There is better AI out there than ChatGTP, though, particularly the free version.
It's interesting that the school claims they're actually spending more on staff. I doubt that will remain the case; two out of those 3 will likely soon find themselves redundant, and if more than one subject is taught through that course the savings can, potentially, be huge.
Will I worry about my job anytime soon? No. For a start, my subject is a practical one. The screens appear to show at least some Science content, and the beauty of that subject lies in the experiment side of things. You need fully-trained staff for the health and safety aspect, as well as the space and equipment. Yes, we can potentially get to the point where camera tracking and AI interpretation of student movement give the system an accurate measure of how a practical is carried out and suggest corrections, but you might as well just pay the staff member (less maintenance needed, too). So the Sciences, PE, Design and Technology, Food Tech, to some extent the Arts, Drama - they all look like reasonably safe options.
Cost of the actual software aside, the cost of the mere hardware needed would be phenomenal. 20 VR headsets, in constant use (so plenty of spares needed when they, inevitably, break), speakers, microphones, key boards (ideally the ones you cannot easily remove keys from), computers themselves, potentially motion trackers and cameras if the experience was actually interactive. Servers capable of dealing with the demand. I'm talking best-case scenario here where there is future software capable of replicating what an average teacher does to bring a subject to life. Dozens of IT technicians, because yes, you suddenly need many more of those.
Reports of progress wouldn't be an issue' though I'd expect it to be more neutral.
Remember that this is all assuming that the software is capable of all of the following:
- captivating interest daily
- being positive, funny and motivational (most gaming is, so I actually don't see the issue here)
- answering questions and explaining things in a variety of ways that can be tailored to the student
- interpreting the content of answers, even if they do not 100% match the pre-defined answers and even if spelling is appalling (something AI is still pretty poor at, but might get better over time)
- identifying and addressing mistakes in working out, for example in Maths-based subjects
- tracking and correcting student hand movements, for example in diagram construction
- interactive 3D or even 4D modelling
- adapting work for students with hearing aids (visual impairment could be less of an ossue with automatic zooming in options) - this is assuming the software is capable of adapting to most other SEND needs
- being unhackable and staying ahead of students who will find a way to game or watch porn in class
- being aware of changes in student motivation/ progress and raising alerts where necessary (though part of safeguarding will no doubt still be done by the supervisors in class)
- tongue in cheek, getting a student to comply with the dress code
- completing all the necessary checks and paperwork to take students on trips (I'd actually love a system that does all that automatically)
- staying updated with current affairs and development and not be completely outdated after 5 years
I sound rather negative here, but AI still has such a long way to go before it is able to replace teachers, even if the hardware was available and we had the IT staff necessary to make this successful. I would also caution against the over-use of screens in general - my optician regularly points out just how bad constant screen use is for our eyes.
AI will be used more and more in the classroom, of that I have no doubt. Any teacher worth their salt uses the free software out there already to make their lives easier. But fully virtual classrooms, especially in mainstream education, is a pipe dream for now.
/essay over