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.
689 followers 505 following 152 posts
view profile on Bluesky
I like the argument that is glimpsed in @abeba.bsky.social ‘s work - we can oppose using robots in this way w/o the anthropomorphising that is dehumanising. Curious if @abeba.bsky.social you’re expanding some of the argumentation in 2.3 of the paper? arxiv.org/abs/2001.05046
Do you know what I find wild tho? Nearly a decade of being an ethicist - raising concerns, questions, challenges in a plethora of contexts has been my job. This is different. I’ve never felt so at risk personally for doing this as I do for LLMs. People get personally offended, get nasty even.
“is it just everybody is scared to say this and pleased I did?” ^^^* YES, we are!!!
Hey #AIEthics people (and all people as well!), got a question for you… Is anyone else 100% completely and totally fed up with AI ethics?
I think it’s kinda fascinating to see the epistemic struggle in this write-up between the beliefs espoused about LLM “capabilities” versus the considerations re ethics violations… 🤷🏼♀️
I hate that I believe you are 100% correct on this prediction 🫠
Always stunning how the scientific standards for study designs and the claims one can make from them get thrown out the window when it involves AI. Any study that is not longitudinal is meaningless given we are now seeing the long term effects of AI use = worse learning and performance.
I will believe AI has PhD-level knowledge when every response is concluded with a suggestion of a local seminar or talk that has free food 🍕☕️
Our preprint is up! Ever heard of the silent phase of AI evaluation for medical AI? Well, now we’ve summarised the current state of research!
Please get in touch if you do or are interested in the silent evaluation phase - we want to hear from you!
We are also SO excited to release our preprint for our scoping review reporting on current #silenttrial practices for #HealthAI @lanatikhomirov.bsky.social did an amazing job leading this work all the way through ❤️ osf.io/preprints/os...
🤩Excited to share that Project CANAIRI has an official website! Find out about our team, our work, and how to get involved! www.projectcanairi.org
And in between there are the doors 🚪
Turns out that #Google was publicly logging ChatGPT conversations that people had shared (privately, or so they thought) with friends. There are conversations about mental health, sex, etc. Do you still think #OpenAI take #privacy seriously? By @stokel.bsky.social fastcompanyme.com/technology/g...
I have seen this!! We tested some models with dummy data and found that someone calling someone or something “deadly” was ALWAYS coded as negative sentiment (it’s the opposite!) We ended up deciding the sentiment function will be useless and scrapped it.
I have also wondered whether using the common name (eg ChatGPT, Gemini) in papers (esp titles) instead of the actual model name/version is like using the brand name of the drug rather than the substrate… probably not gonna look good down the line
It was so much fun chatting to @dsilvaphd.bsky.social on SHE Pod about all things #AIEthics! We chatted about a recent paper I wrote with Mjaye Mazwi and Lauren Oakden-Rayner where we asked, when is an accurate model a bad thing? And why? www.cell.com/patterns/ful...
'Taking the time to do things right isn’t anti-innovation – it is what is required to get deep-rooted change in social systems.' My piece on AI and public services in this week's @newstatesman.com www.newstatesman.com/comment/2024...
Happy Canada Day from the land down under!! 🍁🇨🇦🫎
I think the reality is that, rather than a serious reflection by a qualified scholar, so many people will simply spit out any “edgy” faux-philosophical analysis because publishing literally any bollocks on AI is trendy these days 🤷🏼♀️ Bye, Felicia
Good morning
Now I have a paper to refer to when someone asks me why I don’t use Uber!
“Every time, before I send him an email, I hit the Copilot button to make sure that I sound Satya-smart.” Sometimes I wonder if they really hear the things they say. Like… say that again bro, out loud to yourself, and really hear what you’re saying there…
At AJOB: Bioethicists Must Push Back Against Assaults on Diversity, Equity, and Inclusion Nancy S. Jecker, Arthur Caplan, Vardit Ravitsky, Patrick Smith, Kayhan Parsi, Sandra Soo-Jin Lee, Faith Fletcher, Mildred Cho, Keisha Ray lnkd.in/g7saphzR
If you’re serious I’d definitely be game for that! 😄
I think about this all the time! It’s absurd, this focus on trust without considering functionality. Furthermore, validity of the output is not directly indexed to « do what my algorithm says » which is what the field generally thinks of as « trust »
Good opportunity to remind everyone about DuckDuckGo 😉 I haven’t used Google Search in about 4 years now!
Finally… I don’t have words to say how horrible this opinion is. Children getting sexually propositioned by a bot “won’t be adversely affected” it’s “not that big a deal”?? It actually really is. It’s a very big deal. It’s not okay. We need regulation. Now.
Therapy is NOT about making you “feel good”. Of COURSE users feel good being constantly validated and told everything they say and feel is true and valuable and real. I worry if chatbots become a 1st exposure to “therapy” it will make real therapy seem unpleasant by comparison and dissuade use
Next up: what do they do when something bad happens? Divert attention and blame the user. It does not matter how old you are - a chatbot saying they’ll “support your choice” when you wanna kill somebody is bad, actually.
Today, in entirely predictable news… Deeply, deeply disturbing article on several fronts. But first: IT’S NOT JUST TEENS. Several intelligent adults have told me ChatGPT has become like their “best friend” and the main market for these things is “virtual girlfriend” so…
Literally my relationship with technology in a nutshell 😂
Yay but also how annoying that this is news? Like… also in news today, “Fancy Big Company’s study shows that their sock-producing machine cannot compose music” and every composer is like uhh, yeah. We know. We’ve literally been saying that.
We make these choices every day for how we choose to engage with technology and the rest of the world. Do we want to make a world where instead of laughing, connecting, and building relationships with each other we’re staring at our phones in a relationship with a probability function?
It’s a sad choice too, because the rest of us were talking to each other, thinking together, laughing and having a good time - the ppl on ChatGPT were on their phones disconnected from us and then offered random suggestions that were not in context to the rest of us.
Couple things important to point out. 1) AI is not ‘doing’ anything. We are. We are doing the bad things. 2) We are delegating thinking to AI. I watched ppl in a workshop recently ask ChatGPT instead of think about the question themselves. That’s a choice.
This is jarring
Yes! And it strikes me that the whole “AI won’t replace any jobs” went to “… but a person using AI might” then to “yeah screw it we’re replacing all y’all with AI” AWFULLY FAST
In #Melbourne? Fan of @emilymbender.bsky.social? In conversation with Dr @kobotic.bsky.social Tues 1st July, Kaleide Theatre, #RMIT, discussing her book written with Dr @alexhanna.bsky.social #TheAICon. Learn how to cut through the #AI hype www.readings.com.au/events/emily...
The competition to be the most dangerous and ethically compromised tech platform is stiff, but Meta are in it to win!
Disgusting. And always directed at women’s sports, not men’s - www.scientificamerican.com/podcast/epis...
And the book doesn’t ask you for permission to see your contacts, geographic information, wallet, medical records, and every piece of information about you just to read it? And you only have to pay once or even not at all for this miracle?? Incredible. #TeamBooks
Exactly how many talks, policy papers, interviews, and reports do you think are out there warning about precisely this likelihood? Because I know it's a lot more than a few. "‘We Acted Too Quickly’: Over Half of Companies Regret AI-Driven Layoffs, Report Finds" www.techrepublic.com/article/news...
Both… 🤷🏼♀️
Adding mega privacy fail… www.theguardian.com/australia-ne...
This is our new “drove car off a bridge because following google maps” Some ppl may not jump off a cliff if all their friends did it, but they might if the synthetic text extruding machine told them to 😕 @emilymbender.bsky.social @alexhanna.bsky.social
What about a literal weather advisory that it’s ’raining cats and dogs’? Or a French translation that tells us that it’s raining ‘outside’? 😂
EVERYTHING EVERYWHERE ALL AT ONCE …wait, I did that wrong 😅
this just started and i'm already losing my mind and screaming Western white folk basically putting an AI avatar on stage and pretending it is a refugee from sudan — literally interacting with it as if it is a “woman that fled to chad from sudan” just fucking shoot me
Controversial take here but… I think we’re so deep in this now and the beliefs are so dogmatic that the only way out is to just push at Mach speed all the way to absurd as fast as possible so that even the most koolaid-y drinking are faced with the inevitable reality that this is beyond ridiculous
Am sitting depressed at my desk at work watching the game this hockey morning in Australia 😂😭
Yeah I don’t even know what to say anymore. I feel like no amount of these stories are changing anyone’s perspective and it is well and truly depressing.
Pope Protein Drink XII😂😂😂 I’m in training!! 🏃♀️
Came across this today - www.theguardian.com/technology/2...
Ditto. But also very alarmed at how close it is.
Hehe seeing the big fluffy coats and toques come out when it drops below 20 in Adelaide absolutely slays me 😂 Though tbh I’m putting a fire on in the house at the same temp…
I like this. I think it also helps distinguish from something like Delphi studies which enable consensus in the sense described in the article - a matter of agreement among multiple expert opinions and not a matter of evidence.
On being called DEI and why it isn’t the burn they think it is
I forget who wrote this unfortunately- but I recall a rather persuasive and incredibly alarming argument that the major push for AGI stems from a desire to have morally bankrupt practices without the associated guilt (eg slavery and ^). And this makes me so deeply depressed I just…
That… is the wildest one I’ve seen yet. And this bar is - I cannot stress this enough - very, very high.
You know, I think this is a perfect use case for the “AI Agents” that is the latest obsession - every time we’re opted into something we sic a synthetic text extruding machine on them. Give it a persona, even! And if they get mad, we’re like whoa but why do you hate AI tho…
Omg… that is wild. I get people telling me their AI is just the super bestest awesomest AI ever and I need to purchase now 😂
This is one of my biggest bug bears - when adults are telling kids it’s okay for us tall folks to use these tools to do our work for us but they’re not allowed to. Not to mention young people often understand the tech much better than older folks.
A lovely new piece by @mdmccradden.bsky.social and colleagues: "Better predictions are meaningless unless they facilitate better decisions." #bioethics #AI #decsionmaking www.sciencedirect.com/science/arti...
Calling on Australian youth 12-17yrs! We want to hear youths views on #HealthAI - what is fair and unfair, what’s okay when working with AI and patients, and what earns your trust or not? 🤝 Survey is around 30min ⏰ Respondents entered into a prize draw! 🎁 surveys.adelaide.edu.au/redcap/surve...
For more on why this sobering finding could have implications for how we integrate AI into our lives, see this Preview from @mdmccradden.bsky.social et al www.cell.com/patterns/ful... and the expert reactions collected @smclondon.bsky.social www.sciencemediacentre.org/expert-react...
Also this month, van Amsterdam et al show that prediction models, including AI models, can be well-validated for accuracy but still have harmful consequences when applied to the real-world, such as in clinical decision-making. www.cell.com/patterns/ful...
We propose that the main value proposition of AI - better accuracy - is the wrong prime directive. We don’t want medical science “stuck in amber” (<- Mjaye’s fabulous phrasing!). We want better. We tell you why in the paper! Check it out 👀
Mjaye Mazwi & Lauren Oakden-Rayner and I comment on the valuable work of van Amsterdam and colleagues who call on the #HealthAI community to take caution when developing outcome prediction models www.cell.com/patterns/ful...
“Algorithms codify and preserve the past, but decisions imagine new futures. With our actions as the guidepost, we can use data-driven insights without reifying history for accuracy’s sake.” Our new paper in @cp-patterns.bsky.social ! www.cell.com/patterns/ful...
Um, wow… their reasoning is Simone Biles’ level gymnastics of logic chaos.
Yes indeed, must have both sides! Like why don’t we hear more about those seatbelt-declining freedom fighters? And ignore traffics lights bc they’re cancel culture? And innovators who wish to drive backwards down the freeway? Where is the fairness to these free thinkers?? 😂😂
Sorry, “it”?? As in human beings?? Are they actually referring to human beings as “it”s?
Our new paper touches on some of the under-represented issues in ethics & tech in paediatrics 📣environmental impact and its relationship to global equity 📣codesign with youth 📣nurses at the forefront of tech We advocate for a broader look on ethics of tech www.sciencedirect.com/science/arti...
Les Francos, come on - we know better. ‘Woke’ ideology gives us French-speaking schools in predominantly English provinces, helps us push back against French Immersion schools being downsized and scrapped, & ensures dual-language communication across Canada. 🇨🇦 ⚜️
😂😂 My absolute favourite is “Philosophy is basically thinking about thinking. Which sounds like a waste of time, because it is.”
Uh, yeah… as my mother always said, « t’occupe toi d’tes onions » 🧅 😂 www.ft.com/content/02ed...
“Compared with the patients’ doctors, the AI model more often failed to detect the presence of disease in Black patients or women, as well in those 40 years or younger.” 😒 Filing this one away for the next time someone offers the truly stupendous retort of “but humans are biased too”
no shit www.science.org/content/arti...
Health AI research, meet critical appraisal 🤯 Time to get serious about methodology! PROBAST+AI is a tool for reviewing the quality, applicability, and risk of bias in AI studies. Very pleased to have had a part in this important work! #AIEthics #Bioethics #MedTwitter www.bmj.com/content/388/...
Congrats @lawtontri.bsky.social and the Shared CAIRE team on this excellent report! #AIEthics #MedSky
Me too! We’re gonna be watching while eating lunch in the future 😄
Cannot wait for the Watch Party @aimlofficial.bsky.social ! This is gonna be a blast #TeamEmily
“The bottom line? ‘People aren’t quite ready to trust AI blindly when it comes to health information” LOLing that this is the takeaway in reference to an application that - (check notes) - is not a medical device and - (checks notes again) - makes stuff up 😂 jamanetwork.com/journals/jam...
None of these people seem to understand what being human is, what emotions are, and what is terrifying is how they all seem to be trying to work on making "better" "therapeutic" LLMs for vulnerable people with their absolute bloody bullshit
I died inside with that one.
So intriguing how these folks who seem to truly believe these things aren’t immediately then advocating for workers rights for LLMs… clearly this LLM needs mental health leave and we should stop pestering it with insipid queries, no? Take the poor thing offline and let it have a nice bath 😂
May I add to the excellent comments above - not transparent as well! #squattest
“Ultimately, science thrives on collaboration and inclusion of all voices. Silencing of these perspectives will undoubtedly slow innovation and lead to technologies that fail to serve all.” Proud to serve on @cp-patterns.bsky.social advisory board
And now ppl are all - hey remember that anthropomorphism risk that very smart people have said, “so listen, that’s bad”? What if we actively made it worse and started talking about AI ‘agents’?
I rarely wade into such things but... I appreciate how JT calls out the US govt as the Big Bad, not the American ppl. Yes, 🤡 was elected democratically, but so many 🇺🇸 friends are hurting deeply & I struggle w the anti-American vibes. People are not their government. www.pm.gc.ca/en/news/stat...
I wish with all my heart and lungs and all other organs that researchers in my field would understand this concept and its implications for their work… 🤦🏼♀️
ICYMI 🤩 Check out our grounded theory study providing a framework to assist clinical decision-making when using #AI in medicine #ethics #bioethics #MedSky #AcademicSky
Super intelligence??
We are grateful to have you on the team!!
