The business Edgenuity offers systems that are AI grading tests. Works out that they simply look for key words without doing any real analysis that is semantic.
It’s not artificial cleverness but natural stupidity placed on computers.
I’m maybe not amazed. Should anybody a bit surpised?
In European countries, Article 22 of this General Data Protection Regulations (GDPR) offers at the very least some transparency that is algorithmic. I will be no attorney but We think the rather demonstrably completely automatic nature associated with so-called AI algorithm quoted here will mean that the european pupil could need to start to see the algorithm to which (s)he had been subject. Such transparency would undoubtedly place a quick end to this kind of rubbish software and ideally the firms peddling it.