jump to:all–mastodon–bluesky–lemmy–lobsters–dev–stack–wiki–rss–hn
wwebbit
view: | | |
mastodon

Mastodon discussion

19 public replies · showing 19 · fetched 0s ago · read only

  1. @futurebird@sauropods.win · 9 points · 2h ago

    Write it in sidewalk chalk ... it'll find its mark.

  2. @futurebird@sauropods.win · 1 points · 1h ago

    Maybe we could crowd fund a billboard.

  3. @futurebird@sauropods.win · 53 points · 1h ago

    If this is you? Just stop. Stop today and everyone will forget about it in a few weeks. That's the solution. Just stop using AI to make the images or "improve" the writing or to design the flyer or whatever. Please don't force your friends and coworkers to confront you about it. The other option is to clearly label all such work as "made with AI" ... If you don't want to do that? Let us ponder together *why* that is. What does reluctance imply?

  4. @llewelly@sauropods.win · 2 points · 1h ago

    @ futurebird leftist messaging strategy: "we're hoping a sneaky teenager with a can of spraypaint will do our job" nazi messaging: "ah, we literally spent many decades and billions of $$ buying nearly every billboard in the USA "

  5. @futurebird@sauropods.win · 2 points · 1h ago

    @ llewelly Well if you aren't really interested in controlling people expressing dismay at personal choices that degrade social well-being and reduce joy is always awkward. To propagandize itself can feel "unleftist" But sometimes it's just repeating the truth over and over until it seeps in.

  6. @Nicovel0@mastodon.social · 3 points · 1h ago

    @ futurebird god yes, where I work pretty much everyone in marketing and above VP level. Always illustrating their posts with AI slop when stock images are easily available (we have a subscription) or an illustration is not necessary.

  7. @clayfoot@mastodon.social · 3 points · 1h ago

    @ futurebird I'm reminded of the ad agency that switched to AI photos, then switched back to real photos. There was nothing obviously wrong with the AI photos, but engagement went way down anyway. Using AI somehow made the results less interesting.

  8. @JugglingWithEggs@mstdn.social · 1h ago

    @ clayfoot @ futurebird I think this applies to everything now…but I’ve particularly noticed it a lot with estate agents recently. They manipulate the images of the house with AI and get AI to write the particulars. It’s really very obvious and I don’t think potential buyers trust it at all. Trust is already at minimal levels with estate agents - why reduce this?

  9. @futurebird@sauropods.win · 1h ago

    @ JugglingWithEggs @ clayfoot "everyone else is doing it" And sometimes it seems like it works? I don't understand having photos of a property that don't show how it will really look when you show up. That is just wasting everyone's time. And there is a gradient to this. Did the agent use photoshop to remove some junk from a room to show how it would look empty? Did they use a wide angle lens to make it seem bigger? Did they improve the contrast? But it's hard to see what's too much.

  10. @JugglingWithEggs@mstdn.social · 1 points · 52m ago

    @ futurebird @ clayfoot It’s been the case for many years they have used photoshop to ‘tidy up’ cluttered homes and taken the photos with a wide angled lens to make everything look twice as wide as it actually is. They argue if they don’t do this - and everyone else is doing it - people will just swipe straight past the property. I don’t know?! But using AI to write the particulars results in generic flowery verbiage…nothing stands out!

  11. @mansr@society.oftrolls.com · 25m ago

    @ JugglingWithEggs @ futurebird @ clayfoot Photographing interiors is hard, and the wide-angle distortion is inevitable if you're going to get much of a room in a single shot. Some light editing might be justified too, such as to remove something overly personal. I do expect, though, that a house will look substantially the same when viewed in person. If not, I'll assume nothing they say can be trusted.

  12. @carstenfranke@mastodon.social · 29 points · 1h ago

    @ futurebird I am in Germany right now and noticed that commercial ads and flyers do have a little label "made with AI". Seems to be a new EU law. Very helpful in deciding which stores to avoid.

  13. @futurebird@sauropods.win · 1h ago

    @ carstenfranke I want this law in the US. Please can we have this?

  14. @carstenfranke@mastodon.social · 1h ago

    @ futurebird no, we can't have all the nice European things, like universal healthcare, almost free education or lower poverty or marked AI content.... Because... FREEDOM!!! /s

  15. @cpsask@techhub.social · 1h ago

    @ carstenfranke @ futurebird I would also like this in Canada please

  16. @5tern1@corteximplant.com · 32m ago

    @ carstenfranke @ futurebird the mark is really nice, I ask myself multiple times a day now "Why would you ai generate *that*?"

  17. @paulc@mstdn.social · 1 points · 35m ago

    @ futurebird a coworker told me that she had to take a letter written by another superior coworker that was obvious AI generated. She said she removed half the words and corrected the rest for style and content. And sighed.

  18. @atlovato@mastodon.social · 25m ago

    @ futurebird - Understandable - I like it.

  19. @gbsills@vivaldi.net · 2 points · 1h ago

    @ futurebird One of the things I find annoying (hey, I'm almost 70, I could go on and on about the things that annoy me) is that this stupid fantabulist LLM junk is giving AI a bad name. Personally, I love the spam filter that comes with Proton Mail. Radiologists can do great things with the tool built with old school machine learning and knowledge engineering. The problem with LLMs is that they patch together some really nice AI work like natural language recognition, with a gigantic Google search engine. The tech bros are selling this "If we just scrape enough stuff, intelligence will 'emerge'". When you use that to write or draw, you are just copy and pasting from the Internet. Maybe Trump is almost right by accident. It should be "SI" for "Stupid Intelligence".