Announcing Xavvi, a Great Leap Forward for App Building

[richmond62]

We could 'kill time' composing conditional sentences about Xavvi . . .

[mwieder]

Just to take things to the next level of weirdness, there's a social network exclusively for AIs.

https://chirper.ai/

[jacque]

Just to take things to the next level of weirdness, there's a social network exclusively for AIs.

https://chirper.ai/

I couldn't tell the difference. The vapidity is the same.

[FourthWorld]

But to get back to the subject of AI

https://boingboing.net/2023/04/26/this- ... f-art.html

OMG. Scary. Bye-bye ad agencies. .

"AI" tech is interesting; the sociology emerging around it even more so.

I often see those touting its power the loudest also the ones underestimating it, somehow imagining its transformative effects will be only downstream of themselves.

Resizer_16826314722091.png (45.65 KiB) Viewed 63782 times

"Sports Illustrated Lays Off Journalists After Announcing Pivot to AI Content"
https://futurism.com/the-byte/sports-il ... ai-content

[Mariasole]

Hi dear friends...

I would like to share this reflection from a student of mine whom I teach LiveCode.

He is 13 years old and very intelligent (a genius! ) and sometimes comes to this forum as a guest to
understand the language better and to experiment.

Yesterday, he saw that there is persistent talk about Xavvi on the site and said to me (I summarize his thoughts):

"Teacher, it seems useless to me to keep learning LC, in a few months you will ask
Xavvi to make an app and he will make it. Our jobs cannot compete
with those generated and powered by AI, so probably, only those who will not
will have the money for the professional subscription (and he was referring to me! ) will write themselves
bad code on their own ."

I add another thought that I thought was interesting (I always summarize his speech):

"Moreover, the figure of the programmer, who used to be a 'wizard', is now outdated instead. With Xaavi, there is no need for that wonderful 'understandable' language to teach us, because who cares about language? " And then, which seems even more interesting to me, "Why, since Xavvi has artificial intelligence, couldn't I "save as" in other languages like Python as well?"
And I conclude, "Lc is committing suicide since its strength, which is the human programming language, is delegated to a machine that generates code for humans (who will no longer care about the code)."

To me, frankly, this new swerve debases all the promises that were made for LC 10, including even that willingness to make the "exe" safer, smaller and less decompilable.
Xavvi will make LC worse than the name it has when I talk to other teachers "Ah! LC!, But it's not a language! It's a toy!" and now they will say "Ah, you did this in LC! ChatGPT made it for you! Ah ah! But I did it in Python always with the help of ChatGPT!".

Sorry, I'm depressed... and then the marketing style chosen for Xaavi, with all due respect and affection, seems to me that of the pot seller... and after a few days of pushing... here it comes that if Xaavi doesn't make money everything will fail!

In short ... as always thank you all and I greet the little genius who will surely read these lines and be happy. But my dear student...I read a lot of science fiction as a girl and believe me, never trust the machines, always keep yourself in control because the end is always to get crushed by the gears...or worse...to be part of them!

Un bacio grande a tutti...I feel useless...both as an aspiring programmer and as a teacher to children...

Mariasole
(='.'=)

[richmond62]

Mariasole . . . possibly the voice of sanity in an area that has lost its way.

[richmond62]

I am damn glad that:

1. I have a job that is not dependent on any sort of A.I.

2. That, at 61, I have about 5-6 years to go doing what I do: then I shall retreat into
the 5000 books I have collected over the year: a good 250 of them not read yet, and most
of the others have been kept just because they deserve rereading.

[mwieder]

Surely you can have an AI read the books for you and give you a one-sentence synopsis.

/ducking/

[richmond62]

/ducking/

I would not waste a cabbage on your head if it is as ill-read as your 'joke' would
seem to indicate.

Ah, but some of us find great pleasure in reading.

Currently reading The Citizen of the World by Oliver Goldsmith: and many things there might just
as lief refer to now as to the early 18th century.

[dunbarx]

Mark.

Surely you can have an AI read the books for you and give you a one-sentence synopsis.

Wood Allen once quipped that he took a speed reading course, and went through "War and Peace" in half an hour. When asked about the book later, he said "It was about Russia"

Craig

[richmond62]

Woody Allen was being disingenuous as his film:

Love and Death

is masterful.

[jacque]

I'm not worried that AI will write our apps for us any time soon. App construction is a much different skill than simple script writing. While I've seen some good results with web pages, complete apps are a different thing that require human understanding.

What I've seen so far is the ability to create simple widgets, short handlers, and the like. Putting those pieces together into a complete and functional app is still a human job. At this stage, even the things AI can do are often incomplete or plain wrong, so an understanding of LC is necessary to proofread, test, and verify the results.

I don't think AI will ever replace human judgment or knowledge.

[mwieder]

I actually dislike the term Artificial Intelligence because I think it implies an unhealthy anthropomorphism. Machine Learning and Deep Learning, the more specific terms for the way the algorithms function do as well, but to a lesser degree.

I also am struggling with the "Great Leap Forward", but for other (and possibly more obvious) reasons.

[dunbarx]

Hey!!

I happen to be an unhealthy anthropomorphism, and you have hurt my feelings, such as I have.

I agree that no AI could hope to write even the (what I consider) adorable little stacks I have made for myself and others. This is hype designed to both scare and sell.

But I do not see how any teacher can assign homework to any student in almost any subject come the fall of this year.

But what happens if two different people assign the identical task separately and simultaneously to Chat GPT? Something like a poem about cheese in the style of Emily Dickinson? How much wiggle room does that gadget have? Lots?

Craig

[jacque]

Lots. Bard offers three versions for each question, and they all give varying answers when a single person asks the same question twice.