The article investigates the impact of a futures-oriented teaching module that incorporates generative AI (GenAI) to enhance university students' foresight and action on climate change. The study focuses on how this innovative approach influences students' perceptions of climate futures, their hope, and self-perceived action competence. By integrating GenAI tools, the module aims to foster a more optimistic and solution-oriented mindset, encouraging students to envision and work towards sustainable futures. The research employs a comparative design with pre- and post-tests, utilizing both quantitative and qualitative data to assess the effectiveness of the teaching module. Key findings reveal a significant shift towards positive and sustainability-focused perspectives among students, highlighting the potential of AI-enhanced education in addressing complex global challenges. The article also explores the development of AI literacy among students, emphasizing the importance of critical engagement with AI tools in educational settings. Through detailed analysis and student reflections, the study provides valuable insights into the transformative potential of futures-oriented teaching with GenAI integration, offering a compelling case for its application in climate change education.
AI Generated
This summary of the content was generated with the help of AI.
Abstract
The study examines the development and impact of a futures-oriented teaching module designed to cultivate foresight and engagement with climate change action. The module follows a scaffolding framework that guides undergraduate students through exploring possible future scenarios, identifying their underlying factors and drivers, and envisioning the desirable, sustainable futures connected to their present lives. Building on an earlier version, the module integrates generative Artificial Intelligence (GenAI) to help students visualize and articulate these future scenarios. The study involved two groups: one using Image GenAI and the other using Text GenAI. Through a pre and posttest design (N = 58 and 62, respectively), we assessed changes in students’ perceptions of the future, hope, and self-perceived action competence regarding climate change. Data collection included a word association test and two Likert scales, supplemented by post-instructional reflections on futures-oriented learning and GenAI integration. Results showed that after instruction, both groups associated the future with more positive, sustainability-related words, and their sense of hope significantly increased. Both groups felt more aware of their role and the implications of climate change after exploring future possibilities. Significant improvement in self-perceived competence was observed in the Image Group. While students responded positively to using GenAI, they also recognized its limitations, an important outcome that contributed to their AI literacy. Understanding the limitations of GenAI is a key component of developing critical thinking around AI tools. These findings highlight the potential of integrating GenAI-supported futures thinking into education to empower students with the mindset and competencies needed to address complex global challenges.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
Foresight, or the ability to anticipate and prepare for future scenarios, is crucial in climate change education. There are two key assumptions about the concept of foresight: one is that the future comprises a set of plausible outcomes rather than a single, predetermined path, and the other is that people do have some influence over which future actually occurs (Bishop & Hines, 2012). Educational scholars emphasize that thinking about alternative scenarios broadens perceptions of the future and prepares students for uncertainty (Bishop & Strong, 2010). By fostering a forward-thinking mindset, students are empowered to make informed decisions and take meaningful actions (Pahl & Bauer, 2013; Pahl et al., 2014).
This paper addresses a study on a GenAI-enhanced, futures-focused teaching module (8 h) developed for university general education in Taiwan, aiming to enhance students’ foresight or futures thinking on climate change. The module, intended for all majors and suitable for integration into general science and environmental courses, includes a series of lectures, films, and in-class activities to introduce and apply the concept of scenario development and analysis, both in general and specific to climate change. Students are led to systematically explore future pathways based on scientific evidence, identify factors and drivers underlying these pathways and eventually envision a future scenario with a focus on the desirable and sustainable pathway.
Advertisement
The instructional design of this study is grounded in futures thinking pedagogy, which emphasizes structured explorations of possible, probable, and desirable futures (Hicks, 2012; Hodson, 2011; Paige & Lloyd, 2016). Futures thinking encourages learners to critically reflect on long-term consequences, recognize the role of human agency, and envision sustainable pathways (Wiek et al., 2011). This approach aligns with climate change education, where uncertainty about the future often leads to disengagement and fatalism (Ojala, 2012a). Hope plays a critical role in this process, as it enables students to shift from a passive stance to active participation in climate action. While Snyder’s hope theory (Snyder et al., 2002) conceptualizes hope as goal-directed thinking that involves identifying pathways and maintaining motivation, our study builds onLi & Monroe’s 2018 framework, which emphasizes both individual and collective efficacy in addressing climate challenges.
The futures-focused teaching module was designed following a structured scaffolding framework (Julien et al., 2018; Lloyd et al., 2010) to guide students from conceptual exploration of future pathways to concrete envisioning of a carbon–neutral future. This process was structured into two key stages (as shown in Fig. 1): (1) Exploring futures (steps 1–5): Students analyzed drivers and trends influencing climate futures, examined alternative pathways based on scientific evidence, and critically reflected on past and present influences. (2) Narrating futures (steps 6–7): Students envisioned and articulated their own desirable future scenarios, using visualization to bring these futures to life.
Several researchers have found that using student-generated narratives to teach socio-scientific issues is helpful in promoting positive emotions and moral sensibilities (Otto & Pensini, 2017; Taiwan, 2020). Narrating futures scenarios can strengthen the presentation and interpretation of important scientific findings by bringing futures alive and making the impact of climate change more compelling and relevant (Pahl & Bauer, 2013). Especially, through developing narratives of futures scenarios by themselves, learners have the opportunity to get engaged in envisioning the future, connecting the future possibilities to their current lifestyles and community choices, and contemplating the meaning of decision-making and action in light of their future envisioning.
By scaffolding students’ engagement with future scenarios, the intervention aimed to develop both cognitive and emotional-behavioral competencies, particularly in three key future-related areas: (1) perceptions of climate futures—helping students move from passive acceptance of dystopian narratives to more solution-focused perspectives (Jone set al., 2012; Levitas, 2017); (2) hope regarding climate change—fostering students’ sense of hopefulness about climate change with a belief in both their own capacity and society’s collective capacity to address climate challenges (Li & Monroe, 2018, 2019); and (3) self-perceived action competence—increasing students’ knowledge, confidence, and willingness to effectively engage in climate solutions (Breiting & Mogensen, 1999; Olsson et al., 2020).
Advertisement
An earlier version of this teaching module, without the integration of GenAI, showed a positive effect on students’ perspectives on climate change (Liu, 2022). However, envisioning futures proved to be a difficult exercise for many students. To address this challenge, GenAI was integrated into the teaching module in this study to enhance students’ envisioning of future scenarios. Implemented over two semesters, one class used text-to-image AI (Microsoft Bing), while another used text-to-text AI (ChatGPT). This AI integration aimed to assist students in creating plausible and creative future scenarios.
The research aimed to understand the impacts of this GenAI-enhanced futures-focused learning model on students’ perceptions and attitudes. Specifically, this study was guided by the following research questions:
1.
How does the GenAI-enhanced futures-focused learning model influence undergraduate students’ perceptions of climate futures?
2.
What are the effects of GenAI-enhanced futures-focused learning on students’ hope and self-perceived action competence regarding climate change?
3.
How do students perceive their learning experiences and gains in the GenAI-enhanced futures-focused teaching module?
Integrating Futures in Climate Change Education
One of the most powerful arguments for the inclusion of an explicit future perspective in climate change education is the recognition of the impact of young people’s own thinking and images of the future (Smith, 2010). Previous studies indicate that young people generally feel a loss of faith in humanity’s progress toward a positive future. They often express worry, fear, sadness, anger, and a sense of powerlessness about climate change and its impact on their future lives (Ojala, 2012b, 2015; Pfautsch & Gray, 2017). For example, in a recent large-scale survey of 15- to 19-year-old Australians, one in four reported feeling very or extremely concerned about climate change, with 29% expressing high psychological distress (Teo et al., 2024). Another international survey of 16- to 25-year-old young people also revealed a high number of negative thoughts about climate change, for example, 75% said that they think the future is frightening, and 83% said that they think people have failed to take care of the planet (Hickman et al., 2021). All these negative emotions, in turn, diminish their belief in the possibility of a sustainable and positive future (Ojala, 2012b).
The power to shape or influence the future is closely related to the quality of our visions, especially the images of a preferred state (Bishop & Hines, 2012). Without clear and inspiring visions of a sustainable future, it is challenging for individuals and societies to move toward positive changes. The lack of well-considered images of sustainable futures creates serious barriers to the envisioning of a different way forward (Smith, 2010). Therefore, enhancing the quality and depth of these visions is essential for fostering a proactive and hopeful mindset toward the future.
As climate impacts become increasingly severe and climate change affects all important aspects of our future, education should address the critical question: “Where do we want to go?” This involves highlighting knowledge about alternatives and visions (Jensen, 2002, p. 331). Levitas, 2013, 2017) emphasizes the role of utopian thinking in addressing contemporary social and ecological challenges. Utopia is commonly used to describe dreams for a better life that have always been a part of human imagination. Levitas argues that utopia is not just an unrealistic ideal but a crucial method for imagining alternative societies and futures.
Utopian thinking can be considered a way of organizing an integrated “forward dreaming” (Bloch et al., 1986). This method “allows us not only to imagine what an alternative society could look like, but enables us to imagine what it might feel like to inhabit it, thus giving a greater potential depth to our judgements about the good” (Levitas, 2017, p. 3). It is both affective and cognitive, moving between the present and the future. Engaging in utopian thinking enables a kind of double vision in which we can look not only from present to future but also from (potential) future to the present, fostering a shift in desires and systemic thinking (Nørgård & Holflod, 2024).
Generative Artificial Intelligence (GenAI) as Pedagogical Tools
The use of GenAI in education has gained increasing attention with the release of ChatGPT in November 2022. GenAI is based on algorithms that learn from patterns of existing data (training data) and generate novel data in the form of text, images videos, audios, and even 3D models (Ali et al., 2024). The algorithms are able to engage in coherent human-like conversations and have displayed remarkable proficiency (Grassini, 2023). Their impact on the education sector has elicited mixed reactions, with some educators advocating them as a progressive step and others raising alarms over their potential to reduce analytical skills and promote misconduct (Cooper, 2023; Langran et al., 2024; Nikolopoulou, 2024; Wu, 2023). A general consensus is that integration of GenAI in student learning and assessment is only helpful if educators play an active and effective role in checking the validity, reliability, and accuracy of the generated material (Nikolopoulou, 2024; Zirar, 2023).
University students have shown a generally positive attitude toward GenAI in teaching and learning (Chan & Hu, 2023; Farhi et al., 2023). As ChatGPT as well as other later GenAI tools are made accessible to everyone with internet access, students’ self-initiated adoption has made it almost impossible to ban or control it (Dai et al., 2023). While ChatGPT has been increasingly integrated into a wide range of fields and industries, the job market is constantly evolving, and many future careers will involve working with emerging AI tools. It is necessary to help students develop their AI literacy, which includes understanding the basics of AI, how it works, its advantages, and disadvantages, as well as different uses in higher education in order to thrive in their future workplace (Kong et al., 2023; Ng et al., 2021; Tzirides et al., 2024).
The educational potential of GenAI lies in its ability to enhance learning analytics, generate customized scaffoldings, facilitate idea formation, and eventually expand educational access and resources for social justice (Dai et al., 2023). GenAI can help educators generate teaching and assessment material (Mikroyannidis et al., 2024) and provide students with engaging and compelling learning experiences and academic outcomes (Ali et al., 2024; Grassini, 2023). One most noted potential is that GenAI can be used as an ideation facilitator to spark students’ curiosity and creativity by generating diverse and unpredictable ideas and solutions (Ali et al., 2024; Dai et al., 2023; Langran et al., 2024). Its extensive database enables it to offer different perspectives, which allows students to think outside the box. However, while GenAI’s output is derived from the data it has been trained on, it is not a think tank of novel and original ideas itself; students will need to deploy their expertise and judgement to evaluate the outputs and construct novel and original ideas themselves (Daietal., 2023).
Beyond providing content, GenAI can serve as an interactive tool that supports students in articulating and refining their ideas. In this study, students engaged with GenAI iteratively, adjusting their prompts and reflecting on the generated outputs to develop well-considered future scenarios. This process fostered active learning and deeper cognitive engagement, rather than passive reception of AI-generated content. From a sociocultural perspective, this iterative engagement aligns with scaffolding strategies (Amerian & Mehri, 2014; Van DerStuyf, 2002), where AI functions as a cognitive mediator that helps students externalize and refine their thinking. Furthermore, peer discussions and group interactions played a key role in shaping these future scenarios, reinforcing the role of social mediation in the learning process (Vygotsky & Cole, 1978).
This study explores the intersection of futures-oriented teaching and GenAI in climate change education. We facilitated students’ future envisioning through GenAI tools based on text-to-text (ChatGPT) and text-to-image (Microsoft Bing) algorithms. GenAI is used to create textual or visual future scenarios based on students’ own prompts, helping them better understand and engage with complex concepts related to climate change and sustainability. By combining foresight with AI-enhanced instructional strategies, this research investigates the potential impact on student engagement and learning outcomes, particularly in terms of hope and action competence.
Methodology
This study employed a comparative design with pre and posttests to examine the impact of two variations of a futures-oriented teaching module with AI integration. This teaching module was conducted as the final component of the course, after students have received substantial instruction on climate change, including basic scientific concepts, causes, and consequences in ecological and social dimensions. Two groups of students received identical instructional content and structure, with the only difference being the use of either Image GenAI (Image Group) or Text GenAI (Text Group) to facilitate the envisioning of desirable, sustainable future scenarios (Fig. 1). The pretest was administered immediately before the module began, and the posttest was conducted upon its completion. The study aimed to assess the effects of the futures-oriented teaching module with these different AI integrations on students’ perceptions of the future, hope, and self-perceived action competence regarding climate change. A mixed-methods approach was used, combining quantitative and qualitative data, including word association tests, Likert-scale surveys, and post-instructional reflections.
Participants
The participants were undergraduate students enrolled in an introductory climate change course intended for all majors. They represented a wide range of academic disciplines and study years. The teaching module, delivered as the final component of this course, was introduced after students had gained fundamental knowledge about climate change. This prior knowledge provided a foundation for engaging in futures thinking activities. The study was conducted over two consecutive semesters, with one semester utilizing Image GenAI and the other Text GenAI. The final matched data included 58 students in the Image GenAI group and 62 in the Text GenAI group. Participation was voluntary, and informed consent was obtained from all students prior to the teaching intervention.
Teaching Intervention
The futures-oriented teaching module (ca. 8 h) was designed to guide students through a scaffolded exploration of possible future scenarios related to climate change, the key drivers and underlying factors shaping these futures, and their personal visions for a sustainable and desirable future. The goal is to foster students’ innovative application of their climate change knowledge while encouraging a forward-looking mindset in addressing the complex challenges of climate change. The module was developed using an instructional framework adapted from previous studies on futures thinking and narrative writing (Julien et al., 2018; Lloyd et al., 2010), and it is divided into two major themes: “exploring futures” and “narrating futures” (Fig. 1).
The first theme, “exploring futures,” guided students systematically through a process of examining possible future scenarios related to climate change, based on scientific evidence, and identifying key differences between these scenarios. Students were introduced to an online platform (https://tccip.ncdr.nat.gov.tw/ds_02_01_ar5.aspx) that communicates Taiwan’s climate change projections, derived from rigorous scientific data. The platform allowed students to manipulate parameters and explore future changes across four emission pathways: Representative Concentration Pathway (RCP) 2.6, 4.5, 6.0, and 8.5, as defined by the IPCC’s fifth assessment report.
Through interactive engagement with these scenarios, students discussed the implications and significance of future climate changes, as well as the scientific and societal factors shaping different pathways. Toward the end of this theme, the focus shifted to identifying the most desirable, sustainable pathway—RCP 2.6, which represents a carbon–neutral future. To ground this in real-world contexts, students identified diverse drivers of climate change, covering energy choices, transportation and consumption behaviors, urban development, policy interventions, and technological innovations. In one activity, students reflected on and discussed the figures, events, and factors that have influenced or are currently influencing climate change in both positive and negative directions. By examining these contrasting influences, students were encouraged to think critically about what drives a sustainable future and what steps are necessary to achieve a carbon–neutral society.
The theme of “narrating futures” aimed to guide students in identifying concrete elements of positive change within a key aspect of life and envisioning their own visions of a carbon–neutral future. In this part of the module, GenAI was integrated to help students create desirable, sustainable future scenarios based on their knowledge and imagination. The accuracy and relevance of AI-generated outputs were grounded in scaffolded instruction. Students were guided to iteratively refine their prompts and engage critically with the generated content. Rather than passively accepting AI outputs, students actively adjusted their inputs to arrive at future scenarios that best reflected their ideas. By interacting with AI, sharing and discussing their results, they were expected to develop a more genuine understanding of GenAI’s strengths and limitations.
The activity began with a brief introduction to the GenAI tools students would use. For the Image Group, this tool was Bing, and for the Text Group, it was ChatGPT. Students were instructed that they would use GenAI to help narrate what a sustainable future might look like. More specifically, students were asked to imagine themselves in the year 2053, in their 50 s, living in a city that had achieved carbon neutrality. They were tasked with envisioning what they would see and experience in such a future.
To generate effective prompts for the GenAI tools, students first needed concrete elements as input, which were grounded in prior instruction. Before this activity, they had already engaged in structured discussions on key drivers of carbon neutrality, including energy choices, transportation, consumption behaviors, and policy interventions. Working in groups, they explored how these previously identified could be translated into future scenarios. These discussions helped them shape vivid and detailed visions of a carbon–neutral city. Each group then presented their ideas to the class, encouraging a diverse range of perspectives on what a carbon–neutral city could look like. Following the presentations, students interacted with GenAI by using prompts that began with “Create a future scenario that includes…,” incorporating the features they had agreed upon or liked during their discussions. Students were encouraged to revise and refine their prompts until the GenAI output felt most fitting to their envisioned future. All students shared their final outputs on Jamboard (for the Image Group) and on Line (for the Text Group), allowing the entire class to view each other’s work and enabling the instructor to summarize and conclude the activity. Through this process, students gained a clearer understanding of how specific prompt elements shape AI-generated future scenarios. While many interesting and meaningful visions were created, certain recurring features, such as renewable energy and low-carbon transportation, were frequently emphasized, reflecting the shared foundation of prior discussions. The opportunity to compare multiple AI-generated outputs allowed students to reflect on different real-world interpretations of carbon–neutral futures and gain deeper insights into the role of GenAI in this particular task.
Data Collection
A survey including a word association test and two Likert scales was administered twice, at the beginning and the end of the semester.
The word association (WA) test is an open-ended question that gauges students’ perceptions or feelings connected to the future of climate change by focusing on the spontaneous generation of words. This question was intended to gain a unique insight into students’ perceptions of climate change in terms of hope. The guiding prompt was “What is your feeling when thinking of the future? Please write down three words that come to your mind.”
The Climate Change Hope Scale (CCHS) (Li & Monroe, 2018) was used to measure students’ beliefs in their personal capability, society’s collective capability, and the potential lack of both in addressing the challenges posed by climate change. This scale was chosen because it aligns directly with our stud’s focus on enhancing students’ capacity to engage in climate action. While another well-established climate change hope scale by Ojala 2012a, 2015) is valuable for measuring different emotional responses, it primarily focused on sources of hope, distinguishing constructive hope from denial-based hope. This CCHS scale comprised 11 items, grounded in a three-factor structure: personal-sphere willpower and waypower (PW), collective-sphere willpower and waypower (CW), and a lack of willpower and waypower (LW). PW assesses the degree to which an individual possesses agency and pathway thinking in addressing climate change-related problems because they can think of ways and are willing to take action to help solve problems (e.g., “I know that there are a number of things that I can do to contribute to global warming solutions”). CW reflects the degree to which an individual perceives society as a whole as having agency and pathway thinking in addressing climate change-related problems (e.g., “If everyone works together, we can solve problems caused by climate change”). LW measures the degree to which an individual feels hopeless in addressing these problems because they believe that climate change is too complex and beyond our control (e.g., “Global warming is such a huge problem and I don’t think people can change it”). Participants were asked to rate each statement on a 7-point scale (1 = strongly disagree; 7 = strongly agree). Higher scores, with LW scores reversed, represent higher levels of hope regarding climate change.
Another Likert scale, Self-perceived Action Competence for Climate Change Scale (SACCCS), was employed to measure students’ self-perceived capability to act on climate change. The scale was adapted from an established and previously validated scale on self-perceived action competence for sustainability (Finnegan, 2022; Olsson et al., 2020). It was chosen because it is a well-established and validated scale that aligns with the instructional objectives highlighting action and engagement. It contained 12 items categorized into three components: knowledge of action possibilities (KAP) (e.g., “I know what one can do at home in order to combat climate change”), confidence in one’s own influence (COI) (e.g., “I believe what each person does makes a difference to climate change.”), and willingness to act (WTA) (e.g., “I want to encourage my family and friends to take climate action”). Each component contained four statements. Participants were asked to rate their agreement with each statement on a 5-point scale (1 = strongly disagree; 5 = strongly agree). Higher scores indicate higher levels of self-perceived action competence for climate change.
For the post-survey, two open-ended questions were added to gauge students’ perceptions of their learning experiences and gains. The first question asked students how exploring carbon–neutral scenarios influenced their thinking about climate change. The second question prompted students to describe the pros and cons of using the GenAI tool to visualize future scenarios in the final activity.
Data Analysis
The word association test elicited a collection of words from the students related to the future of climate change. Two experienced research assistants categorized all word responses into three categories: “positive,” “negative,” and “neutral.” This categorization follows prior research using WA tests to analyze emotional and cognitive responses to science or environmental concepts (Finnegan, 2023; Vlasák-Drücker et al., 2022). In this study, words reflecting optimism, sustainability, or solutions (e.g., “green-energy,” “hopeful”) were categorized as positive. Words indicating fear, crisis, or loss (e.g., “fearful,” “disaster”) were considered negative. Those without a strong emotional connotation (e.g., “technology,” “change”) were placed in the neutral or uncertain category. Examples of categorized words and their frequencies are presented in Table 1. During the analysis, there were minor disagreements between the coders, which were resolved through discussions involving the first author. All words were initially Mandarin Chinese and were translated into English for publication.
Table 1
Each group’s summary of word association results
Pretest
Posttest
Category
N
Frequent words (n > 2)
N
Frequent words (n > 2)
Image Group
Negative (N)
50
Hot (5), lost (4), hopeless (3), extreme climate (3), warming (3)
Using Python dictionaries, we stored all unique words along with their count and the words they were associated with (i.e., mentioned by the same student) in the Image and Text GenAI pre and posttests. These unique words amounted to 85, 79, 84, and 84 words in the Image pretest, Image posttest, Text pretest, and Text posttest. respectively. We then used this stored information to draw four network diagrams in which each unique word is depicted as a node, its size indicating its frequency of occurrence. The connecting lines between words, referred to as edges, represent the co-occurrence of two words and their width highlights how frequently these words co-occurred in the responses. Each word is colored according to its category (“positive,” “negative,” and “neutral”). The network diagrams were drawn using the spring layout algorithm from the NetworkX library (Hagberg et al., 2008).
We additionally formulated a numerical variable for each student by assigning a value of + 1 for every “positive” word, a value of − 1 for every “negative” word, and a value of 0 for every “neutral” word in their responses. These values were then summed together. A higher resulting value signifies a more positive feeling toward the future. This variable was generated to facilitate further statistical analysis with other measures.
Student responses to open-ended questions in the posttest were generally short, consisting of 2–5 sentences for each response. An initial examination of the texts revealed several recurring meanings, leading to the selection of thematic analysis as the method to analyze these qualitative data. This process involves identifying topics or ideas (themes) that come up repeatedly.
Results
Associated Words to the Future
Students have used a wide range of words to describe their perceptions of the future in relation to climate change. Table 1 provides an overview of the categorization of these words under “negative,” “positive,” and “neutral,” including word frequencies and the most frequently used words provided by the students. Notably, there was an increase in the use of positive words, along with a thematic shift toward environmental sustainability and technological innovation in the posttest phase for both groups.
The frequency of positive words increased substantially from the pretest to the posttest, with counts rising from 44 and 47 to 95 and 89 for the Image and Text Groups, respectively. The qualitative change is evident when comparing the most frequently mentioned word in the pretest, “convenient,” which was overtaken by “green” and “green-energy” in the posttest. Additionally, the word “sustainability” was broadened into a variety of related words such as “environmental-protection” “carbon–neutral,” or “carbon-reduction,” “energy-saving,” and “electric-car.”
By contrast, word frequencies in both “negative” and “neutral” categories decreased from the pretest to the posttest, as indicated in Table 1. Words like “hot,” “extreme-climate,” and “fearful” are dominant thoughts in the “negative” category before and after the instruction. Similarly, “technology” continued to dominate the “neutral” category, while the word “unknown” clearly decreased after the instructional intervention.
A visual representation of the data in the form of network diagrams was included for both the pretest and posttest in Fig. 2 (Image Group) and Fig. 3 (Text Group). A shift to a more positive and sustainability-focused perspective can be observed by comparing the pre-post diagrams. One notable result is that, although the word “technology” appeared as a dominant node in all tests, it was frequently connected with words such as “rapid” (Image Group) and “progress” (Text Group) in the pretest, while it exhibited stronger connections with “green-energy” (Image Group) and “carbon-reduction” (Text Group) in the posttest. Another example is the word “extreme-climate,” which initially had co-occurrences with words like “rising-sea-level,” “hot,” “lost,” and “despair.” After the instructional intervention, it was mentioned together with words like “innovation,” “green-building,” “energy,” and “green,” indicating a change toward a more mitigation-oriented outlook among students. These comparative results provide a clear view of the changes and trends in student responses following instruction.
Fig. 2
Image Group’s pretest and posttest co-occurrence network diagrams for word association (green: positive; red: negative; grey: neutral)
Hope and Action Competence Regarding Climate Change
Perceptions of the Future (Based on WA)
For each student, a WA score was calculated by assigning values to each word provided in their responses: a positive word received a score of + 1, a negative word received a score of − 1, and a neutral word received a score of 0. A higher WA score indicates a more positive outlook on the future.
Overall, the results of this variable (Table 2) revealed significant and substantial improvements in both the Image and Text Groups. This indicates a positive shift in their outlook toward the future following the instructional interventions. The large effect sizes reported (Cohen’s d = 0.75 for the Image Group and 0.51 for the Text Group) highlight the instructional interventions’ substantial influence on enhancing participants’ future perspectives.
Table 2
The summary of paired-sample t-tests and effect sizes
Pretest
Posttest
Variable
Group
N
M
SD
M
SD
t
df
p
Cohen’s d
WA
Image
58
− 0.12
1.73
1.19
1.79
5.67
57
< 0.001
0.75
Text
62
− 0.05
1.68
0.97
1.68
3.99
61
< 0.001
0.51
H
Image
58
4.96
0.82
5.32
0.90
4.04
57
< 0.001
0.53
Text
62
4.77
0.82
5.08
0.82
3.13
61
0.003
0.39
PW
Image
58
5.47
1.04
5.72
1.04
2.59
57
0.012
0.34
Text
62
5.23
0.99
5.56
0.97
2.44
61
0.018
0.31
CW
Image
58
4.83
1.02
5.28
0.99
4.07
57
< 0.001
0.54
Text
62
4.70
1.11
5.13
1.03
3.18
61
0.002
0.41
LW (r)
Image
58
4.57
1.49
4.95
1.47
2.35
57
0.022
0.31
Text
62
4.42
1.36
4.53
1.57
0.61
61
0.541
–
AC
Image
58
3.96
0.66
4.04
0.66
1.68
57
0.098
–
Text
62
3.88
0.66
3.94
0.76
0.61
61
0.544
–
KAP
Image
58
4.18
0.61
4.21
0.83
0.36
57
0.719
–
Text
62
4.04
0.59
4.08
0.82
0.33
61
0.745
–
COI
Image
58
3.68
0.83
4.04
0.92
3.77
57
< 0.001
0.50
Text
62
3.67
0.89
3.82
0.89
1.32
61
0.193
–
WTA
Image
58
4.03
0.72
4.22
0.72
2.74
57
0.008
0.29
Text
62
3.93
0.69
3.93
0.81
0.00
61
1.000
–
R = reversed; higher scores for LW reflect reduced disempowerment and greater hope
Hope (H)
H represents participants’ climate change hope levels. An overall improvement was observed in both the Image and Text Groups, with the former exhibiting stronger effect sizes than the latter (Cohen’s d = 0.53 and 0.39). Specifically, the Image Group showed significant improvements in all three subcategories: personal-sphere will and way (PW), collective-sphere will and way (CW), and lack of will and way (LW). As LW was reverse-scored in the analysis, a higher score indicates a reduction in perceived lack of hope and a greater sense of empowerment. These findings suggest that participants felt more empowered and motivated to engage in both individual and collective actions on climate change, with a notable reduction in the perception of climate change as an overwhelmingly complex issue beyond our control. The moderate effect sizes (Cohen’s d = 0.34, 0.54, and 0.31) reflect the meaningful impact of instructional interventions. For the Text Group, significant improvements were also observed in PW and CW (Cohen’s d = 0.31 and 0.41), but not LW, indicating a persistent feeling of disempowerment or a lack of efficacy in addressing climate change.
Self-Perceived Action Competence (AC)
AC represents participants’ self-perceived competence to take action on climate change. Only the Image Group showed significant improvements in two subcategories: confidence in one’s own influence (COI) and willingness to act (WTA) after instruction (see Table 2). This indicates that participants in this group felt more confident in their capacity to influence climate change development and became more willing to engage in creating a positive impact. The moderate effect sizes (Cohen’s d = 0.50 and 0.29) underscore the meaningful impact of the instructional intervention. Although neither group improved on one subcategory knowledge of action possibilities (KAP), the average scores are highest among all subcategories. This suggests that students generally perceived themselves as well-informed about action possibilities regarding climate change. The ceiling effect may explain why no significant improvement was observed.
Perceived Learning Experiences and Learning Gains
1.
What do you think exploring the carbon–neutral futures has affected your thinking about climate change?
We thematically analyzed all student responses to this question in the posttest, resulting in the following major themes:
Optimism about the future: Many passages expressed feelings of hope and positivity about the future, leading to an elevated willingness to make changes. The following are two excerpts:
I hadn’t really thought about what the future would look like before, but through discussions with the group, I came up with a lot of interesting points, such as new professions, such as carbon neutral inspectors. (IS71)
We often think about how bad the future will be if we don’t limit [emissions] now, but we rarely consider what it would be like to achieve a carbon-neutral future. I feel more empowered when led to think in a more positive and hopeful direction. (TS44)
Personal responsibility and desire to change: Some statements emphasized personal actions, self-reflection, and the desire to make a positive impact. The following excerpt vividly illustrates this theme:
I feel that during the exploration process, I have learned many different perspectives, allowing me to think and reflect from multiple angles. I also believe that only by achieving carbon neutrality as quickly as possible can we provide a good environment for future generations to live. (IS35)
Awareness and understanding: Some student responses provided insights related to gaining knowledge, understanding the implications of climate change, and identifying action possibilities. The following are illustrative excerpts:
I can identify areas in my daily life where there might be too much carbon emission, and then reflect and address them. (IS6)
I realized that energy saving and carbon reduction can actually be implemented all-around in our life. When the concept of energy saving or practical devices are integrated into our daily life, perhaps it will not be so difficult to improve the climate problem, and everyone can do his/her part. (TS86)
These themes illustrate students’ perspectives on their learning experiences and gains from the futures-focused module (see Appendix for category descriptions and response examples).
2.
In what way do you think using the genai tool helped or limited the visualization of your future scenario?
For this question, we analyzed two groups separately in order to identify the benefits and limitations of integrating image or text GenAI in future thinking activities. It was observed that both groups emphasized “efficiency” and “enhanced creativity or imagination” as key strengths of using GenAI. The Image Group also frequently mentioned “fun” while the Text Group considered ChatGPT is helpful with “information synthesis.” Interestingly, the Image Group focused on the benefits rather than the limitations as compared to the Text Group.
Image Group
Strength: Efficient Visualization, Enhanced Creativity, and Fun
Students in the Image Group highlighted how GenAI helped them efficiently generate images that aligned with their ideas, saving time and enhancing creativity. Some illustrative comments include.
It integrates the scenes in my mind with the concepts I’m thinking of in a short time and present them (IS64)
It saves the time needed for drawing and materializes our thoughts (IS21)
Using AI allows us to have a more concrete image of the future for further discussion and effort. (IS33)
Some students found that the use of AI added an element of fun, making the activity more engaging and interesting overall. For example,
Although I had some ideas in mind, AI used my keywords to create images that were really surprising. They were more interesting and better pictures, and added creativity to my thoughts. (IS62)
Through the tangible images, we can receive the imaginations of each student, a very interesting activity! (IS7)
Limitations: Mismatch Between AI Visualization and Mental Images
Despite these strengths, some students pointed out that AI-generated images often did not fully capture the ideas they had in mind. Some noted that the images felt unrealistic or lacked important details, while others observed that the GenAI occasionally misinterpreted their input, resulting in visuals that did not match their envisioned scenes. Students recognized that providing more specific and detailed prompts could help reduce these mismatches. They commented, for example,
It feels more like animation, lacking a bit of a sense of realism (IS3)
AI might not always be able to accurately represent the keywords we provide, or it might misinterpret our keywords, resulting in generated images that don’t quite match what we had in mind. (IS42)
If you want to generate a more detailed and accurate representation of the future city you have in mind, you’ll need to provide more descriptions and metaphors to achieve this. (IS24)
Text Group
Strengths: Enhanced Imagination, Efficiency, and Information Synthesis
Students in the Text Group valued ChatGPT’s ability to synthesize information and provide imaginative ideas for their narratives. They appreciated its efficiency and how it helped them focus their thinking. Some examples include.
It is very efficient and very imaginative, it can quickly consolidate our requests and respond immediately, even if it is not always correct, it helps me to focus on more issues. (TS36)
It provided some ideas for story situations to supplement my imagination. When it becomes an article, the picture will come out even more! I think this part is great, but the viewpoint of the story is shaped by what I feed it, and so still a bit limited in terms of ideas. (TS63)
Limitations: Lack of Depth and Dependence on Specific Inputs
Many students in this group recognized limitations in AI-generated content, noting that it sometimes lacked originality or depth and that it relied too heavily on existing internet-sourced information. Some students expressed concerns that this dependence could limit creative thinking and prevent the generation of truly innovative ideas. For example,
AI-generated narratives help to instantly stimulate our thoughts, but content-wise seem to lack depth and authenticity. (TS39)
…Using AI may limit our imagination, not only we rely too much on it, but also the data that AI refers to is only limited to the information available on the Internet, so there may be a lack or shortage of out-of-the-box frameworks and thinking. (TS81)
Discussion
This study investigated the impact of a futures-oriented teaching module incorporating Image and Text GenAI on students’ perceptions of the future, their hope, and self-perceived action competence regarding climate change. The results provide valuable insights into how AI tools can facilitate futures thinking and foster engagement with complex global challenges like climate change.
Shift toward Positive, Sustainability-Oriented Futures Outlook
The significant increase in positive words and the thematic shift toward sustainability and technological innovation in both groups underscores the potential of the teaching intervention to foster a more optimistic and solution-oriented mindset. This aligns with prior research, which shows that futures-oriented teaching helps students transition from seeing the future as distant and abstract to perceiving it as something tangible and within their control (Liu, 2022; Fletcher, 2019; Pahl & Bauer, 2013). In this study, students explored possible futures through the lens of IPCC-projected pathways. By systematically examining these scenarios, identifying key drivers of change, and envisioning their own desirable futures, students engaged in a structured process that strengthened their ability to connect present actions with future consequences. Through this process, students develop “future-scaffolding skills,” which integrate futures thinking and agency, enabling them to envision actionable paths toward sustainable futures (Levrini et al., 2019, 2021). The increase in words like “green-energy” and “carbon–neutral” further reflects the effectiveness of the intervention in promoting a more positive and proactive view of future challenges.
The observed changes in the network diagrams also illustrate this shift. Terms like “technology,” initially associated with rapid development or progress, later became linked to sustainability concepts such as “green-energy” and “carbon-reduction.” These shifts highlight how futures thinking, when scaffolded effectively, allows students to reframe challenges as opportunities for innovation and action.
Impact on Hope and Action Competence
The improvement in students’ hope and self-perceived action competence further reinforces the value of the module. These findings align with prior research on futures-oriented teaching and learning, which indicates that systematically guiding students through futures exploration helps them feel more empowered and capable of influencing the future (Liu & Lin, 2018; Liu, 2022; Levrini et al., 2021; Paige et al., 2018). As Gidley, 2004 concluded in a systematic review of futures/foresight education, examining diverse future scenarios can help counteract young people’s fears about the future, foster a greater sense of hope and possibility, and inspire meaningful actions toward a hopeful future. More specifically, envisioning life within a positive, sustainable future scenario through scaffolded learning helped ensure that their sense of hope was grounded in understanding—what Ojala, 2012a, 2015) terms “constructive hope”—rather than in ignorance or denial.
The finding that the Text Group did not significantly improve in action competence may support the idea that “a picture is worth a thousand words.” This suggests that the visual tool used by the Image Group, with its simpler, more direct representation of future scenarios, may encourage a more immediate emotional or action-oriented response than text alone. While ChatGPT allowed the Text Group to articulate their ideas in detailed, sequential narratives, the Bing-created visuals likely made future scenarios more tangible and compelling, potentially fostering a stronger drive toward action. This suggests that combining both AI formats may be one way to enrich futures-oriented teaching, offering students both conceptual depth and motivational impact.
Developing AI Literacy through Meaningful Interaction
A crucial outcome of integrating GenAI tools in this study was the development of AI literacy among students in an applied, authentic learning context. Rather than using AI for its own sake, students engaged with GenAI to generate real-world future scenarios that aligned with their ideas, helping them articulate and visualize complex futures that might have been difficult to construct independently. By interacting with GenAI in a structured learning environment, students were able to critically evaluate both the benefits and limitations of these tools. This hands-on engagement allowed them to see first-hand how GenAI can enhance creativity and efficiency, as well as where it falls short, such as in producing less realistic visuals or relying too heavily on pre-existing internet data. By iteratively refining their AI-generated outputs and comparing results across peers, students not only observed how AI-generated outputs changed in response to their prompts (becoming better or worse) but also how different inputs led to future scenarios that were often interesting and meaningful, yet sometimes repetitive. This mirrors findings from broader research, where GenAI tools are used not only to foster creativity and problem-solving but also to help students develop critical thinking skills by evaluating AI-generated content (Kong et al., 2023; Wood & Moss, 2024; Xia et al., 2024).
Through this experience, students in both groups gained a more genuine understanding of GenAI’s potential and constraints, recognizing that AI tools, while powerful, have limitations that must be approached with caution. This type of interaction can contribute meaningfully to AI literacy, which has been highlighted as an essential skill in modern education (Casal-Otero et al., 2023; Walter, 2024). Given that students are frequently exposed to AI-generated futures through media without engaging in deep, critical thinking, these findings suggest the importance of integrating AI into science and environmental education. In contexts where future envisioning is key, this approach can foster both critical AI literacy and more thoughtful engagement with sustainability challenges, empowering students to actively shape the future rather than passively consuming it.
Conclusion
This study assessed the impact of a futures-oriented teaching module, integrated with GenAI, aimed at fostering positive outlooks on climate change, enhancing students’ hope, and building their self-perceived action competence. The scaffolding module, which incorporated activities focused on exploring possible futures and envisioning sustainable futures, effectively shifted students’ perspectives from a predominantly negative outlook to a more solution-focused and optimistic view. This shift is evident not only in word counts but also in the increased use of sustainability-related terms and the thematic changes in students’ word associations. The overall findings indicated that the module supported students in developing constructive hope and a sense of agency in addressing climate challenges.
In examining the specific tools used within the module, both the Image and Text GenAI tools were integrated into the module, with different groups of students using either Image GenAI or Text GenAI. This distinction allowed us to explore whether these tools would have varying impacts on desired learning outcomes, though their impact was not the sole focus of the intervention. Post-instruction reflections provided valuable insights into students’ perspectives on the benefits and limitations of these tools, indicating the learning gain on their AI literacy.
Taken together, the study sheds light on the effectiveness of structured, futures-oriented teaching with the integration of GenAI in empowering students to envision actionable paths toward sustainability. The findings reinforce the importance of educational approaches that not only equip students with knowledge but also inspire hope and a sense of competence in addressing global challenges like climate change. They further illustrate GenAI’s potential to engage students in meaningful future envisioning. Future research could further explore how various educational tools, including but not limited to AI, can support these goals.
Declarations
Ethical Approval
The Human Research Ethics Committee under the Taiwan Ministry of Education granted ethical approval for this study (approval number 110–224).
Consent to Participate
All participants were informed about the data collection and gave consent to the collection and use of data. All data was anonymized. All participants took part in the study voluntarily.
Consent for Publication
Not applicable.
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Increased hopefulness and a positive outlook on climate action
“I hadn’t really thought about what the future would look like before, but through discussions with the group, I came up with a lot of interesting points, such as new professions, such as carbon neutral inspectors.” (IS71)
“I started to feel that the future is actually easier to achieve than we imagine. The goal of carbon neutrality feels more attainable.” (T26)
“We often think about how bad the future will be if we don’t limit [emissions] now, but we rarely consider what it would be like to achieve a carbon–neutral future. I feel more empowered when led to think in a more positive and hopeful direction.” (TS44)
Personal responsibility and desire to change
Emphasis on individual roles in addressing climate change and motivation to take action
“(…) helped me reflect on future development, and the role we play as individuals. I feel this more deeply.” (IS19)
“I feel that during the exploration process, I have learned many different perspectives, allowing me to think and reflect from multiple angles. I also believe that only by achieving carbon neutrality as quickly as possible can we provide a good environment for future generations to live.” (IS35)
“The biggest impact of exploring carbon neutrality on me is that it has allowed me to see that my own changes [of behaviour] can indeed improve the environment. This has also motivated me to take action in my daily life.” (TS66)
Awareness and understanding
Greater recognition of climate-related challenges and solutions
“I can identify areas in my daily life where there might be too much carbon emission, and then reflect and address them.” (IS6)
“It made me think more about what we can achieve within our capabilities and what current technologies can realistically support.” (TS17)
“I realized that energy saving and carbon reduction can actually be implemented all-around in our life. When the concept of energy saving or practical devices are integrated into our daily life, perhaps it will not be so difficult to improve the climate problem, and everyone can do his/her part.” (TS86)
Electronic supplementary material
Below is the link to the electronic supplementary material.
