I agree. Since your article 5 years ago, we have built massive high-fidelity neural encodings of text and images. And transformers can draw from the state at any point in a sequence. But we need more work toward learning/encoding logical and probabilistic
RT @cabitzaf: @curtlanglotz And context. In a JAMA piece (2017) we basically made the Cassandra point that we face the risk of a "demise of…
RT @cabitzaf: @curtlanglotz And context. In a JAMA piece (2017) we basically made the Cassandra point that we face the risk of a "demise of…
@curtlanglotz And context. In a JAMA piece (2017) we basically made the Cassandra point that we face the risk of a "demise of context" in medical reasoning, if we think that everything needed to make a diagnosis can be traced/reduced to text. And LLMs see
@mireillemoret @andrewrens_ria @bernardionysius @davidgraeber Indeed! However that article is more about the deskilling as degradation of expertise *required* in a profession due to its automation. Deskilling as loss of skills in technology-supported deci
@AllisonCGardner @ncganderup @IntHealthAI We wrote this in 2017 on @JAMA_current, and doing so has drawn us some criticism (see related letters to the editor) (https://t.co/iM6PoEQ5H1). However, the question is: how to identify this effect? Few studies ha
@JohnSumser @vendorprisey @cabitzaf And much earlier in brilliant short discussion of many of the assumptions that simply don’t fly when applied without proper awareness https://t.co/cJxPXlh0RR, see also https://t.co/qjQQli2pLW 4/
Talking about the unexpected consequences of IT is hard and often self-harming. I know something about it: little impact for my 2017 piece on JAMA (https://t.co/iM6PoEQ5H1) and a chapter I wrote on the so-called 'Cobra AI' has been under review for 1 year
RT @CMichaelGibson: Unintended Consequences of Machine Learning in Medicine | JAMA https://t.co/EIgjUdB49H
At the end of the day, we demand that "magic" of randomisation in impact studies. https://t.co/JZag8XhCgb https://t.co/wRLqQZJdes
Unintended Consequences of Machine Learning in Medicine [Jul 20, 2017] Cabitza et al. @JAMA_current https://t.co/ICCsPnLRNM #DeepLearning
RT @mtmdphd: Unintended Consequences of Machine Learning in Medicine [Jul 20, 2017] Cabitza et al. @JAMA_current https://t.co/ICCsPnLRNM #D…
Unintended Consequences of Machine Learning in Medicine [Jul 20, 2017] Cabitza et al. @JAMA_current https://t.co/ICCsPnLRNM #DeepLearning
@ginorthshore @RogueRad We can take an educated guess in the meantime. https://t.co/iM6PoEQ5H1
RT @CAM_Paddison: Another interesting paper here on the potential unintended consequences of machine learning https://t.co/GckgXRXJ5U HT @R…
RT @CAM_Paddison: Another interesting paper here on the potential unintended consequences of machine learning https://t.co/GckgXRXJ5U HT @R…
Another interesting paper here on the potential unintended consequences of machine learning https://t.co/GckgXRXJ5U HT @RasoiniR
@CAM_Paddison I agree. We also wrote about some potential pitfalls of #AI. https://t.co/ix86pdiVUC
@kncukier And yet, I'm wondering if words like impact, limitations and challenges (occurring several times in that piece) don't end up by sweetening concepts that would have been better rendered as caution, prudence, shortcomings, drawbacks. But I'm probab
Unintended Consequences of Machine Learning in Medicine | Clinical Decision Support | JAMA | JAMA Network 🔗 https://t.co/d7mwdHmQal #artificialintelligence
Unintended Consequences of Machine Learning in Medicine [7/20/17] Cabitza et al. @JAMA_current https://t.co/ICCsPnLRNM #DeepLearning
RT @cabitzaf: Getting a glimpse of the dark side of #AI and #MachineLearning in medicine on JAMA: https://t.co/qsuvOoH3Uz https://t.co/Hd6h…
RT @cabitzaf: If only Cassandra had had less passion and a better strategy.... /n https://t.co/iM6PoEQ5H1
If only Cassandra had had less passion and a better strategy.... /n https://t.co/iM6PoEQ5H1
Well articulated and important concerns about some "Unintended Consequences of Machine Learning in Medicine". Take in mind: Medicine depends a lot on experience and tacit knowledge of doctors! https://t.co/U9UWKITxJA
Well articulated and important concerns about some "Unintended Consequences of Machine Learning in Medicine". Take in mind: Medicine depends a lot on experience and tacit knowledge of doctors!
RT @Rad_Nation: Selection bias is a consideration for studies. Is algorithmic bias a risk for #AI in #RadOnc? https://t.co/kDwtij8IUC https…
RT @Rad_Nation: Selection bias is a consideration for studies. Is algorithmic bias a risk for #AI in #RadOnc? https://t.co/kDwtij8IUC https…
RT @Rad_Nation: Selection bias is a consideration for studies. Is algorithmic bias a risk for #AI in #RadOnc? https://t.co/kDwtij8IUC https…
RT @Rad_Nation: Selection bias is a consideration for studies. Is algorithmic bias a risk for #AI in #RadOnc? https://t.co/kDwtij8IUC https…
Selection bias is a consideration for studies. Is algorithmic bias a risk for #AI in #RadOnc? https://t.co/kDwtij8IUC https://t.co/BZsynKhIom #jc #bias T3
RT @cabitzaf: Getting a glimpse of the dark side of #AI and #MachineLearning in medicine on JAMA: https://t.co/qsuvOoH3Uz https://t.co/Hd6h…
RT @raymondpeck3: Points about quality & validity of data, inc. hidden confounders, are relevant to many #machinelearning applications http…
@azeem @tabithagold @joannashields @DrHughHarvey @MattHancock You decide if this research line fits your interests, by me and @RasoiniR. https://t.co/JTRiP6bjHp
RT @cabitzaf: Getting a glimpse of the dark side of #AI and #MachineLearning in medicine on JAMA: https://t.co/qsuvOoH3Uz https://t.co/Hd6h…
Unintended Consequences of Machine Learning in Medicine https://t.co/jIkRDfu9tn
@RslewisSally Definitely agree. We wrote about this here: https://t.co/5VwmgU1RHE
@DrLukeOR @EnricoCoiera That would be a critical signal to detect as soon as it emerges from AI-supported clinical practices, Luke. One year ago we speculated on JAMA that AB could induce forms of deskilling and this is still the most controversial and dis
RT @pash22: Unintended Consequences of Machine Learning in Medicine https://t.co/cKKiegVJla via @cabitzaf & @JAMA_current
Coming to a clinic near you - #Health3: Rise of the Machines https://t.co/K9vmvzyLx4
Unintended Consequences of Machine Learning in Medicine https://t.co/cKKiegVJla via @cabitzaf & @JAMA_current
RT @cabitzaf: Getting a glimpse of the dark side of #AI and #MachineLearning in medicine on JAMA: https://t.co/qsuvOoH3Uz https://t.co/Hd6h…
Technology is not a panacea for a health care system in crisis. https://t.co/ChwauSlxoa
RT @cabitzaf: Getting a glimpse of the dark side of #AI and #MachineLearning in medicine on JAMA: https://t.co/qsuvOoH3Uz https://t.co/Hd6h…
RT @UREssien: My initial thoughts would be that machine learning algorithms would reduce provider bias and ultimately disparities in care.…
My initial thoughts would be that machine learning algorithms would reduce provider bias and ultimately disparities in care. Not sure how much this has been studied. https://t.co/WJVqmoergJ
RT @eugeniosantoro: Dequalilificazione, deskilling e unintended consequence of machine learning in medicine. I temi di @cabitzaf a #4Words1…
RT @eugeniosantoro: Dequalilificazione, deskilling e unintended consequence of machine learning in medicine. I temi di @cabitzaf a #4Words1…
RT @eugeniosantoro: Dequalilificazione, deskilling e unintended consequence of machine learning in medicine. I temi di @cabitzaf a #4Words1…
Dequalilificazione, deskilling e unintended consequence of machine learning in medicine. I temi di @cabitzaf a #4Words18 https://t.co/4DMyMYARzz
RT @Vincenzo__Marra: Retweeted Federico Cabitza (@cabitzaf): Getting a glimpse of the dark side of #AI and #MachineLearning in... https://…
Retweeted Federico Cabitza (@cabitzaf): Getting a glimpse of the dark side of #AI and #MachineLearning in... https://t.co/Hdr6nlXyaW
RT @cabitzaf: Getting a glimpse of the dark side of #AI and #MachineLearning in medicine on JAMA: https://t.co/qsuvOoH3Uz https://t.co/Hd6h…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @cabitzaf: Another question I'd love to talk about with you in Rome at #4Words2018! (I'm counting the minutes! :) @G__Recchia @oziadias…
RT @cabitzaf: Another question I'd love to talk about with you in Rome at #4Words2018! (I'm counting the minutes! :) @G__Recchia @oziadias…
RT @cabitzaf: Another question I'd love to talk about with you in Rome at #4Words2018! (I'm counting the minutes! :) @G__Recchia @oziadias…
RT @cabitzaf: Another question I'd love to talk about with you in Rome at #4Words2018! (I'm counting the minutes! :) @G__Recchia @oziadias…
RT @cabitzaf: Another question I'd love to talk about with you in Rome at #4Words2018! (I'm counting the minutes! :) @G__Recchia @oziadias…
Another question I'd love to talk about with you in Rome at #4Words2018! (I'm counting the minutes! :) @G__Recchia @oziadias @trishgreenhalgh https://t.co/F2ypQQYTYV
@techreview Uh oh, someone is beginning to take the deskilling argument seriously. And the best (?) is yet to come! https://t.co/O8k45sUbbS
Machine learning and artificial intelligence are not miracle solutions to healthcare challenges - Doctors and machines must collaborate to reach the best diagnosis. Brought to you by #PraavaHealth - Bangladesh's first Family Health Center https://t.co/E9
While healthcare has been slow to adopt AI, it’s important to understand that machine learning and artificial intelligence are not a miracle solution on their own. Doctors and machines must work in tandem to reach the best decision. https://t.co/sxRVLI6Fn
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
良くある議論ですが、盛り上がり! https://t.co/on2AoTKn52
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
@jama @erictopol Everything is a piece of the puzzle. Nothing works alone. Does it make sense to the physician? To the patient? To the subject matter expert? Data does not speak truth inherently--does it fit in the context? https://t.co/3Jghtmvi4q
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…
RT @EricTopol: A @JAMA_current negative viewpoint on #AI in medicine gets a backlash of letters https://t.co/9eBu70nJNR https://t.co/Z7dJ3G…