Inside San Francisco’s new AI school: is this the future of US education? | Artificial intelligence (AI)

An “AI-powered” private school at the world’s epicenter of technological innovation, headlines for shamelessly embracing technology.

Opening its doors to K-8 students this fall, Alpha School San Francisco is the newest outpost of a network of 14 private schools across the country. The learning model requires just two hours of focused academic study per day, and the school says students can learn twice as fast as their counterparts in traditional schools with the help of artificial intelligence.

Alpha says AI is central to the school’s learning philosophy, brand and impact on students.

Alpha is not alone in its efforts to introduce AI into the classroom. Schools across the U.S. are experimenting with AI as a tool that can help teachers develop curriculum, identify learning trends, and even increase student engagement.

But the strong emphasis on AI at schools like Alpha and its claims of success – and some eye-popping annual tuition fees – is giving education and technology researchers pause, even though they agree that AI will play an important and inevitable role in the future of pedagogy.

“There is evidence that AI can have exciting applications in education, but we have also seen educational experiments in the past that have not worked well for children, including those pushed by technologists,” said Emma Pierson, an assistant professor of computer science at the University of California at Berkeley and a member of the university’s Berkeley Artificial Intelligence Research (BAIR) Laboratory. Pierson, who describes himself as a “cautious optimist” when it comes to artificial intelligence, says there needs to be more critical and objective research on the effects of artificial intelligence on students of all backgrounds.

“Careful evaluation of these programs is really important.”

Alpha School’s website It proposes a futuristic learning model: AI tools enable students to spend just two hours a day on traditional subjects such as history and mathematics. This is done individually on screens using software that adapts to students’ unique pace and learning styles. The remainder of the day is devoted to learning life skills, accomplished through creative activities such as designing, permitting, and operating a food truck that promotes teamwork, social skills, and financial literacy. Rather than receiving instruction from teachers, students are supervised by “guides” who coach them through a self-directed process.

These elements are not completely new pedagogical approaches, experts said. Take the number of hours devoted to traditional schoolwork. Ying Xu, an assistant professor of education at Harvard University, explained that between working on group projects, socializing and taking breaks, students in traditional schools spend the same amount of time, on average, focusing on core curriculum requirements.

Self-directed learning models also have a long lifespan. Xu sent her son to Montessori school because of the adaptable, customizable, independence-encouraging model it offered. “A lot of schools use this type of strategy,” he said.

Some of the software Alpha uses has been widely used in public and private education settings for years to create personalized learning paths, said Chris Agnew, director of Stanford University’s Center for Generative Artificial Intelligence for Education. This includes the following programs: IXL And Mathematics Academy. Historically, schools have also been tasked with: Han Academy and on Duolingo.

Alpha says other apps are new; develops and distributes new software through the network’s affiliate brand. 2 Hours of Learning More than half of the students perform application-based learning through special software.

The school says this specialized software tracks the learning of all students and can deliver lessons to students at the right pace and at the right grade levels, allowing it to truly tailor lessons to each child’s needs, including correcting learning gaps, progressing slower with some materials and faster with others, or tailoring learning materials to children’s interests.

The primary use of AI therefore appears to be as a tool to help Alpha’s guides understand how and how quickly students are learning, so that the material they are given is at the appropriate level, experts said. “AI is included as a layer to help understand the pace of students’ progress and then make suggestions about what might be best to do next,” Agnew explained. “Most learning AI is not student-facing.”

“It’s not exactly a non-stop conversation with a personalized ChatGPT bot,” echoed Victor Lee, an associate professor at the Stanford Graduate School of Education. Alpha confirmed that chatbots play no role in their pedagogy.

In this sense, Alpha’s relationship with technology is not imperceptibly far from the relationship Americans are used to seeing in elementary school classrooms. Screens are used but not constantly; practices are assigned but not without supervision. Perhaps most importantly, the presence of a qualified adult in each classroom encourages a degree of interpersonal learning; Agnew says this should remain a vital part of the school experience. The key difference is that whereas teachers have historically been viewed as subject matter experts giving direction from the podium, at Alpha they are coaches on the field. “There are still adults in the room who know the kids,” he said.

Agnew wonders whether the school’s AI component has been deliberately gamed to turn a profit and attract parents and students who fear their children might miss the AI ​​boat. “They’re leading the way with AI because they’re following the wave, and that’s attracting a lot of attention,” he said.

KWhen it comes to personnel, Alpha is well positioned to benefit from the AI ​​boom. The brand’s founder, education podcaster and social media influencer MacKenzie Price, is married to software executive Andrew Price. He works with Texas billionaire Joe Liemandt, the principal of Alpha Schools. Hedge fund billionaire Bill Ackman He is said to be an unofficial Alpha ambassador. Families who send their students to the San Francisco school also belong to elite circles; highest tuition fee of any private school in the city.

2 Hour Learning website He notes that Alpha students consistently rank in the top 1-2% nationally and claims that 90% of Alpha students love going to school. But experts worry that Alpha’s demographics could cloud assessments of whether Alpha’s methods improve learning outcomes at a nearly universal rate. First, students from unusually affluent backgrounds are likely to have access to other resources that prepare them for academic success. Additionally, parents who choose Alfa demonstrate a high level of interest in their child’s education, another indicator of success.

Alpha rejects criticism that the school’s high achievement record is due to students having high-resource backgrounds and says its model has the potential to increase academic achievement for all students.

Experts warn that there is much to work out and that there is no perfect pedagogy that will work for every child.

Xu’s own research at Harvard shows that learning with AI can lead to very different outcomes depending on the individual’s disposition. Children who are more self-directed and confident in their abilities are more likely to use AI to improve their understanding of material, he said, while children who lack self-confidence are more likely to use AI to generate answers and bypass critical thinking. This means “a student who is inherently less motivated to learn may struggle in this environment,” he said.

“Just like in Montessori schools, the format does not work for all students,” Xu said. “We must not give up on direct instruction,” he said.

Rose Wang, an OpenAI researcher who studies machine learning applications in doctoral educational settings, raised questions about whether an application-based curriculum can serve all students equally, especially given that K-8 students are still developing basic skills. Basic skills like reading, writing and math will be extremely important for children growing up in the age of AI, he says, as a way to ensure they know how to interact critically and effectively with these new technologies.

“This educational model would probably work really well for kids who are already pretty advanced,” Wang said, noting that both younger students and students who need more learning support respond well to face-to-face, collaborative methods, such as solving a math problem on a whiteboard with classmates. “While our AI systems are truly amazing in many ways, these types of collaborative learning and teaching environments [is] It’s really hard to turn it seamlessly into an automated interaction.”

To truly know how widely applicable the method is and for whom it works, experts say more rigorous study is needed. Flagging risks like AI hallucinations and algorithmic bias, Pierson recommends developing a small pilot with a randomized control trial before scaling the model. (Alpha says its proprietary apps have been developed to eliminate the possibility of AI hallucinations.)

Lee calls for examination of issues of privilege and fair access; These are the first issues that come to mind in a city like San Francisco, where the wealth gap is so wide. “Who can do this, where do the resources come from, and what advantages are already behind these programs like Alpha School?” he asked.

Other Alpha campuses offer financial aid, including one in Brownsville, Texas, where the $10,000 annual tuition is heavily subsidized. Alpha says the San Francisco campus can’t afford to provide financial aid yet, given that only 10 students are actively enrolled (and a total of 25 families have put down deposits for their children to enroll between 2025 and 2026).

Experts also warn against over-regulating education in a way that discourages experimentation. This is especially important at a time when technology is advancing rapidly, they said, and young people should be given opportunities to engage with technology in a meaningful way. “We want to have the freedom to try different things, and I think schools and teachers would like that too,” Lee said, pointing out that school districts are doing just that.

Washington’s Peninsula school district is exploring how AI can make learning experiences more inclusive; In Orange County, the superintendent hired artificial intelligence experts to guide schools on how to best incorporate technology into classrooms; Public schools across California are integrating AI literacy into their curricula as part of state-mandated digital media literacy requirements.

“Finding new ways to get students excited about learning, learning better, and learning faster is always a welcome thing,” Wang said.

Agnew says that at the end of the day, improvements are needed in the current system, and artificial intelligence will inevitably be part of that conversation. The challenge for educators and policymakers will be to cut through the noise to ensure that the benefits of AI in the classroom outweigh its risks.

“With so many products being sold… it’s really hard to make decisions based on what’s right for kids and what’s good for learning,” he said. “But it is not an option for schools to ignore AI.”