteur a besoin de son architecte puissance IA. ๐Ÿ‘‰ https://lnkd.in/eaJd3bZ8
๐ŸŽฏ Masterclass gratuite le 16 avril : passez du prompting ร  l'IA agentique en 1h. ๐Ÿ‘‰ https://lnkd.in/eZGGrvvY
๐Ÿš€ Nos bootcamps tournent sur Claude. Un projet IA ? ๐Ÿ‘‰ https://decisionia.com/rdv

15 industries, 60 gรฉants, 15 tests de survie. Lequel vous a le plus frappรฉ ? ๐Ÿ‘‡ | 79 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 6, 11:28 AM
Scoop.it!

#earthatnight #artemisii | Peter Stumpf

#earthatnight #artemisii | Peter Stumpf | Notebook or My Personal Learning Network | Scoop.it
Earth at Night, illuminated by full moon - a dark sphere, with editing or longer exposure transformed to pale blue, show bright spots which are cities at night across iberian peninsula and Africa's mediterranean sea coast line & South America on the right.

Dark shot: https://lnkd.in/d6gmWgiF

Colorful shot https://lnkd.in/dr_4iEEm

#earthatnight #artemisII
No comment yet.
Scooped by Gilbert C FAURE
April 3, 11:38 AM
Scoop.it!

Academic Research vs Case Study: Key Differences | Docadeson R. posted on the topic

Academic Research vs Case Study: Key Differences | Docadeson R. posted on the topic | Notebook or My Personal Learning Network | Scoop.it
๐—ช๐—ต๐˜† ๐—–๐—ต๐—ผ๐—ผ๐˜€๐—ถ๐—ป๐—ด ๐—•๐—ฒ๐˜๐˜„๐—ฒ๐—ฒ๐—ป ๐—”๐—ฐ๐˜๐—ถ๐—ผ๐—ป ๐—ฅ๐—ฒ๐˜€๐—ฒ๐—ฎ๐—ฟ๐—ฐ๐—ต ๐—ฎ๐—ป๐—ฑ ๐—–๐—ฎ๐˜€๐—ฒ ๐—ฆ๐˜๐˜‚๐—ฑ๐˜† ๐—–๐—ฎ๐—ป ๐— ๐—ฎ๐—ธ๐—ฒ ๐—ผ๐—ฟ ๐—•๐—ฟ๐—ฒ๐—ฎ๐—ธ ๐—ฌ๐—ผ๐˜‚๐—ฟ ๐—ง๐—ต๐—ฒ๐˜€๐—ถ๐˜€.

Many graduate students weaken their thesis by confusing ๐—ฎ๐—ฐ๐˜๐—ถ๐—ผ๐—ป ๐—ฟ๐—ฒ๐˜€๐—ฒ๐—ฎ๐—ฟ๐—ฐ๐—ต with ๐—ฐ๐—ฎ๐˜€๐—ฒ ๐˜€๐˜๐˜‚๐—ฑ๐˜†โ€”yet the two serve fundamentally different academic purposes.

๐—”๐—ฐ๐˜๐—ถ๐—ผ๐—ป ๐—ฟ๐—ฒ๐˜€๐—ฒ๐—ฎ๐—ฟ๐—ฐ๐—ต is initiated to solve an ๐—ถ๐—บ๐—บ๐—ฒ๐—ฑ๐—ถ๐—ฎ๐˜๐—ฒ ๐—ฝ๐—ฟ๐—ผ๐—ฏ๐—น๐—ฒ๐—บ It focuses on ๐—ถ๐—บ๐—ฝ๐—น๐—ฒ๐—บ๐—ฒ๐—ป๐˜๐—ถ๐—ป๐—ด ๐˜€๐—ผ๐—น๐˜‚๐˜๐—ถ๐—ผ๐—ป๐˜€, often within the ๐—ณ๐—ถ๐—ฒ๐—น๐—ฑ ๐—ผ๐—ณ ๐—ฒ๐—ฑ๐˜‚๐—ฐ๐—ฎ๐˜๐—ถ๐—ผ๐—ป, where researchers may also ๐—ฎ๐—ฐ๐˜ ๐—ฎ๐˜€ ๐—ฝ๐—ฎ๐—ฟ๐˜๐—ถ๐—ฐ๐—ถ๐—ฝ๐—ฎ๐—ป๐˜๐˜€ in the research process. This approach is practical, intervention-based, and solution-oriented.

๐—–๐—ฎ๐˜€๐—ฒ ๐˜€๐˜๐˜‚๐—ฑ๐˜†, by contrast, involves ๐—ถ๐—ป-๐—ฑ๐—ฒ๐—ฝ๐˜๐—ต ๐—ฎ๐—ป๐—ฎ๐—น๐˜†๐˜€๐—ถ๐˜€ of a ๐—ฝ๐—ฎ๐—ฟ๐˜๐—ถ๐—ฐ๐˜‚๐—น๐—ฎ๐—ฟ ๐—ฒ๐˜ƒ๐—ฒ๐—ป๐˜ ๐—ผ๐—ฟ ๐—ฐ๐—ฎ๐˜€๐—ฒ ๐—ผ๐˜ƒ๐—ฒ๐—ฟ ๐—ฎ ๐—น๐—ผ๐—ป๐—ด ๐—ฝ๐—ฒ๐—ฟ๐—ถ๐—ผ๐—ฑ ๐—ผ๐—ณ ๐˜๐—ถ๐—บ๐—ฒ. It emphasizes ๐—ผ๐—ฏ๐˜€๐—ฒ๐—ฟ๐˜ƒ๐—ถ๐—ป๐—ด ๐—ฎ๐—ป๐—ฑ ๐—ฎ๐—ป๐—ฎ๐—น๐˜†๐˜€๐—ถ๐—ป๐—ด ๐—ฎ ๐˜€๐—ถ๐˜๐˜‚๐—ฎ๐˜๐—ถ๐—ผ๐—ป, is ๐˜‚๐˜€๐—ฒ๐—ฑ ๐—ถ๐—ป ๐—บ๐—ฎ๐—ป๐˜† ๐—ณ๐—ถ๐—ฒ๐—น๐—ฑ๐˜€, and ๐—ฑ๐—ผ๐—ฒ๐˜€ ๐—ป๐—ผ๐˜ ๐—ฝ๐—ฟ๐—ผ๐˜ƒ๐—ถ๐—ฑ๐—ฒ ๐—ฎ ๐˜€๐—ผ๐—น๐˜‚๐˜๐—ถ๐—ผ๐—ป ๐˜๐—ผ ๐—ฎ ๐—ฝ๐—ฟ๐—ผ๐—ฏ๐—น๐—ฒ๐—บ. Researchers typically ๐—ฑ๐—ผ ๐—ป๐—ผ๐˜ ๐˜๐—ฎ๐—ธ๐—ฒ ๐—ฝ๐—ฎ๐—ฟ๐˜ in the research setting.

Misunderstanding this distinction leads to flawed methodology, weak research design, and inconsistent findingsโ€”common issues in rejected proposals.

๐Ÿ“ฒ If you need thesis help, WhatsApp DocAdeson on: +14243487554

โ™ป๏ธ find this useful? follow + like + repost + comment.

#DrAdeson
#AcademicResearch
#ResearchMatters
#ResearchCommunity
#AcademicWriting
#PhDLife
#PostdocLife
#GradSchool
No comment yet.
Scooped by Gilbert C FAURE
April 3, 11:29 AM
Scoop.it!

Reliability of LLMs as medical assistants for the general public: a randomized preregistered study - Nature Medicine | Vitaly Herasevich

Reliability of LLMs as medical assistants for the general public: a randomized preregistered study - Nature Medicine | Vitaly Herasevich | Notebook or My Personal Learning Network | Scoop.it
LLM are ok for medical diagnoses, but AI chatbots for public are not.
LLMs complete the scenarios accurately, correctly identifying conditions in 94.9% of cases and disposition in 56.3% on average. However, participants using the same LLMs identified relevant conditions in fewer than 34.5% of cases and disposition in fewer than 44.2%, both no better than the control group.

https://lnkd.in/gRCrBSkE

#LLM #AI
No comment yet.
Scooped by Gilbert C FAURE
April 2, 1:18 PM
Scoop.it!

Reddit veut scanner votre oeil pour vous laisser écrire un commentaire. On dirait de la science-fiction. Mais non. C'est une annonce officielle du 25 mars 2026. Reddit a un problème. Des millions ...

Reddit veut scanner votre oeil pour vous laisser écrire un commentaire. On dirait de la science-fiction. Mais non. C'est une annonce officielle du 25 mars 2026. Reddit a un problème. Des millions ... | Notebook or My Personal Learning Network | Scoop.it
Reddit veut scanner votre oeil pour vous laisser รฉcrire un commentaire.
On dirait de la science-fiction. Mais non. C'est une annonce officielle du 25 mars 2026.

Reddit a un problรจme. Des millions de faux comptes automatisรฉs envahissent la plateforme. Des programmes qui postent, commentent, likent ร  la place de vrais utilisateurs. Reddit en supprime 100 000 par jour. Et รงa ne suffit plus. Digg, son ancien concurrent, vient de fermer, submergรฉ par les machines.

La solution ? Demander aux comptes suspects de prouver qu'un humain se cache derriรจre. Avec du Face ID, des passkeys, ou mรชme World ID, un systรจme qui scanne votre iris.

Le dรฉtail ร  connaรฎtre : World ID, c'est un projet cofondรฉ par Sam Altman, le CEO d'OpenAI. Le mรชme Sam Altman qui a investi plus de 60 millions de dollars dans Reddit, qui a siรฉgรฉ ร  son conseil d'administration pendant sept ans, et qui possรจde plus d'actions que le CEO de la plateforme. Sa participation vaut plus d'un milliard de dollars. Lรฉger conflit d'intรฉrรชt.

Nous voilร  donc coincรฉs entre deux risques : laisser mourir l'internet ouvert sous les faux comptes, ou le sauver en confiant nos donnรฉes biomรฉtriques au propriรฉtaire de ChatGPT.

Je n'ai pas la solution. Mais si la seule faรงon de prouver qu'on est humain, c'est de sacrifier son anonymat, alors on a un sรฉrieux problรจme de conception. | 36 comments on LinkedIn
No comment yet.
Scooped by Gilbert C FAURE
April 1, 10:39 AM
Scoop.it!

#epidemiology #publichealth #evidencebasedpractice #clinicalresearch #datascience #globalhealth #researchmethods | Collins Ogweno MPH, MSc, PMP

#epidemiology #publichealth #evidencebasedpractice #clinicalresearch #datascience #globalhealth #researchmethods | Collins Ogweno MPH, MSc, PMP | Notebook or My Personal Learning Network | Scoop.it
Not all evidence is created equalโ€”and in public health, that distinction saves lives.

Understanding clinical study designs is the foundation of evidence-based decision-making:

1. Observational studies (cohort, case-control, cross-sectional) help us detect patterns, associations, and disease burdenโ€”critical for surveillance and hypothesis generation.
2. Experimental studies (randomized vs non-randomized) go a step furtherโ€”establishing causality through controlled intervention.

But hereโ€™s the nuance many overlook:

โœ”๏ธ Cohort studies track exposure โ†’ outcome (powerful for incidence & risk)
โœ”๏ธ Case-control studies work backward (efficient for rare diseases)
โœ”๏ธ Cross-sectional studies provide snapshots (essential for prevalence)
โœ”๏ธ Randomized trials minimize bias and remain the gold standardโ€”but are not always feasible in real-world public health settings

The real expertise lies not in choosing the โ€œbestโ€ designโ€”but in choosing the right design for the right question.

In an era of data abundance and rapid policy decisions, strengthening our understanding of study designs is not optionalโ€”it is a professional responsibility.

#Epidemiology #PublicHealth #EvidenceBasedPractice #ClinicalResearch #DataScience #GlobalHealth #ResearchMethods
No comment yet.
Scooped by Gilbert C FAURE
April 1, 6:12 AM
Scoop.it!

Can you really publish 20-40 articles a day?

Can you really publish 20-40 articles a day? | Notebook or My Personal Learning Network | Scoop.it
Introducing my new white paper: The myth of the academic superstar - or why name disambiguation is crucial
No comment yet.