That’s the outrageous part of it. They didn’t test it against historical data to find it performed worse. They deployed it at scale, rather than trial it in a small sample. They didn’t notice it leaving large numbers of teachers as “drop-outs” despite only being considered for one position. They didn’t move to correct it, when the problem was noticed. Then to top it all off, they did it again, knowing it does this. Absolutely absurd.