"ChatGPT Better Than Dr. Google." Former Minister on the Future of Self-Diagnosis on the Internet
The guest of the Health Market podcast "Stechtoskop" was Janusz Cieszyński , MP from Law and Justice, previously Minister of Digital Affairs and Deputy Minister of Health.
We provocatively asked him what is better for checking disease symptoms on the internet: an internet search engine (e.g. Google ) or an artificial intelligence-supported language model (e.g. ChatGPT )?
He replied that he thought ChatGPT was the best choice because it allows for conversations with the tool, which is more natural for someone wanting to check their health. Meanwhile, the search engine is "static"—you have to navigate the maze of information contained within individual links.
Janusz Cieszyński noted that both tools have serious limitations. Primarily, they can provide information that is inconsistent with established medical knowledge, or even downright inaccurate. Despite this, the former minister believes that the future of healthcare lies in the increasing use of language models.
And this, according to our interlocutor, could have interesting consequences. Today, such modern technologies are used, for example, in call centers, where consultants' salaries are not very high. Meanwhile, in healthcare, especially when it comes to doctors, we are dealing with the replacement of some of the work of highly qualified individuals, who, consequently, receive very good salaries.
"The benefits will be much greater. Not to mention the fact that there are no queues at the machine and you can contact it at any time of day or night," praised Janusz Cieszyński.
We asked the former minister if he was concerned that language models would use information from questionable sources and thus become an uncontrolled tool that could spread medical disinformation. And that the very design of language models could cause patients who use them to suggest the AI the answer they want.
According to Cieszyński, "the world's best brains" are working to ensure that the results presented by language models are based on the best possible sources. Our interlocutor emphasized that he is far from techno-optimistic, believing that models will solve all our problems for us and that humans will be unnecessary.
Currently, attempts to commercially offer language models in the healthcare sector raise significant legal concerns (primarily regarding personal data protection and the lack of certification for such tools). In Rynek Zdrowia , we reported on the MyMediScan service, which purports to offer modern, free analysis of laboratory tests using AI. However, its declarations raised serious concerns from the Office for Personal Data Protection and lawyers, primarily regarding ensuring user privacy. Following our inquiries, the website promptly suspended its operations.
The entire interview with Janusz Cieszyński can be viewed at Rynek Zdrowia.
All episodes of the "Stechtoskop" podcast are available for listening on Spotify.
Write to the author: [email protected]
Copyrighted material - reprint rules are specified in the regulations .
rynekzdrowia
