Another school year is starting. Hunter College's classes started last week and public school K12 teachers went back yesterday. I've been having some "back to school dreams" and, yeah, already missing being with the kids but not missing those opening meetings and the rest of the nonsense.

Still, I'm thinking about what this school year will bring in terms of CS.

The elephant in the room is just the same as what we closed out with last Spring - AI - specifically generative AI.

I think AI is particularly interesting because it's going to be a big topic in CS, in other subjects and also in school and district policy.

Unfortunately, we're already seeing the education community - CS and otherwise play the usual game - anoint the usual suspects experts and give them a platform and eventually those opinions will become the "right ones." Please stop. There are AI experts - people who've created, worked with, analyzed and studied these systems and they can and should be engaged as experts to communicate what we're dealing with on a technical level but nobody's been exploring and experimenting with these tools long enough in school and class settings to be an expert. All we have are people trying some new things now. Results will start to come in and we might eventually have "experts" in AI in education but we don't know so just stop. I wrote about this here where I was at a conference and there were two speakers - both given a platform as "experts." One was way off base and should never have been given the platform. The other just shared "here's what I'm trying and here's what others are trying - give it a go yourself." That second speaker was great.

In non CS classes, I suspect we'll see the usual play out. Some teachers and schools will go draconian with concerns over cheating and copying and others will come up with clever ways to use generative AI to help build classroom exercises and student experiences.

I saw one person talking about redefining how he taught his English classes and that concerns me. He said that with generative AI, writing is less important so he could now focus on reading.

This is a problem. Already there are too many HS English teachers who fancy themselves college literature professors. Writing, as in pen to paper or finger to keyboard might change because we have all these newfangled tools but it's not really about writing. It's about communication. That's the needed skill and that isn't going away. Becoming a literature class won't help there and I see this attitude as problematic if it takes hold.

I'm also concerned about the AI tutor replacing teachers. This one's tricky. On the one hand, automated feedback can be a great addition in terms of student support but it can also be used, as I spoke about when writing about Harvard's CS50 and their using AI tutors. This will probably be good for the students but it's also a way of glossing over serious deficiencies in either course design or delivery. I'm also concerned about students - particularly introverts connecting more with AIs than with people.

On the CS front, we'll see a rush of new AI units. I haven't had a chance to review what's already been put out there but we'll see. It's a tough thing to get right for a CS class. K12 students by and large won't have the background to get their heads around a lot of what's going on so how can you give a fair inkling without overwhelming.

I've seen good and bad with this over the years. A good example in another topic is networking - a complex topic that middle school kids aren't by and large ready for, at least for a deep dive. That said, I've seen teachers do wonderful units where the class becomes a packet switching network breaking up messages, writing down packets, putting them in envelopes and passing them through the room. In the end they can't sit down and code a networking stack but they have a good idea, in the general sense, how it works.

On the bad side, was a set of Google lessons that I saw at SIGCSE a few years ago. All fancy with graphics and sliders but in the end really communicated nothing about how the system (a neural net in this case) worked.

It'll be interesting to see what comes but my money is on more weak stuff than strong.

Within CS and tech we'll see continuations of the usual issues - lack of teachers, digital divide, potential changes in curriculum and the like but in education as a whole we'll see some other big issues to pay attention to.

The trends in red states like Florida and how their policies are downright dangerous to minorities. Book bannings, policies harmful to marginalized communities - particularly the Black and LGBTQ communities and privatization. The CS community has mostly been silent on these with a short lived limited exception when the College Board / DeSantis APAAS thing first came up. On the one hand, the CS community appears to do lot - in our conferences and events inclusion is a big topic but that's coffee cooler talk. To the outside the voices in the community are happy to go along to get along. I'd love to see that change this year - stronger voices by conference organizers and power players. Take a stand - don't give me the CS or education isn't political because these days, everything is.

Anyway, before I get myself too riled up, I think I'll stop. It should be an interesting school year. I'll be on the outside looking in and figuring out how I can remain involved. For those of you in the classroom - good luck for your openings, have fun with your kids and try not to let the administrations bring you down.