Thanks for articulating this! I would be concerned too about the narrowness and possible bias of an AI career counselor.
However, I think that might be more in the software that uses AI than in the nature of AI itself. I have found that it’s possible to give all kinds of prompts to GPT-3 in the OpenAI playground. For example, I wonder what I would get if I prompted it to give me some surprising ideas for courses that might interest me and explain why they might be interesting.
Hello all, I’m jumping in very late here with some responses–hope that’s okay! I just got so excited to find a group of OER folks interested in AI because I’ve just been discovering AI and thinking about how to connect it to my OER textbook. Here’s a conference proposal I just did on the topic.
I believe, the serendipity also has some pattern/ algo working behind it. Maybe subconscious.
One needs to go on that path to be able to discover, isn’t it?
It can’t be just magic though it does seem to be magical.
My stance of Che-Mystery was to bring out the hardwork, failures, immersion into a topic before the success story. Sceince textbooks miss this angle. Acc to the textbook portrayals, what the students see may be this-
Suddenly, Mendeleev stumbles upon the arrangement, Archemedes has the ‘gotcha’ ‘Eureka,Eureka’, Newton discovers gravity because of the apple falling, Kekule dreams of the structure of benzene…the list is endless. It makes science discoveries more of serendipity than what they actually are- scientific thought.
Those one paragraph, in a box in a corner type of text/image just fails to communicate the process of what happened before Mendeleev got that arrangement, or how Archmedes, Newton were so immersed in the topic to be able to realize what was being manifested.
Just to share with you, here is the video link of the presentation of the course Che-Mystery for OE4BW program and also my podcasts titled Che-Mystery- unheard stories of scientists. (This year, I have also started another series, bedtime stories (folktales)
@wernerio - yes, that is an interesting area- who is accountable of automated decisions? I say this, because I recently came across many algorithms to train children for decision-making and I was and still am a bit apprehensive about the ways where AI is heading to and is becoming so unpredictable in itself, but what if in the long run, a person’s decision taken due to AI was not really the appropriate one and was just based on the data (which could be incomplete)…seems like a rant? 
@annarmills, yeah, ineteresting ideas would and definitely do pop up, but it depends on the prompts we give, I suppose. So, we need advocacy for that as well. 