Summary of those papers: 'This calculator was able to solve the equation faster than a math professor!'.
-
-
-
Yes! Exactly.
- Još 2 druga odgovora
Novi razgovor -
-
-
These conversations also never seem to extend to an analysis of how well the humans were educated, trained, resourced, and supported in the work prior to being put up against the AI.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
There is also a quaint contradiction in claiming that an AI is outperforming humans when the ground truth it is measured against was, in fact, manually determined by other humans.
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
I'm a radiologist and have spent quite some time over the last 4 yrs going to AI meetings. But if I listen to all the presentations and assume they are all going to work perfectly we are still only talking about 20 to 25% of my job. That is only a couple of years worth of demand
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Increase at present rates. I would argue that the value I add is to a very large extent about what I ignore, diminish and reduce. Focus is everything. The combination of widely disseminated AI systems working on radiological data will likely lead to mass confusion and failure to
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Focus on the main issues. In this era where multimorbidity and frailty are major drivers of cost, then adding AI systems to the diagnostic pathway will likely lead to the system being overwhelmed. I teach younger doctors that they need to cultivate wisdom as that is our fundament
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.
Machine learning outperforms judges at predicting recidivism, but only because judges have multiple objectives -- they don't want to incarcerate young people.
How many other machine learning "wins" are actually misspecified objective functions?