Anthropic’s Claude 3 is here, and with it comes much-hyped rhetoric of true artificial intelligence, which in turn leads to the predictable pattern of early adopters seeing ghosts of AGI in every output, followed by legions of skeptics poking holes in these flimsy observations. We should pause and remember the people at Anthropic and OpenAI see current transformer-based systems as mere stepping stones on the road to true Artificial Intelligence, and playing into boosting or bashing creates more engagement for a technology that hasn’t slowed down enough for people to actually adopt it as part of their daily practice.
"the point is each human being should have the opportunity to think, to dream, to learn, to exist outside of the dichotomy between man vs. machine"
I really liked this point. I think the first half of your posts into how powerful Technopoly as defined by Neil Postman truly is in the US. We are unable to really influence values as powerfully as the creators of this technology it seems.
I really like the quote about dignity in writing and believe that part of the challenge in educating young writers is helping them to realize the strength, power, and importance of their own writing.
A reading seems to be that AI is anti-process - enter a prompt, get a paper or story - yet in composing, it is the process, which definitionally features considering ideas and racing down blind alleys and occasionally having breakthroughs and insights, that reveals to ourselves and others the "dignity" described above: Writing becomes more about the messy studio and less about the crisp museum. And yet as a teacher of writing, I know that in the world outside my classroom/bubble, Final Product does matter - how do we include that disjuncture in our teaching and assessment?
The basic idea, it seems, is that AI will be so superior to human writing that it will be the equivalent of "AI art" compated to a baby throwing ink at a wall. As AI bros like to imply, you can always keep up your embarrassing little efforts, but it will be as meaningless as life, art, hope and existence itself.
I have often said that AI is the anti-life equation, essentially the soulless demon that seeks to kill all meaning before killing life itself(perhaps creating a mockery of it via simulation later).
It is a monster. We don't oppose it and use its temptations(like cheap "art") to our own doom
As the abilities of genAI models have slightly plateaued in recent months, it may give us a chance to have more of these kinds of philosophical conversations about the overall role AI will play, not just with regard to writing, but also with respect to other important educational areas like scientific research (see the NY Times this morning - https://www.nytimes.com/2024/03/10/science/ai-learning-biology.html). But to Marc's point, the observation by Scott Aaronson about his daughter's writing is just incredibly sad and anyone who has spent significant time with AI text output knows that it is not even close to exceeding the kind of professional writing celebrated by authors, reviewers, and readers. GenAI writing at the moment, from a teaching standpoint, is useful for a variety of tasks, including idea generation, outlining, and demonstrating structure and clarity, but originality and style are not its strong suits. In some fields where the writing is often turgid, dense, and tedious (law for example), genAI will likely have a significant impact, especially with its ability to scan and summarize large volumes of information, such as case law. But for more creative and open ended writing projects, genAI is not good enough and may never actually be "good enough" to surpass the best human writing. My focus as a teacher at the moment is trying to harness what genAI writing programs can do "right now" and not get caught up speculating about the future. That is more than enough of a challenge for most teachers in the current AI hype climate.
I think there is fascination with offloading the work of teaching (and writing) to technology, whether AI or edtech/Web 1.0 and 2.0 tools, because there is a fear that perhaps technology can give students more of what they need than teachers can. Writing is hard. Teaching is hard! As a teacher, how can I compete with machine learning: a gamified computer program helping students master a particular concept? If I have a classroom of 30 students, all with different learning gaps, needs, and strengths, there is an allure to the idea that I can plug them into the computer and after 15 minutes they will re-emerge with more/better understanding.
What people don’t realize is that while machine learning may be more adaptive, teaching and learning is a human experience requiring human to human interaction. I may be able to learn something from a computer, but I won’t remember it nearly as well as a conversation I had with another person. I won’t be able to make connections to other aspects of my life. Learning is in an innately social act, and AI will always fail in replicating the social humaneness of learning.
"the point is each human being should have the opportunity to think, to dream, to learn, to exist outside of the dichotomy between man vs. machine"
I really liked this point. I think the first half of your posts into how powerful Technopoly as defined by Neil Postman truly is in the US. We are unable to really influence values as powerfully as the creators of this technology it seems.
I really like the quote about dignity in writing and believe that part of the challenge in educating young writers is helping them to realize the strength, power, and importance of their own writing.
Great post!
A reading seems to be that AI is anti-process - enter a prompt, get a paper or story - yet in composing, it is the process, which definitionally features considering ideas and racing down blind alleys and occasionally having breakthroughs and insights, that reveals to ourselves and others the "dignity" described above: Writing becomes more about the messy studio and less about the crisp museum. And yet as a teacher of writing, I know that in the world outside my classroom/bubble, Final Product does matter - how do we include that disjuncture in our teaching and assessment?
The basic idea, it seems, is that AI will be so superior to human writing that it will be the equivalent of "AI art" compated to a baby throwing ink at a wall. As AI bros like to imply, you can always keep up your embarrassing little efforts, but it will be as meaningless as life, art, hope and existence itself.
I have often said that AI is the anti-life equation, essentially the soulless demon that seeks to kill all meaning before killing life itself(perhaps creating a mockery of it via simulation later).
It is a monster. We don't oppose it and use its temptations(like cheap "art") to our own doom
As the abilities of genAI models have slightly plateaued in recent months, it may give us a chance to have more of these kinds of philosophical conversations about the overall role AI will play, not just with regard to writing, but also with respect to other important educational areas like scientific research (see the NY Times this morning - https://www.nytimes.com/2024/03/10/science/ai-learning-biology.html). But to Marc's point, the observation by Scott Aaronson about his daughter's writing is just incredibly sad and anyone who has spent significant time with AI text output knows that it is not even close to exceeding the kind of professional writing celebrated by authors, reviewers, and readers. GenAI writing at the moment, from a teaching standpoint, is useful for a variety of tasks, including idea generation, outlining, and demonstrating structure and clarity, but originality and style are not its strong suits. In some fields where the writing is often turgid, dense, and tedious (law for example), genAI will likely have a significant impact, especially with its ability to scan and summarize large volumes of information, such as case law. But for more creative and open ended writing projects, genAI is not good enough and may never actually be "good enough" to surpass the best human writing. My focus as a teacher at the moment is trying to harness what genAI writing programs can do "right now" and not get caught up speculating about the future. That is more than enough of a challenge for most teachers in the current AI hype climate.
Excellent post!
I think there is fascination with offloading the work of teaching (and writing) to technology, whether AI or edtech/Web 1.0 and 2.0 tools, because there is a fear that perhaps technology can give students more of what they need than teachers can. Writing is hard. Teaching is hard! As a teacher, how can I compete with machine learning: a gamified computer program helping students master a particular concept? If I have a classroom of 30 students, all with different learning gaps, needs, and strengths, there is an allure to the idea that I can plug them into the computer and after 15 minutes they will re-emerge with more/better understanding.
What people don’t realize is that while machine learning may be more adaptive, teaching and learning is a human experience requiring human to human interaction. I may be able to learn something from a computer, but I won’t remember it nearly as well as a conversation I had with another person. I won’t be able to make connections to other aspects of my life. Learning is in an innately social act, and AI will always fail in replicating the social humaneness of learning.