Trust is earned by actions matching words spoken that get the individuals in charge of our healthcare system in the first place. Trust needs to be earned by the programmers of Ai. Trust needs to be earned by all elected officials. Let’s start there.
Generative AI is best explained with the metaphor of being an "automated polling institute". Instead of directing a query at thousands of individuals by conducting interviews using a call center and then generating the statistical consensus from the interview data as response, Generative AI does it through its statistical algorithm from the several thousands/millions of sources (texts, images, videos, sounds). Everytime a chatbot responds with "I think that ..." replace it with "The statistical consensus in the text data suggest the thought that ...". The variations come from a random number embedded in the query (per session) - if you remove that random number, the responses would always be the same. Generative AI itself is also not intelligent, it is "only" the moderator between your intelligence and the intelligence of the authors of the source data (like texts). The results from Generative AI are almost as trustworthy as the results of polling institutes.
Trust is earned by actions matching words spoken that get the individuals in charge of our healthcare system in the first place. Trust needs to be earned by the programmers of Ai. Trust needs to be earned by all elected officials.
Let’s start there.
Generative AI is best explained with the metaphor of being an "automated polling institute". Instead of directing a query at thousands of individuals by conducting interviews using a call center and then generating the statistical consensus from the interview data as response, Generative AI does it through its statistical algorithm from the several thousands/millions of sources (texts, images, videos, sounds). Everytime a chatbot responds with "I think that ..." replace it with "The statistical consensus in the text data suggest the thought that ...". The variations come from a random number embedded in the query (per session) - if you remove that random number, the responses would always be the same. Generative AI itself is also not intelligent, it is "only" the moderator between your intelligence and the intelligence of the authors of the source data (like texts). The results from Generative AI are almost as trustworthy as the results of polling institutes.
There is no one that you can tquestion.Everyone lies yes no one tells the truth they know how to lie
Classic