We learn from making mistakes. A mistake recognized and criticised by a good teacher has a bigger effect on the pupil that were a machine or an AI device to make the comment. This is because, unlike the good teacher, the machine has no emotional capacity and responds in a pre-set manner. So good teachers will always be needed or edumacation will never be well done.
Very thoughtful post. Thank you, Marc. I was unaware of the Alpha school experiment and am curious to learn more. I think your empasis on the friction, or struggle, behind learning is very important.
So, I don't think the spectre of AI replacing teachers is an either-or proposition. Teachers will be needed. Which teachers and how they will be needed are up for debate.
Oganizations polled by the World Economic Forum – including the education sector – said they were planning to incorporate more AI technologies. They estimate that 34% of all jobs-related tasks in their organizations are already being performed by machines and they expect that number to reach 42% by 2027.
I don't think it's unfathomable that 42% of education related tasks will be performed by AI by 2027. But that doesn't mean that teachers are out of work. As you've alluded to, there is keen interest in the automation of non-classroom teaching tasks, such as grading and feedback, lesson planning, school communication, and more. (And school administrators are in need of enhanced automation systems.) An Education Week study in 2022 found that a typical U.S. teacher works a median of 54 hours per week, but just 46 percent of their time in the school building is spent teaching. So, it's possible that AI will likely have its greatest impact outside of the classroom, not inside, by automating time-consuming tasks.
That said, I don't think all teachers are safe. LLMs have achieved “stunning results" on standardized educational assessments and its abilities continue to improve. The Organisation for Economic Co-operation and Development (OECD) predicts that AI will soon completely master its test of adult literacy, numeracy, and problem-solving skills. AI has also made great strides in perceiving and responding to human emotional states and even reading our minds. (according to a New York Times report). And, as you illustrate, students are increasingly embracing AI tutors.
Teachers point out that human-to-human interaction is at the core of learning. In an Education Week survey greater than 90 percent disagreed that learning would improve where “chronically low-performing” human teachers were replaced by AI robots. It underscores a commonly held belief amongst educators that robots will “never” replace teachers since human interaction is vital to the educational experience.
This last sentiment, while noble, strikes me as wishful thinking. For one, a 2019 PEW Research Center report indicated that Americans viewed teachers as the sixth most likely profession to be at risk of automation. In addition, a 2021 Frontiers in Education study found that students were “generally excited and positive” about the idea of being taught by robots.
If I'm a primary school teacher I'm generally unconcered about AI because the social and emotional learning that comes from interactions between human beings is at the core of student development in my classroom. If I'm a university lecturer or an upper-level secondary school teacher I'm more concerned, because AI knows my academic content better than I do and can deliver it in varied, multimodal ways to reach diverse learners. And if I'm a teacher-focused, stand-up-and-deliver type, than I'm even more concerned.
In the end, the most likely scenario is that teachers will be forging collaborative partnerships with AI. AI will support educators by providing personalized learning content and recommendations, analyzing student performance data, and offering insights into optimal teaching strategies.
Teachers will use this information to tailor their lessons and better address individual student needs, while still maintaining the human connection and empathy critical to fostering a supportive learning environment.
But, as you've depicted in Austin, there will be variations in AI adoption along the way.
I really like the phrase “Friction matters in learning”. Learning faster should not be associated with learning better, it should not be a North Star when developing educational tools. Learning takes time, no amount of technology will make our brain assimilate complex concepts 2x, 5x, 10x faster (expect the hypothetical implant, Matrix style). It requires repetition, productive struggle, social interactions, etc. Certain solutions may provide more optimal ways to learn, but the promise to keep learning faster and faster, instead of better and better, is misleading at best.
After reading these essays, I've decided that students should be reading them, so I'm thinking about a way to make them into an assignment on Perusall. Last fall I had students watch Ethan and Lilach Mollick's videos on LLMs and AI but of course they're outdated (somewhat) now. Any of the TTs mentioned here students who are savvy to this have seen - but the philosophical and ethical issues you raise here are things we should be talking about with our students. These are so thought provoking and useful.
So far I like what I've heard from Khan; he has a lot of experience with ed tech and a good reputation, and seems clear-eyed about the need for guard rails around AI. I have far less confidence in the for-profit AI education newcomers, local school boards and State legislatures. I'm curious as to how textbook publishers will react - my guess is with lawsuits. AI content could replace textbooks, allowing school boards to assign bespoke versions that align with their viewpoints. With AI tutors to enforce teaching consistency, all that will be left for 'teachers' is classroom management.
I think you may have put the blame for the problem in the wrong place. In an education system that relies on knowledge transactions translated into GPA it should not surprise anyone that the bot would fit right in. Teachers are very powerful figures who are mandated to fill out report cards and reduce learning to accumulation of evidence. People of all ages who are treated as pawns will behave as pawns and seek the path of least resistance
Education should be liberatory. This looks like entrapment.
Its just part of their effort to replace humans for everything
We learn from making mistakes. A mistake recognized and criticised by a good teacher has a bigger effect on the pupil that were a machine or an AI device to make the comment. This is because, unlike the good teacher, the machine has no emotional capacity and responds in a pre-set manner. So good teachers will always be needed or edumacation will never be well done.
It is all about saving money. That is the bottom line. And the teacher shortage. In K-12, the shortage is going to get serious in the next 5-10 years.
I really like this series.
Very thoughtful post. Thank you, Marc. I was unaware of the Alpha school experiment and am curious to learn more. I think your empasis on the friction, or struggle, behind learning is very important.
So, I don't think the spectre of AI replacing teachers is an either-or proposition. Teachers will be needed. Which teachers and how they will be needed are up for debate.
Oganizations polled by the World Economic Forum – including the education sector – said they were planning to incorporate more AI technologies. They estimate that 34% of all jobs-related tasks in their organizations are already being performed by machines and they expect that number to reach 42% by 2027.
I don't think it's unfathomable that 42% of education related tasks will be performed by AI by 2027. But that doesn't mean that teachers are out of work. As you've alluded to, there is keen interest in the automation of non-classroom teaching tasks, such as grading and feedback, lesson planning, school communication, and more. (And school administrators are in need of enhanced automation systems.) An Education Week study in 2022 found that a typical U.S. teacher works a median of 54 hours per week, but just 46 percent of their time in the school building is spent teaching. So, it's possible that AI will likely have its greatest impact outside of the classroom, not inside, by automating time-consuming tasks.
That said, I don't think all teachers are safe. LLMs have achieved “stunning results" on standardized educational assessments and its abilities continue to improve. The Organisation for Economic Co-operation and Development (OECD) predicts that AI will soon completely master its test of adult literacy, numeracy, and problem-solving skills. AI has also made great strides in perceiving and responding to human emotional states and even reading our minds. (according to a New York Times report). And, as you illustrate, students are increasingly embracing AI tutors.
Teachers point out that human-to-human interaction is at the core of learning. In an Education Week survey greater than 90 percent disagreed that learning would improve where “chronically low-performing” human teachers were replaced by AI robots. It underscores a commonly held belief amongst educators that robots will “never” replace teachers since human interaction is vital to the educational experience.
This last sentiment, while noble, strikes me as wishful thinking. For one, a 2019 PEW Research Center report indicated that Americans viewed teachers as the sixth most likely profession to be at risk of automation. In addition, a 2021 Frontiers in Education study found that students were “generally excited and positive” about the idea of being taught by robots.
If I'm a primary school teacher I'm generally unconcered about AI because the social and emotional learning that comes from interactions between human beings is at the core of student development in my classroom. If I'm a university lecturer or an upper-level secondary school teacher I'm more concerned, because AI knows my academic content better than I do and can deliver it in varied, multimodal ways to reach diverse learners. And if I'm a teacher-focused, stand-up-and-deliver type, than I'm even more concerned.
In the end, the most likely scenario is that teachers will be forging collaborative partnerships with AI. AI will support educators by providing personalized learning content and recommendations, analyzing student performance data, and offering insights into optimal teaching strategies.
Teachers will use this information to tailor their lessons and better address individual student needs, while still maintaining the human connection and empathy critical to fostering a supportive learning environment.
But, as you've depicted in Austin, there will be variations in AI adoption along the way.
Is anyone else as afraid of the future of our society as I am?
I really like the phrase “Friction matters in learning”. Learning faster should not be associated with learning better, it should not be a North Star when developing educational tools. Learning takes time, no amount of technology will make our brain assimilate complex concepts 2x, 5x, 10x faster (expect the hypothetical implant, Matrix style). It requires repetition, productive struggle, social interactions, etc. Certain solutions may provide more optimal ways to learn, but the promise to keep learning faster and faster, instead of better and better, is misleading at best.
After reading these essays, I've decided that students should be reading them, so I'm thinking about a way to make them into an assignment on Perusall. Last fall I had students watch Ethan and Lilach Mollick's videos on LLMs and AI but of course they're outdated (somewhat) now. Any of the TTs mentioned here students who are savvy to this have seen - but the philosophical and ethical issues you raise here are things we should be talking about with our students. These are so thought provoking and useful.
Data science is going to ruin everything.
So far I like what I've heard from Khan; he has a lot of experience with ed tech and a good reputation, and seems clear-eyed about the need for guard rails around AI. I have far less confidence in the for-profit AI education newcomers, local school boards and State legislatures. I'm curious as to how textbook publishers will react - my guess is with lawsuits. AI content could replace textbooks, allowing school boards to assign bespoke versions that align with their viewpoints. With AI tutors to enforce teaching consistency, all that will be left for 'teachers' is classroom management.
I think you may have put the blame for the problem in the wrong place. In an education system that relies on knowledge transactions translated into GPA it should not surprise anyone that the bot would fit right in. Teachers are very powerful figures who are mandated to fill out report cards and reduce learning to accumulation of evidence. People of all ages who are treated as pawns will behave as pawns and seek the path of least resistance