The AI Tsunami: Reorienting CSCL Toward a Humanistic, Justice-Oriented, and Planet-Conscious Future

For decades, a specific branch of educational research has studied how people learn best when they work together. This field, known as computer-supported collaborative learning, investigates how technology can help groups of students, colleagues, or community members build knowledge as a team. The core idea is simple but profound: learning is not just a solo activity where a person absorbs facts; it is a social process where understanding grows through conversation, disagreement, and the shared effort of making sense of the world. Researchers in this area have spent twenty years figuring out how to design digital tools that support this messy, human process, moving from early experiments that focused on how groups solve problems to more recent work that considers how culture, identity, and fairness shape those interactions. As artificial intelligence becomes a daily part of life, the question has shifted from whether technology can help us learn to what happens when that technology becomes so smart it can do the thinking for us.

A new paper by researchers from the University of Haifa and Clark University argues that the arrival of generative artificial intelligence is not just another step forward for this field, but a sudden, overwhelming force that threatens to undo decades of progress. The authors describe this shift as a tsunami, a metaphor for a wave so large and fast that it does not simply follow the path of previous changes but has the power to wash over them entirely. They contend that while earlier technologies acted as tools to help humans communicate, modern artificial intelligence acts as a participant that can generate fluent, polished answers on its own. This creates a dangerous illusion: when a group uses these tools, the final product may look perfect, but the actual learning process—the struggle to understand, the friction of differing viewpoints, and the effort to resolve confusion—can disappear. The researchers warn that if the field does not change its direction immediately, it risks losing the very human qualities that make collaborative learning meaningful.

To understand the urgency of this warning, the authors look back at how their field has evolved over the last two decades, describing its history as three distinct waves of development. The first wave focused on the mechanics of how groups think together, treating the team as a single unit of intelligence supported by digital tools. The second wave shifted the focus to the human experience, emphasizing how people become part of a community and how their unique voices and personal histories shape the learning process. The third wave, which was just beginning to take hold before the rise of advanced artificial intelligence, started to confront difficult questions about power, fairness, and who gets left behind in these digital spaces. The authors argue that the field was finally finding its ethical footing, ready to tackle issues of justice and identity, when the rapid arrival of generative artificial intelligence disrupted the entire landscape.

The central problem identified in the paper is that artificial intelligence tends to smooth out the rough edges of human thought. When people learn together, they often start with different, imperfect ideas. They argue, they get confused, and they have to work hard to explain their thinking to one another. This struggle is where real learning happens. However, generative artificial intelligence is designed to produce the most statistically probable, fluent, and coherent text. When learners use these tools, the AI can polish their rough ideas into a perfect-sounding paragraph, effectively erasing the unique voice and the specific confusion that led to the learning moment. The result is a group that appears to agree and move forward quickly, but in reality, the members may still hold completely different, unexamined understandings of the topic. The tool has hidden the differences rather than helping the group resolve them, creating a false sense of shared understanding.

Beyond the classroom, the authors point to three massive, interconnected trends that make this issue even more critical. First, the technology is being developed with the goal of creating artificial general intelligence, systems that could eventually match or surpass human thinking in every area, raising questions about who controls these powerful tools. Second, artificial intelligence is moving out of screens and into the physical world through robots that can work alongside humans in hospitals, factories, and homes, changing the nature of physical collaboration. Third, the infrastructure required to run these systems consumes a staggering amount of energy, with data centers projected to use over 1,000 terawatt-hours of electricity as the technology grows, posing a direct threat to the planet's climate goals. The authors argue that a field dedicated to how humans learn together cannot ignore these forces, as they fundamentally alter the conditions under which learning takes place.

In response to these challenges, the paper proposes a new direction for the next twenty years of research, built on three commitments. The first is a renewed focus on humanism, which means designing technology that protects and values the messy, imperfect, and authentic parts of human experience. Instead of using artificial intelligence to make learning faster or smoother, the goal should be to use it in ways that encourage deep thinking, genuine disagreement, and the kind of struggle that leads to real understanding. The second commitment is a clear focus on justice. The authors argue that because artificial intelligence is trained on data that often reflects existing biases, it can easily reinforce inequalities. Researchers must actively work to ensure that these tools do not silence marginalized voices or make it harder for people from different backgrounds to participate fully. This involves designing systems with the communities they are meant to serve, rather than just for them.

The final commitment is a planetary mandate, which asks the field to expand its view of what it is studying. This means looking beyond the immediate classroom to consider the broader environmental and social costs of the technology. It also means collaborating with other groups, such as environmental scientists, ethicists, and indigenous knowledge keepers, to tackle the complex problems facing humanity, like climate change and social division. The authors believe that the unique strength of their field is its deep, practical understanding of how people actually build knowledge together in real-world settings. By applying this expertise to the challenges posed by artificial intelligence, the field can help ensure that technology serves human flourishing rather than replacing it.

The paper concludes with a call to action for researchers to develop new ways of measuring success. Instead of just looking at how well a group performs a task or how polished their final product is, they need to find methods to see if the artificial intelligence is actually helping people think together or if it is simply doing the work for them. They must study how different groups experience these tools, paying close attention to those who might be harmed by them, and they must engage in the public conversation about how these powerful systems should be governed. The authors suggest that the next twenty years of this field will not be about inventing a new type of software, but about clarifying what collaboration is truly for. It is a time to decide whether technology will be used to deepen human connection and understanding, or whether it will be allowed to flatten the rich, complex reality of how we learn and grow together.