Jacque, well said. And I am also happy. It is a giant step forward for the
platform (especially when you add in the other new capabilities). Congrats to
everyone who contributed, and to Kevin and all the LC team. We made it happen.
> On 18 May 2023, at 6:22 pm, J. Landman Gay via use-livecode
My day to day work is in medical education – and assessment specifically. I’m
interested in using LLMs to (1) help generate assessment items (well draft
versions of them at least) and (2) individualised descriptive feedback based on
assessment performance.
But, rather than trying to get
I used ChatGPT to generate some test data for me.
It took several tries before it was able to give me
a dozen lines of useable data.
I told it to separate each item with a comma and
each line with a carriage return.
Then I just copied and pasted the data into a
spreadsheet program,
If you watch some of the videos posted on the challenge page it becomes
clearer. LC AI will be integrated into the script editor and will create
handlers as expected. You can accept, reject, or revise its output. LC AI
knows all about what's available and how to use it. Bard, on other hand,
In Jacque’s “test” AI first had to create a list of ten “people”, garnered, I
assume, from research into our planet’s knowledge base. It knew what a “name”
should be. That is cool.
That list could have been a list of actual humans, derived from, say, the staff
of a particular bank.
If the
On 5/17/23 5:38 PM, Terry Judd via use-livecode wrote:
Let’s say I have a whole lot of input statements/info, some of which is generic
but some of which varies according by user. Any thoughts on whether it might be
possible to leverage Xavvi to process and combine that info into useful prose