In this era of AI excitement, how do we ensure surgeons' expectations are appropriately aligned with the actual capabilities of new surgical AI tools? In @JAMASurgery, @Laparoscopes & I comment on some important dimensions of user calibration to new surgical AI systems.
Collaborative project b/w @pcasolab@laparoscopes and @kostasPenn has been awarded 1 of 4 Discovering the Future of AI awards and will use #AI to reconstruct four-dimensional surgical environments from video to assess skill, predict risk, and improve patient outcomes.
More information at:
ai.upenn.edu/discovering-fut…
So excited for the 6th Annual William H. Pearce, MD Research Symposium tomorrow, showcasing the research done by our amazing trainees this year, a talk by Edelstone-Bendix Visiting Professor @laparoscopes, and the graduating chiefs residents!
It was such an incredible privilege to have been asked to give a keynote at the 125th Japan Surgical Society & to cover how #computervision#robotics#endoscopy will change surgery
Thanks @pennsurgery@PCASOLab@GRASPlab for support of multidisciplinary CS+surgery research
Great job by @pennsurgery@PCASOLab Harrison scholar @shubha_vasisht on her @AmCollSurgeons#ACSCC24 oral presentation on PPI use in patients who underwent POEM vs Heller
Great job presenting as a premed student!!!
Fyi med schools, she's applying this year!
Many people believe that AI advances will dramatically increase inequality.
In a paper with two Nobel laureates, Daron Acemoglu and Simon Johnson, plus 30 multidisciplinary experts, we argue that it’s more complex than a simple “rich-get-richer” story.
For example, we coined the term “inverse skill bias” to describe an emerging pattern: generative AI seems to benefit low-skilled workers more than high-skilled ones.
We also suggest generative AI may reduce racial and gender bias in healthcare and education.
However, some inequalities could indeed worsen.
For example, companies with access to more data may gain an anticompetitive advantage, exerting market power over smaller firms. Additionally, companies may be incentivized to automate work rather than invest in enhancing and complementing human capabilities.
Gender bias in career achievement may also worsen, as preliminary evidence shows that men are using chatbots more than women, leading to an increase in productivity among men but not women.
We argue institutions will play a critical role in sharing AI’s benefits equitably. Unfortunately, current regulations fall short of addressing inequalities and fostering shared prosperity.
Our paper ends with six policy suggestions we believe can help reduce socioeconomic inequality:
1) Create a more balanced tax structure, equating marginal taxes on hiring, training, and AI investments.
2) Engage workers and civil society in AI shifts, and establish data unions for control over data.
3) Boost support for research into human-complementary AI tools to enhance productivity and skillsets.
4) Train professionals, especially in healthcare and education, in AI use, including ethical aspects.
5) Invest in tools to counter AI-generated misinformation and in education on misinformation.
6) Embed AI expertise in government for sector-wide decision support.
Read the full paper here: academic.oup.com/pnasnexus/a…
Thank you to an amazing list of coauthors, without whom this work wouldn’t have been possible:
@AustinLentsch@DAcemogluMIT@SelinAkgun9 Aisel Akhmedova @EBilancini @JFBonnefon @BehSnaps @lu_butera@Karen_Douglas@JimACEverett Gerd Gigerenzer @chrisgreenhow@Laparoscopes@PCASOLab@jholtlunstad@jetten_j@baselinescene@werkunz@longoni_chiara Pete Lunn @simone_natale Stefanie Paluch @iyadrahwan Neil Selwyn @viveksinghmed@ssuri Jennifer Sutcliffe @JoePTomlinson @Sander_vdLinden@PaulvanLange@FriederikeWall@jayvanbavel Riccardo Viale
Many people believe that AI advances will dramatically increase inequality.
In a paper with two Nobel laureates, Daron Acemoglu and Simon Johnson, plus 30 multidisciplinary experts, we argue that it’s more complex than a simple “rich-get-richer” story.
For example, we coined the term “inverse skill bias” to describe an emerging pattern: generative AI seems to benefit low-skilled workers more than high-skilled ones.
We also suggest generative AI may reduce racial and gender bias in healthcare and education.
However, some inequalities could indeed worsen.
For example, companies with access to more data may gain an anticompetitive advantage, exerting market power over smaller firms. Additionally, companies may be incentivized to automate work rather than invest in enhancing and complementing human capabilities.
Gender bias in career achievement may also worsen, as preliminary evidence shows that men are using chatbots more than women, leading to an increase in productivity among men but not women.
We argue institutions will play a critical role in sharing AI’s benefits equitably. Unfortunately, current regulations fall short of addressing inequalities and fostering shared prosperity.
Our paper ends with six policy suggestions we believe can help reduce socioeconomic inequality:
1) Create a more balanced tax structure, equating marginal taxes on hiring, training, and AI investments.
2) Engage workers and civil society in AI shifts, and establish data unions for control over data.
3) Boost support for research into human-complementary AI tools to enhance productivity and skillsets.
4) Train professionals, especially in healthcare and education, in AI use, including ethical aspects.
5) Invest in tools to counter AI-generated misinformation and in education on misinformation.
6) Embed AI expertise in government for sector-wide decision support.
Read the full paper here: academic.oup.com/pnasnexus/a…
Thank you to an amazing list of coauthors, without whom this work wouldn’t have been possible:
@AustinLentsch@DAcemogluMIT@SelinAkgun9 Aisel Akhmedova @EBilancini @JFBonnefon @BehSnaps @lu_butera@Karen_Douglas@JimACEverett Gerd Gigerenzer @chrisgreenhow@Laparoscopes@PCASOLab@jholtlunstad@jetten_j@baselinescene@werkunz@longoni_chiara Pete Lunn @simone_natale Stefanie Paluch @iyadrahwan Neil Selwyn @viveksinghmed@ssuri Jennifer Sutcliffe @JoePTomlinson @Sander_vdLinden@PaulvanLange@FriederikeWall@jayvanbavel Riccardo Viale