91直播

Do schools need artificial intelligence that can 鈥渄etect and predict learning styles associated with different personality traits?鈥�

What about 鈥渁n intelligent learning companion that would ask students to teach a simulated digital peer while tracking students鈥� mathematical understanding?鈥�

Or a new digital tool based on advances in 鈥減recision medicine鈥� that would gather information about students鈥� working memory, attention spans, and ability to manage their thoughts and feelings?

Those are just three of the 465 ideas that left the Bill & Melinda Gates Foundation and the Chan Zuckerberg Initiative 鈥渢remendously inspired and encouraged,鈥� according to a recently released summary of responses to the groups鈥� joint request for state-of-the-art approaches to education, originally announced last May.

Gates and CZI 鈥渟hare a view that there is enormous unrealized potential for students and that breakthroughs driven by innovation can help students and teachers,鈥� according to the report, titled 鈥�Education Research and Development: Learning From the Field.鈥�

Ongoing investments in such R&D work are expected, although the groups have not yet made any funding decisions, according to a Gates Foundation spokesperson.

The underlying premise of the effort is bold: Founded with the multibillion-dollar fortunes amassed by two of the biggest figures in the technology industry, the Gates Foundation (started by Microsoft founder Bill Gates) and the Chan Zuckerberg Initiative (started by Facebook founder Mark Zuckerberg) have teamed up to create an R&D model for K-12. According to the report, the effort will be modeled on DARPA, an agency of the federal Department of Defense that invests in new technologies related to national security, and Bell Labs, a private research and development company now owned by Nokia.

The goal is to support work that 鈥渃ollapses the boundaries between basic and applied research and is focused on immediate problems of practice,鈥� the report says.

Outside observers pointed to the potentially outsized effect that groups as large and well-resourced as Gates and CZI can have on K-12 policy, for better and for worse.

鈥淏oth funders seem attached to a fairly technocratic model of change in education, [believing] that a new research-derived strategy will make it possible to 鈥榝ix鈥� something they see as wrong with schools,鈥� said Sarah Reckhow, an assistant professor of political science at Michigan State University who tracks philanthropic giving in education. 鈥淭he trouble is that many of these philanthropically funded and technocratic approaches have run aground in recent years, either failing to achieve anticipated gains for students or facing serious political headwinds.鈥�

The three 鈥渋mmediate, high-leverage challenges鈥� that Gates and CZI aim to tackle are:

鈥� Improving students鈥� ability to write the kind of nonfiction that is increasingly required in college and the workplace;

鈥� Deepening students鈥� understanding of key math skills and concepts; and

鈥� Strengthening children鈥檚 鈥渆xecutive functions,鈥� such as the ability to think flexibly, consider multiple ideas, and regulate their own thoughts, emotions, and actions.

About half the ideas submitted in response to the groups鈥� Request for Information came from nonprofits or universities. Another 20 percent came from corporations and ed-tech developers. A total of 12 percent came from educators and schools.

Writing, Math, Executive Functions

Poor writing scores on the 2011 National Assessment of Educational Progress highlight the first challenge, according to the report.

鈥淪uccess in college, work, and life depends upon communicating and expressing ideas effectively through writing,鈥� the document notes.

Among the existing barriers to improving writing, Gates and CZI say: instruction that is too often geared toward the five-paragraph essay and other formats that align primarily with state tests, and a lack of time for teachers to provide students with adequate feedback.

What might help, the groups asked?

Responses highlighted in the report include: programs that connect students with 鈥渁uthentic鈥� audiences, like journalists or people in their neighborhoods, to read and offer feedback on their work; initiatives that support students in 鈥渦sing their voices to advocate for social-justice issues"; online platforms that track the comments and edits students make during peer revisions or 鈥渃apture qualitative data from students鈥� 鈥榤essy drafts,鈥� converting words and sentences into patterns that could provide additional insights for teachers into how their students craft arguments or narratives.鈥�

In math, meanwhile, the two groups highlighted a number of 鈥減romising approaches.鈥� Some sounded quite similar to the kinds of personalized-learning software, platforms, and programs already on the market鈥攕ome of which Gates and CZI already support and a few of which already have a fairly robust body of evidence behind them.

Among the examples in the report: 鈥淚ntelligent鈥� tutoring systems; a personalized-learning platform that can 鈥渁nalyze student diagnostics in minutes, enabling teachers to quickly understand whether a particular explanation or intervention is working;" and artificial intelligence that can glean data and produce recommendations by scanning student responses on classroom assignments. Also highlighted were some non-tech efforts to create fun, real-world learning environments by grounding math instruction in chess or contemporary policy debates around issues like raising the minimum wage.

And the third big area was improving students鈥� executive functions鈥攁 broad set of non-academic mental and emotional abilities, such as working memory and paying attention鈥攁 relatively new area of study. The Gates Foundation was excited to receive 171 submissions on that topic, the group鈥檚 spokesperson indicated.

One big existing problem, the report indicates, is that many students, parents, and educators aren鈥檛 familiar with the concept. In addition, the field has yet to develop valid, reliable measures for gauging skills like managing one鈥檚 emotions. And only a few of the submissions discussed contentious issues of privacy and ethics related to the collection, analysis, and use of such sensitive information.

Still, Gates and CZI highlighted proposals to conduct new research, develop 鈥渢ablet-based measurement and intervention platforms,鈥� and explore 鈥渂ehavioral 鈥榥udges鈥� that target a specific daily behavior and can be taught quickly.鈥�

Many of the suggested approaches highlighted in the new report could also raise concerns about the downsides associated with the expanding role of technology in K-12.

Educational publisher Pearson, for example, recently sparked a controversy when it tested behavioral nudges in one of its instructional software programs that is used exclusively in higher education. Questions abound about the efficacy of AI- and algorithm-driven recommendation engines, some of which have been found to reflect and reinforce racial, gender, and other biases. And many of the solutions touted by Gates and CZI are predicated on further expanding the nature and volume of data collected about students, a trend that is already taxing K-12 officials and lawmakers alike.

R&D for the Long Haul

At the moment, there鈥檚 no timetable for funding decisions or official next steps, the Gates spokesperson said.

The hope, though, is that the R&D effort that ultimately takes shape will be able to support efforts to expand promising early-stage research to new applications and contexts.

And the long-term idea is to bring together interdisciplinary teams of educators, developers, experts in human development and learning measurement, technologists, neuroscientists, and others.

鈥淏reaking down the wall between research and practice, and between basic and applied research, offers the potential to dramatically accelerate solutions to problems that stand in the way of delivering better outcomes for millions of students,鈥� the report says.