The Minions of Avian Intelligence
-
@davidrevoy
I found it interesting that on all the panels the birds are flying they are in the "wings open" phase of their flight. My head cannon is that that they've been trained on pictures of birds with their wings open (probably the ones people share the most) so when they fly they open their wings wide and just hover around like this
.@GeoffreyFrogeye Haha, I'll keep the idea somewhere if I ever animated an episode.
-
@SmartmanApps @Stem (2/2) This opportunistic usage comes also with understanding of the ecological issues, the training data issue, the big corp owning it issue; but they often rationalize that for them, using it provide them an individual edge, and give them more time.
So, this big majority tends to like reading things against AI and things pro AI too; for the entertainement of seeing it as inside an arena of idea. They feel outside of the debate.
It's like that for many topic, imo.
@davidrevoy @Stem
"a large majority are using AI by opportunity" - 2 different polls, both with thousands of votes, both put the figure at 10%, which is a small minority - only 1 out of every 10 people - and that's consistent with what I see in my feed as well -
@davidrevoy @Stem
"a large majority are using AI by opportunity" - 2 different polls, both with thousands of votes, both put the figure at 10%, which is a small minority - only 1 out of every 10 people - and that's consistent with what I see in my feed as well@SmartmanApps @Stem I don't think the poll is representative. People in our era ready to even participate to a poll and take action or getting a Pro/Cons mindset are the minority.
Most individual prefer to keeps anonymous, don't express publicly any preferences that could be used against them later, just pass their way and maximize their adaptation to the most dominant culture later, especially if this one become so dominant it starts a witchhunt of the adversarial group later. -
@encukou Thank you very much! TIL about the perspective effect. Do you know if this effect works on object like (dynamic, not converted as path) texts?
That was my main struggle in this episode: https://www.peppercarrot.com/en/miniFantasyTheater/038.html , where I had to use simple "shearing" (not sure of the term) deformation, to fake perspective.
@davidrevoy Sadly, it's only for paths.
-
@davidrevoy Sadly, it's only for paths.
@encukou Thank you for confirming, but great to know it will do for QRcodes!
-
@SmartmanApps @Stem I don't think the poll is representative. People in our era ready to even participate to a poll and take action or getting a Pro/Cons mindset are the minority.
Most individual prefer to keeps anonymous, don't express publicly any preferences that could be used against them later, just pass their way and maximize their adaptation to the most dominant culture later, especially if this one become so dominant it starts a witchhunt of the adversarial group later.@davidrevoy @Stem
"I don't think the poll is representative" - and you would be wrong about that. The DuckDuckGo one had 175,000 votes!"People in our era ready to even participate to a poll" in a random sample are representative of the population, once you have 25 votes. The DuckDuckGo one had 175,000 votes. Even the Mastodon one had 5,000 votes, and both got the same result
"Most individual prefer to keeps anonymous" - both polls were anonymous
-
The Minions of Avian Intelligence
Bonus speedpaint: https://www.peppercarrot.com/en/miniFantasyTheater/067.html#bonus
@davidrevoy
Is there any actual evidence that AI companies are doing that? -
@davidrevoy @Stem
"I don't think the poll is representative" - and you would be wrong about that. The DuckDuckGo one had 175,000 votes!"People in our era ready to even participate to a poll" in a random sample are representative of the population, once you have 25 votes. The DuckDuckGo one had 175,000 votes. Even the Mastodon one had 5,000 votes, and both got the same result
"Most individual prefer to keeps anonymous" - both polls were anonymous
@SmartmanApps@dotnet.social @davidrevoy@framapiaf.org @Stem@framapiaf.org But people using DDG or being on the Fediverse is already a huge bias…
-
@davidrevoy @Stem
"I don't think the poll is representative" - and you would be wrong about that. The DuckDuckGo one had 175,000 votes!"People in our era ready to even participate to a poll" in a random sample are representative of the population, once you have 25 votes. The DuckDuckGo one had 175,000 votes. Even the Mastodon one had 5,000 votes, and both got the same result
"Most individual prefer to keeps anonymous" - both polls were anonymous
@Stem@framapiaf.org @SmartmanApps@dotnet.social @davidrevoy@framapiaf.org I feel like that’s sampling bias, if the same polls ran on Twitter or LinkedIn it would give different results.
my anecdotes anyways but people around me are pro-AI than anti -
@SmartmanApps@dotnet.social @davidrevoy@framapiaf.org @Stem@framapiaf.org But people using DDG or being on the Fediverse is already a huge bias…
@breizh @SmartmanApps @Stem +1 Breizh
Smartman, don't get me wrong, I would love to believe in this poll's results. But in my local reality (at weddings, in my repair garage, at the restaurant, on week-end) when I listen to my non-technical peers, it contradicts this poll.
To me, this 175K of votes (minus the 10% who voted "pro AI") is just the result of active link shared in communities informed about AI issues, and activists (love them, btw, no problem here). Also, 175K is not a lot.
-
@davidrevoy
Is there any actual evidence that AI companies are doing that?@light Quick evidences are sort of easy to find 'on the go' while browsing sometime on viral topics; you just need to check the profiles (profile name with a random word+four or five letter hash, generated avatar or no avatars, recently joined, have only history of comments like that and no 'normal' social media interaction).
But knowing exactly how many bots profiles percents on a social media, no. Knowing who own them, no. Knowing why, no. It's very opaque... -
@davidrevoy if anything cheers me up in the “Age of Avian Intelligence(*)” it is how your comics face the issues with a smile. The fact some friends and some admired bloggers are now enthousiast about this technology is depressing.
Can’t wait to see Cepper in a cyberpunk setting
we know you lurked the genre for a long time. Just one strip, neon lights, some robots and black sunglasses! 
@jak Thank you! Oh yes, I hope I'll one day find the right angle (and time!) to express my take on the Cyberpunk genre, mixing Fantasy genre in it a bit and also SolarPunk.
I still have this directory on my disk, but so far, it's only a set of two or three illustrations, sketches, and a *.md text file I edit twice a year.
-
@davidrevoy oh now we're just on green bird slander now i see how it is david (/jk)
-
@davidrevoy I see nobody posted a 6-7 joke yet...
@poleguy Hey, I had a couple on other places; but I'm bit lost at understanding it as an (now old) Y gen; please tell me about it
-
@davidrevoy Oh, the new art style is permanent. Well it's not what I'm used to, but it certainly has its charm!
And "AI learns like humans do!" is actually true in some instances, strictly speaking. Though it would be more straightforward to describe those situations as "some people never learn."
@Linebyline Thank you. For the art style, I'm still unsure about the thick line-art and sketchy line; it felt like outlining my art with a drank truck and I wasn't happy about myself for that. But for the coloring, the little random textures, saturation inversions, and noise, I might keep that; I sort of start to like them. So, I'll keep this evolving, thank you for the feedback about it!
-
@davidrevoy use nightshade! poison your images!
@PyrexContainer Hey, first, Nightshade is proprietary and not even has a Linux release; so big red flag; and also unfortunately, it's a scam. The method was defeated a week after release, but I love the intent (not the method) and I'm impressed by the Press coverage they got, and how many non technical artist still think this "armor of paper" will protect their art of anything. There is a lot of needs of a psychological protection into this product.
-
@SmartmanApps@dotnet.social @davidrevoy@framapiaf.org @Stem@framapiaf.org But people using DDG or being on the Fediverse is already a huge bias…
@breizh @davidrevoy @Stem
"But people using DDG or being on the Fediverse is already a huge bias" - no it isn't, not to mention it wasn't a poll of DDG users anyway - I already pointed out that I don't use it - it was merely a poll they published on the web, and the link was shared by thousands of people across different platforms. I saw it here. Other people saw it elsewhere -
@breizh @davidrevoy @Stem
"But people using DDG or being on the Fediverse is already a huge bias" - no it isn't, not to mention it wasn't a poll of DDG users anyway - I already pointed out that I don't use it - it was merely a poll they published on the web, and the link was shared by thousands of people across different platforms. I saw it here. Other people saw it elsewhere@SmartmanApps @breizh @davidrevoy That doesn't undermine what David said when he montioned that the silent majority is ok with AI, and for them, convenience and instant gratification comes before ethics.
-
@Stem@framapiaf.org @SmartmanApps@dotnet.social @davidrevoy@framapiaf.org I feel like that’s sampling bias, if the same polls ran on Twitter or LinkedIn it would give different results.
my anecdotes anyways but people around me are pro-AI than anti@AmyIsCoolz @davidrevoy @Stem
"if the same polls ran on Twitter or LinkedIn it would give different results" - the DuckDuckGo poll was probably shared in both those places too.
I saw it because someone shared it here. There's no bias in it, and we know there's no bias in the Mastodon poll because it came up with the exact same results -
@breizh @SmartmanApps @Stem +1 Breizh
Smartman, don't get me wrong, I would love to believe in this poll's results. But in my local reality (at weddings, in my repair garage, at the restaurant, on week-end) when I listen to my non-technical peers, it contradicts this poll.
To me, this 175K of votes (minus the 10% who voted "pro AI") is just the result of active link shared in communities informed about AI issues, and activists (love them, btw, no problem here). Also, 175K is not a lot.
@davidrevoy @breizh @Stem
"in my local reality" - which is no longer a random sample."just the result of active link shared in communities informed about AI issues" - which you have no evidence to support
"175K is not a lot" - as I said, in a random sample you only need *25* to get an accurate result.
