ducators know that when humans learn we make connections, we move between the concrete and the general, and we create and use concepts. The current generation of AI machine-learning systems are incapable of thinking in this manner. But what they are capable of is providing a reminder of what should really be at the heart of education.
With generative AI like ChatGPT and Bard, the lack of understanding is profound—which is precisely why we are right to worry about the dangers of the brave new world that is upon us. At best, AI devices process data and arrange information. Though we do seem to be getting better at training AI to navigate complex situations, the discernment and wisdom that allow humans to leverage values and select actions remain far outside the learning and reasoning capabilities of our machines.
The year ahead in education will remind us that schools and teachers adapt and change much faster than it sometimes appears. Just as education has adjusted to huge quantities of information and knowledge being readily available on pocket devices (for example by teaching new information and media literacies), we will see a reckoning with the fact that the A in AI better stands for “augmentation”. In place of breezy whither-the-classroom angst, educators are already well advanced in sussing out the uses and abuses of these steroidal spell-check technologies.
We learn in a world that is always, already ongoing. Our initial efforts to assimilate the craziness and beauty of the moment we have landed in do mean that early-stage learning can resemble the imitative assembling done by large-language model AIs. Think of that college essay riven with passive voice, strings of adjectival nouns, and tremendous effort at “sounding academic.” But at its necessary best, education moves us toward owning our abstractions and understanding our concepts. ChatGPT and Bard never move beyond pastiche.
We may no longer need schools to train us to write three-act musical comedies about old running sneakers featuring limerick duets the way AI can (okay, maybe we never needed this!). But we do need education that helps us think through pathos and aspiration, waste and regeneration. And this is an education that cuts deeper than the performative—that builds powers of understanding, abstraction, particularization, and discernment. Today’s AI can be a great aid to educators, but it remains unable to replace human thinking—or education.
a global affairs media network
Learning Beyond Pastiche
Image via Adobe Stock.
January 14, 2024
When humans learn we make connections, we move between the concrete and the general, we create and use concepts. Current generations of AI operate without these levels of understanding, and that is why we are right to worry about the dangers of AI in education, writes Noah W. Sobe.
E
ducators know that when humans learn we make connections, we move between the concrete and the general, and we create and use concepts. The current generation of AI machine-learning systems are incapable of thinking in this manner. But what they are capable of is providing a reminder of what should really be at the heart of education.
With generative AI like ChatGPT and Bard, the lack of understanding is profound—which is precisely why we are right to worry about the dangers of the brave new world that is upon us. At best, AI devices process data and arrange information. Though we do seem to be getting better at training AI to navigate complex situations, the discernment and wisdom that allow humans to leverage values and select actions remain far outside the learning and reasoning capabilities of our machines.
The year ahead in education will remind us that schools and teachers adapt and change much faster than it sometimes appears. Just as education has adjusted to huge quantities of information and knowledge being readily available on pocket devices (for example by teaching new information and media literacies), we will see a reckoning with the fact that the A in AI better stands for “augmentation”. In place of breezy whither-the-classroom angst, educators are already well advanced in sussing out the uses and abuses of these steroidal spell-check technologies.
We learn in a world that is always, already ongoing. Our initial efforts to assimilate the craziness and beauty of the moment we have landed in do mean that early-stage learning can resemble the imitative assembling done by large-language model AIs. Think of that college essay riven with passive voice, strings of adjectival nouns, and tremendous effort at “sounding academic.” But at its necessary best, education moves us toward owning our abstractions and understanding our concepts. ChatGPT and Bard never move beyond pastiche.
We may no longer need schools to train us to write three-act musical comedies about old running sneakers featuring limerick duets the way AI can (okay, maybe we never needed this!). But we do need education that helps us think through pathos and aspiration, waste and regeneration. And this is an education that cuts deeper than the performative—that builds powers of understanding, abstraction, particularization, and discernment. Today’s AI can be a great aid to educators, but it remains unable to replace human thinking—or education.
