I love that we get to finish this course with a task related to AI, especially how it pertains to the future and predictions AI is able to create. I’ve never thought about asking AI models about it’s prediction for the future, but I think this really does open up a door for it to use it’s information and data to find similarities in order to make an ‘educated guess’. Out of curiosity, I decided to approach this concept by looking at multiple models ideas.
For this task, I wanted to not only use one AI GPT, but I wanted to see what some of the most popular and advanced GPT services provide. Using this website Tracking AI, it actually constantly checks the AI IQ using daily tests. This means that it runs many tests, like the Turing Test, to consistently collect data on the intelligence and accuracy of the responses in order to compare the GPT quality of each provider. It provides data like questions asked, responses, and progress over time. With this idea in mind, I decided to compare the top 5 GPTs listed on TrackingAI (excluding Microsoft CoPilot) to see how and what each AI model predicts based on the same prompt, both written.
Here are my results:
One thing that stood out across all 5 responses is how each model handled the idea of a future shaped by “progress” and what that does emotionally. Dunne and Raby (2013) say speculative narratives should open up possible futures rather than try to predict them. When I look at these stories together, each one sits in a different corner of that speculative space. Copilot and Grok lean toward the darker, high intensity possibilities where their futures feel more like warnings. Copilot imagines a society built on covered up genetic failures, while Grok goes into full panic, with drones, blood samples, and a sense that the state is ‘always watching’. Both tap into the “provocative scenario” side of speculative design that tries to ‘wake up’ the viewer. From my understanding, this is the zone that Mitrovic et al. (2021) mention that caution can sometimes drift toward spectacle, and you can feel that tension in Copilot and Grok.
Identity also becomes a shared thread. The quieter stories, especially Claude and ChatGPT, explore what progress means for personal identity. Their protagonists are anxious not because the world is collapsing but because they see what was lost. Dunne and Raby’s (2013) idea of using speculation to create reflective space fits perfectly here. These stories feel less like predictions and more like mirrors reflecting our present attitudes about optimization and perfection.
Across all the photos, what stands out is how they visualize the range of speculative futures that Dunne and Raby (2013) describe, from provocative possibilities to quieter reflective spaces. Excluding the vague photo by Claude, some images lean into dystopian tensions while others focus on intimate, everyday materials, reflecting the readings’ point that speculative design can either confront viewers with dramatic scenarios or use small and familiar objects to spark reflection (Mitrovic et al., 2021). Together, the photos are examples that represent how visual design can guide the emotional tone of a narrative, shaping whether we experience the future as a warning, a question, or a mirror.
The variation and similarities between all 5 GPTs make me think about the types and focuses of each’s corpus. Since AI is currently only able to use information and data first created and inputted by human intelligence, it makes me curious on what each one used to base this off of. A lot of the themes and characteristics of each seem familiar and potentially taken from futuristic, Sci-Fi, and dystopian work. It makes me question: how much of this is just the models stitching together pieces of our own cultural imagination?
In relation to Core77, I also wanted to mention that I found it odd that there was no specific category dedicated to education. Although many of the categories’ descriptions mention and make broad connections to education, like Speculative Design, Tools, and Toys & Play, this lack of education specific category I believe to reflect the consistent delay between education and keeping up with emerging technologies. As I’ve mentioned many times, I’m a strong advocate for the potential of AI in education, especially the need for development, investment, and seriousness for more development in this field. If employers want employees who are knowledgeable, comfortable, and creative with AI technologies, they need to be part of the development of technology within institutions. In connection to Harari (2017), supporting education in terms of development and progress will address the future need of new job creation and lifelong retraining. I believe that the investment will result in a more advanced future and society.
References:
Anthropic. (2025). Claude. https://claude.ai
Core77. (2025). Speculative design award: Core77 design awards 2025. https://www.core77.com/posts/137473/The-2025-Core77-Design-Awards-Speculative-Winners
Dunne, A., & Raby, F. (2013). Speculative everything: Design, fiction, and social dreaming. The MIT Press. https://muse-jhu-edu.ezproxy.library.ubc.ca/book/28148/
Google. (2025). Gemini. https://ai.google.dev/gemini
Harari, Y. N. (2017). Reboot for the AI revolutionLinks to an external site.. Nature International Weekly Journal of Science, 550(7676), 324-327. https://doi.org/10.1038/550324a
Microsoft. (2025). Copilot. https://copilot.microsoft.com
Mitrovic, I., Auger, J., Hanna, J., & Helgason, I. (2021). Beyond speculative design: Past – present – future. SpeculativeEdu. https://speculativeedu.eu/wp-content/uploads/2021/06/Beyond-Speculative-Design.pdf
OpenAI. (2025). ChatGPT. https://openai.com/chatgpt
Tracking AI. (2024). Tracking AI: Monitoring artificial intelligence. https://www.trackingai.org/home
xAI. (2025). Grok. https://x.ai
