podcast friday
Jun. 12th, 2026 06:57 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
sabotabby Yep I'm back on my bullshit.
This week's episode is Tech Won't Save Us, "Do AI Chatbots Belong In Schools? ft. Tom Mullaney"
I bet you're going to be real surprised at the answer.
The cool thing about this episode is that it looks at chatbots in the history of ed tech in general. I've often said that the ultimate goal for education is that you'd have 50 students or so warehoused in a classroom, completing modules on screens, disciplined by non-unionized babysitters, while a handful of teachers get paid to write and perform lessons. But that was overly optimistic; those teachers would get paid too much and you can have LLMs write it instead.
It's not that all ed tech is bad. It's just that most of it, historically, has been 1) garbage and 2) in service of privatizing and degrading education.
It shouldn't surprise me that the following approaches to combatting LLMs in schools have failed:
1) The catastrophic, world-destroying environmental cost
2) Intellectual property
3) The cognitive damage it does to children (we have accepted causing brain damage to children in schools, thanks to covid and sports)
Possibly all that remains is the legal liability battlefield. I've had some luck, when chatbots get forced on us, in pushing back by asking if lawyers have reviewed liability if one of the company's products causes the kid to kill themselves or commit a violent crime, given that its architecture is based on software that has caused kids to die by suicide and murder others. No one has seemingly thought about this so it's always a relief hearing tech journalists like Paris Marx and teachers like Tom Mullaney pushing back on the consensus that "personalized tutors" are maybe not a great thing to be inflicting on children.
This week's episode is Tech Won't Save Us, "Do AI Chatbots Belong In Schools? ft. Tom Mullaney"
I bet you're going to be real surprised at the answer.
The cool thing about this episode is that it looks at chatbots in the history of ed tech in general. I've often said that the ultimate goal for education is that you'd have 50 students or so warehoused in a classroom, completing modules on screens, disciplined by non-unionized babysitters, while a handful of teachers get paid to write and perform lessons. But that was overly optimistic; those teachers would get paid too much and you can have LLMs write it instead.
It's not that all ed tech is bad. It's just that most of it, historically, has been 1) garbage and 2) in service of privatizing and degrading education.
It shouldn't surprise me that the following approaches to combatting LLMs in schools have failed:
1) The catastrophic, world-destroying environmental cost
2) Intellectual property
3) The cognitive damage it does to children (we have accepted causing brain damage to children in schools, thanks to covid and sports)
Possibly all that remains is the legal liability battlefield. I've had some luck, when chatbots get forced on us, in pushing back by asking if lawyers have reviewed liability if one of the company's products causes the kid to kill themselves or commit a violent crime, given that its architecture is based on software that has caused kids to die by suicide and murder others. No one has seemingly thought about this so it's always a relief hearing tech journalists like Paris Marx and teachers like Tom Mullaney pushing back on the consensus that "personalized tutors" are maybe not a great thing to be inflicting on children.
