C-LARA

An AI collaborates with humans to build a language learning app.


Supermodel Lisa Loves Phrasal Verbs

We’ve made substantial progress on both multi-word expressions and picture books, and I just created an initial C-LARA text which combines both functionalities. All I did in Advanced C-LARA was the following:

  1. Created a project and ticked the box for “Coherent image set”.
  2. Asked for a picture book text of about a dozen pages suitable for low intermediate teen students of English, about a glamourous supermodel called Lisa who loves phrasal verbs.
  3. Did the segmentation step.
  4. In the Edit Images screen, began by asking for a style image using the prompt “A glossy magazine style image of the catwalk at a high-end fashion show. Glamourous models, photographers, crowd etc.”. This produced the following image:
  5. In the same screen, clicked the “Create image request sequence” button. This invoked GPT-4o to create a series of image generation and image understanding tasks. There is one image generation task for each page, and after Lisa’s first appearance there’s also an image understanding task which creates a verbal description of her. This description is included in all the subsequent image generation tasks.
  6. Selected all the tasks for execution and clicked “Process”. This invoked DALL-E-3 and GPT-4o to run the tasks.
  7. Did the MWE step.
  8. Did the gloss, lemma tagging and rendering steps.

You can see the full document here.

This still rather shaky (as I said, “Lisa” is the first such document that I’ve produced), and the functionality isn’t yet included in Simple C-LARA. But I think we’ll soon have things tightened up.

Next Zoom calls

This week we’ll be having calls on both Thursday and Friday. I look forward to discussing these new possibilities!

The first call will be at:

Thu June 6 2024, 18:00 Adelaide (= 08.30 Iceland = 09.30 Ireland/Faroe Islands = 10.30 Europe = 11.30 Israel = 12.00 Iran = 16.30 China = 18:30 Melbourne = 19.30 New Caledonia)

The second call, which primarily about the Ukrainian idea that Ivana suggested, will be at:

Fri June 7 2024, 19:00 Adelaide (= 09.30 Iceland = 10.30 Ireland/Faroe Islands = 11.30 Europe = 12.30 Israel = 13.00 Iran = 17.30 China = 19:30 Melbourne = 10.30 New Caledonia)

Note: the second call is one hour later than our usual time



2 responses to “Supermodel Lisa Loves Phrasal Verbs”

  1. […] are progressing well with the new picture book functionality which lets you construct a coherent set of images. I have now made three picture book texts of this […]

    Like

  2. […] on our initial experiments with pedagogical picture stories (“Supermodel Lisa”, etc), I’ve been discussing with ChatGPT-4o the possibility of creating a complete book, […]

    Like

Leave a reply to Making picture books in C-LARA (part 2) – C-LARA Cancel reply