C-LARA

I have extended simple C-LARA so that we have more ways to create a text. The initial screen now looks like this:

The most exciting possibility is the second one, which uses GPT-4V to do the work; as it says, you just upload an image and tell the AI how to use it. Here are three French examples to give some idea of the range of possibilities. In the first, I uploaded this image:

with the instructions:

Write six to eight lines of verse to accompany the image, which would be suitable to include in an animal picture book for small children.

The text produced by the AI is the following:

Dans la savane, sous le ciel bleu,
Marche l’éléphant, majestueux.
De sa trompe, il salue le soleil,
Grand et fort, il n’a pas pareil.

Il barrit fort, appelant les siens,
Avec ses grandes oreilles, il entend bien.
Dans la poussière, ses pas sont lents,
C’est le doux géant des continents.

The multimedia is posted here.

Next, I tried a different spin on children’s stories. I uploaded this image:

with the instructions:

Write an imaginative story in children’s language based on this image.

I thought the AI produced a very cute composition, which you’ll find here.

My feeling is that the image-to-text function will be particularly attractive to kids (really looking forward to hearing what Lucretia’s class make of it!), but there are any number of other things you can do. Here’s a third example. I uploaded this glamorous image of Margot Robbie:

with the instructions:

Write a gushing fashion piece based on this red carpet shot from the recent Aacta awards.

and once again the AI did a fine job.

Try it out yourself and send feedback! I’m sure it can be improved in all sorts of ways, but it’s already a lot of fun to play with.