Read on Twitter
Oddly: A-Levels is a good example showing that 'algorithms' have impacted our lives for a long time. A-levels themselves are basically an algorithm to categorise students.
And their outcome has always fed into other algorithms: eg "If >BBB, inc science = accepted on course".
And their outcome has always fed into other algorithms: eg "If >BBB, inc science = accepted on course".
The problem with the A-Level regrading algorithm was not that it was an algorithm, but rather that it was bad.
(& it was actually very crude: think of the number of data points per student. It was overly basic rather than overly sophisticated)
(& it was actually very crude: think of the number of data points per student. It was overly basic rather than overly sophisticated)
& the 'problem' with algorithms impacting lives generally is not that they're new, or that they are 'bad' as a concept, but rather that we've gone from relatively few (A-Levels, etc), to pretty much constant impact on our lives.
And among that constant impact, there are of course many that impact negatively as well as many that impact us very positively.
And it's worth figuring out how to make those with negative impact better, rather than just thinking 'algorithms are bad'.
And it's worth figuring out how to make those with negative impact better, rather than just thinking 'algorithms are bad'.
