In New York City schools, teachers can use artificial intelligence to generate ideas for lesson plans. They are also encouraged to deploy it for research and drafting some documents. But it should never be used to determine disciplinary action against a student or to assign grades to homework or tests.
Those examples of how artificial intelligence should or should not be used by teachers across New York City’s nearly 1,600 public schools are laid out in an initial A.I. playbook released on Tuesday by the school system, which is the country’s largest.
The guidance represents the first major step in establishing ground rules and safeguards for artificial intelligence in New York City classrooms, in what is shaping up to be a larger embrace of the technology, including the possible creation of an A.I.-focused high school.
The city’s guidance comes after some other large districts across the country — including Chicago, Denver and Charlotte-Mecklenburg in North Carolina — have released their own A.I. playbooks during the past 18 months. Some states, such as Ohio, have statewide rules on A.I. in schools.
“We know what is happening in our schools already, which is A.I. is being used by our students and by teachers,” Kamar Samuels, the New York City schools chancellor, said in an interview. “We have to make sure that we have strong guardrails, safety for our young people and data privacy that’s critically important to our families.”
As artificial intelligence has captured the public’s attention and its creators have marketed it as a fundamental reshaping of people’s lives, particularly in the lucrative industry of education, school leaders across the United States have grappled with how much of it belongs in classrooms. Some have embraced it, believing it will be transformational in a positive way. Others argue that it inhibits critical thinking and comprehension, and encourages cheating.
Justin Reich, an associate professor at M.I.T., said that as the technology has become more powerful and ubiquitous, it is necessary for districts to provide guidance for its use by teachers and students. But it should be done with the understanding that artificial intelligence is a nascent technology whose abilities and tools will evolve greatly in the coming years and decades, he said.
“Historically, when we try to guess the best ways of using new technologies, we’re often wrong,” said Mr. Reich, who helped write a guidebook, “A Guide to A.I. in Schools.” “There are lots of people who are out there who will say, ‘This is what we need to do, this is best practice,’ and they’re making stuff up.”
Until now, teachers across the city’s public schools have largely set their own terms for using artificial intelligence. Some have dabbled with it to create lesson plans or draft emails. Others have avoided it. Some teachers, especially those who have caught students using it to cheat, have been frustrated by the lack of clearer rules from school leaders.
The playbook also lands at a time when some New York City parents are urging school leaders to postpone any rollout of A.I. until more rigorous guidelines for teachers and students emerge.
The 30-page guide provides a foundational understanding of artificial intelligence and explains when the technology is not acceptable, when it should be used with caution and when it should be considered a helpful tool. Examples are presented like a stop light: red (prohibited), yellow (proceed with caution) and green (go ahead!).
The guidebook reminds teachers that artificial intelligence should support their work but never replace it and that they should be careful about entering students’ private information. School leaders acknowledge that the technology’s long-term impact on society and students — on how they think and learn — isn’t fully understood.
While the document is intended for teachers, it offers an example for students on an acceptable use. They can consult with A.I. for research and exploring topics, but teachers should check the responses for accuracy, appropriateness and bias.
“Our students do not need technology for its own sake,” the playbook says. “They need accurate instruction, meaningful practice and adults who know them well enough to decide when A.I. belongs in their learning, and when it does not.”
For teachers, there are clear examples of when the technology is prohibited. It should not be consulted for crafting specialized learning plans for students with disabilities.
But teachers can use it to draft sensitive communication — if they tread carefully.
What’s fully acceptable: engaging with artificial intelligence as a guide or sounding board.
Teachers are encouraged to use it to find different ways to teach topics, as well as get assistance with rote tasks such as scheduling and translation. Even then, they are urged to use their own professional judgment because the technology tends to hallucinate — the phenomenon of producing incorrect information, including citing, often with great confidence, sources that do not exist.
During the coming months, school leaders will hold public events to receive feedback and recommendations about the new rules, and incorporate that into an updated guide to be released in June.
Natalie Milman, associate dean of the Graduate School of Education and Human Development at George Washington University, said that it would be imperative for New York City schools to provide training and development for teachers to better understand A.I. In conversations with teachers across the United States, Dr. Milman said that the technology was a polarizing topic; some despise it while others use it every day, with a lot of views in between.
“There are still many teachers who don’t fully understand what A.I. is, or what it’s capable of doing,” she said. “Even if they’re using it, they may not fully comprehend or understand ways that it can be used pedagogically.”
Matthew Haag is a Times education reporter focusing on New York City schools.
The post In N.Y.C. Classes, Teachers Can Use A.I. to Plan but Not to Assign Grades appeared first on New York Times.
