@mdmccradden
Bioethicist specializing in ethical and evidence-based integration of technologies in health care. AI Director, Women’s and Children’s Health Network | THRF Clinical Research Fellow, Australian Institute for Machine Learning. Adelaide, South Australia.
One thing I can absolutely guarantee:
This is not a dispute over “ethics of AI.” It’s not a dispute over “guardrails.”
It’s about power.
www.theguardian.com/us-news/2026...
This work is part of Project CANAIRI and if you’re interested in the silent phase, please get in touch with me to get involved!
www.projectcanairi.org
We advocate for the silent phase as potentially the MOST IMPORTANT stage of AI translation because it gives you specific, local evidence for performance in your setting. And we can do them better than how we are now - this means less wasteful translation and research, and more reliable evidence
Some devices approved by regulators had substantial performance drops in new settings. Thankfully, these papers described SILENT evals - so, they found out the tool was worse than the status quo before they changed patient care!
Major gaps concern failure case analysis, fairness, and clinical verification of outputs.
You know, the stuff that is most important to clinicians calibrating their use of the tool 😳
So what is a “silent trial”?
Apparently, everyone has a different definition (just ask our reviewers 🤭)
We made a taxonomy in Table 1 to promote consistency
What is a silent evaluation in #HealthAI ? And how do you do one?
👀 check out our new paper!
We performed a scoping study re published practices in silent evals for AI/ML tools - everything from hospital-based prediction algos to regulatory-approved medical devices
www.nature.com/articles/s44...
I’m just so shocked that these technologies we were promised would free up our time for relaxing and enjoying life seem to… exacerbate our rapacious productivity obsession?? I am shocked, shocked i tell you! #leopardatemyface
Baffling how many AI-pushers still arguing “well, humans make mistakes all the time!” It’s like, uh yes, well done on that utterly scintillating observation, but perhaps the KIND of errors matters somewhat, no?
Fortunately the handy image provided illustrates a healthy, moderate pour for your reference
“Okay, but like… why?”
I wonder if folks will start caring when they realise their h indexes will take a hit from failing to cite their real papers? #VibeCiting #AIEthics
www.theregister.com/2026/01/22/n...
By contrast, is a pharma company was talking about how their new drug worked really hard to pass phase 1, 2, 3 trials to learn how to be a good little drug and help those humans…
Heading into December with energy matched by this koala…
Placed my order! Can’t wait to read, and congrats @carissaveliz.bsky.social !
Drumrolls, please! I’m thrilled to share the cover for my forthcoming #book from @doubledaybooks.bsky.social
(Penguin Random House), #PROPHECY. This is the best book I’ve written: the boldest, the most innovative, the most personal. I have poured my soul into it like never before. 1/
Unpopular opinion: claiming that“human-in-the-loop” is your governance of AI is the equivalent of claiming drinking orange juice is your governance of not getting a cold
it is worth remembering that cigarette companies also spent decades selling products they knew would harm and kill people and marketed them specifically to young people to create dependence. let's stop this without having to go through a generation of death and several class actions.
I move for immediate approval to cancel phrases such as “AI pioneers/godfathers,” and all variants
Another casualty of the LLM revolution.
In "arXiv’s CS category, review articles and position papers must now be accepted at a journal or a conference and complete successful peer review."
Today, in “Bad things that were 100% completely and totally foreseeable and they just let er rip” news
www.msn.com/en-gb/public...
Oh wow. This is the very definition of facepalm 🤦🏼♀️
The worse part though is that as their costs go up for adding these nonsense features, they inevitably offload it onto us, the consumers. 😡 I’m now looking into old school options again to get away from this stuff.
And the fact that they even think this is possible shows they have absolutely zero understanding of any concepts and principles related to “bias” 🤨 If they cannot understand the very concepts they invoke, then they have zero credibility to address the problem and should sit this one out.
Not arrogant - 100% accurate. The very fact that these folks can switch up their buzz words on a dime shows they don’t understand the words in the first place.
I've yet to see a single discussion of "Canadian digital sovereignty" include any Indigenous experts, at best we're given a token mention that at some point they'll "consult with Indigenous Peoples" like we're some sort of monolith. Umm, y'all need to consider all our lands, and our distinctiveness.
Gonna pump my own paper here which is basically saying it’s the wrong problem to focus on for making good decisions in medicine 😅
www.sciencedirect.com/science/arti...
We’ve also pre-registered our study on OSF to encourage transparency and sharing with the scientific community
osf.io/p6tm5
This shared accountability is important to limiting the expanding problem of moral crumple zones, per -
estsjournal.org/index.php/es...
