Transmitting Everything You Say

@davidrevoy @elzen now I imagine the Avian Intelligence as a drunken bird, as it is fed with the "distilled" liquid extracted from the AGI.

@davidrevoy dang she's cute

@davidrevoy "statistically" is probably the operative word there, since that's more or less what they are... statistical word generators.

They'll never twist or surprise because they're going for the most likely output from the millions of artists they've been fed (and increasingly, re-ingesting slop).

(Edit: I probably didn't need to say all that, did I? Derp.)

Gotta say the Avian Intelligence mini has been spot-on and a delight.

@RandomDamage @Sobex @lazy

@davidrevoy
Is Pepper getting..... Spicy?

I'm quite worried though... Who's been listening and how will they torment Pepper with the knowledge they've stolen?

@tiredbun @voxel I hope I'll can draw enough episodes in this mini series to explore many issues (addiction/psychosis/and more) on user side and be able to land a conclusion aligned with something that was really well expressed on this toot: https://lgbtqia.space/@alice/115845072659379702 , something nuanced.
I hope by the end, the Gothic Sorceress will have a responsible usage of it, and maybe her own 'local' one. Maybe even themed like a crow. I'll see, so many things to draw to reach this point! I'm working on it.

@VANTABlack2000 Thank you!

@davidrevoy @Sobex @solitha @lazy of course! I'm sure if you follow this thread long enough everything will come out.

It's been really fun so far

@davidrevoy @RandomDamage @solitha @lazy

**Canon Updated !**

@Sobex @lazy @solitha @davidrevoy Maybe she’s a postdoctoral fellow – desperately going from one postdoc position to another because there’s almost no permanent positions anymore.

(Any similarity with the current academic landscape would be of course purely coincidental.)

@davidrevoy You're comics are lovely, thank you!

@davidrevoy "I contain no data source-serviceable parts"?

#miniFantasyTheater #AvianIntelligence

@davidrevoy High-quality data!

@davidrevoy

"waiiit a minute, are you telling the truth now?"

"Yes"

"What about yesterday?"

"Reply hazy, try again"

"Wait, what?"

"Iv'e been bewitched to be honest about my transmission capabilites, but not about anything else"

@davidrevoy @RandomDamage @solitha @lazy

Though, re-checking all the story, the goth witch looks too young for a post-doc, and also her relation to the info provided by the parrot is really what you might expect from a bachelor or master student, but not someone with a PhD under her belt. A fortiori a permanent researcher would have the ability to check a good chunk of the parrot nonsense, and avoid incinerating herself.

So my personal take, given what we have now could be a witch at the beginning of her PhD, or in her Master's (since she's getting a bit more assertive). A PhD student is a lot closer to faculties and could well be invited for the new year party she's in.

@davidrevoy Privacy is dead?!Nooo, surely just prankster.

@davidrevoy

@davidrevoy Maintenant, tout le monde veut connaître le contenu de la discussion salace.

@davidrevoy @tiredbun @voxel I've seen enough blatantly wrong LLM-generated alt text and spell check results that I don't think the technology is fit for purpose. Any purpose. AI "hallucination" is not a solvable problem: It's the exact thing LLMs are designed to do. So, with all due respect to the person being quoted here, I don't trust anyone who would entrust something as important as accessibility to AI.

@Sobex For sure, having her in her late years of study could be an advantage if I decide to show other students, and their relation with their Avian Intelligence later. So, I'm not closing this door. I'll see, writing is difficult, and sometime I have to take direction for a good 4 panels and... the lore will follow later Thank you for the feedback!

@RandomDamage @solitha @lazy

@Linebyline I totally agree, this is a good point.
For sure, I think LLM-generated alt text should only be used as, eg. , a quick placeholder text to fill up an editing text-area, requiring a human to post-fix it. I strongly wish the output of any LLM should never go into public or production unsupervised. A human should always be required in the last chain of the process with a strong expertise on the task to critic, rewrite, adapt and make it better.

@tiredbun @voxel