During the SuperMemo Morning session, we discussed the role of new technologies in language education. We explored the idea of enhancing the Extreme language course series with graphic illustrations for all vocabulary items. This addition would help learners associate new words with images, making the learning process more effective. Previously, financial and technical limitations made this impossible, but advancements in generative artificial intelligence are opening up entirely new opportunities.
One of the AI models we are testing is Flux, which enables the generation of high-quality graphics. We’re working on our own server equipped with an appropriate graphics card, and the initial results are promising. For example, with vocabulary such as “Full Moon” or “Breakfast at 10 am,” Flux generates images of satisfying quality—perhaps even better than popular tools like DALL-E 2 or DALL-E 3, which can be costly for generating a large number of images.
While Flux produces impressive results, it is not without flaws. Occasionally, distortions appear, such as in the positioning of hands or body shapes in illustrations. These are areas we are actively addressing, and the system is continually being optimized. If all goes well, users of the Extreme series courses could start enjoying these new illustrations within a few weeks, enriching their language learning experience.
Stay tuned for more updates—we hope you’ll soon see the results for yourself. We also invite you to join SuperMemo Mornings, held in Polish, Monday through Friday at 8:50 AM on Google Meet – LINK.
These sessions aim to help develop the habit of daily learning and provide insights into new SuperMemo features.
