463
Google Earth's new AI image generation function didn't survive a day after users deepfaked disasters (www.businessinsider.com)
submitted 3 months ago by to c/technology@lemmy.world
57 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] 210 points 3 months ago* (29 children)

What was even the point of this; what was the intended use case?

I'm serious, I have no idea who this is for. Other than to show people that their AI can badly superimpose pictures of Godzilla over maps.

  • permalink
  • fedilink
  • source
  • hideshow 29 child comments
    •
  • [–] 32 points 3 months ago (2 children)

    In large organizations when the order comes down from on high that 'we are all in on whatever' AI in this case, middle managers and product owners start looking for ways to ram that thing into their product so they can keep playing the game.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 90 points 3 months ago (4 children)

    It's easier for the news and the government to deepfake things like "huge mobs of Mexican narcoterrorists just over the border" or "there's a nuclear missile silo in Iran" or whatever with this.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 28 points 3 months ago (3 children)

    It's weird to make it public if that's the case.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 31 points 3 months ago (2 children)

    Nah, that just helps to normalize it and let people think it's just harmless and fun.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 13 points 3 months ago (8 children)

    I work in this space: you can use it to mock up commercial developments, changes in land use, etc., to see what it would look like. Somewhat useful when presenting rather dry architectural diagrams into realistic looking scenes.

    Now, why it was broadly opened to the public is unclear to me other than being a flex?

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 8 child comments
    •
  • [–] 6 points 3 months ago (7 children)

    you can use it to mock up commercial developments

    Fair, but aren't there better solutions for this already?

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 7 child comments
    •
  • [–] 7 points 3 months ago (6 children)

    Depends on how you define “better.” You can generate many variants of a scene this way, at night say, or in snowy conditions. You can mock up emergency response situations. It’s part of the story you’re telling but not a straight up replacement.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 6 child comments
    •
  • [–] 4 points 3 months ago (4 children)

    I guess by better I mostly mean "without AI"

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 2 points 3 months ago (3 children)

    Machines should work. Humans should think.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 0 points 3 months ago (2 children)

    Seems to imply people don't use Ai to shortcut their thinking and people don't think when they work.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 3 points 3 months ago (1 child)

    The quote comes from long before this age of LLMs. We need robots (which do not spy on us) to wash our dishes and fold our laundry, freeing us to think and create.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 1 point 3 months ago

    I suppose better in this case means accurate.

    Unless you just need the image to be off to the side of a PowerPoint presentation and no one's really supposed to pay any attention to it I can't really see how I could rely on an AI to produce an output that I want.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 16 points 3 months ago (6 children)

    Isn't this really the only valid reason for any AI image generation? Things I've AI image generated: 1) Godzilla riding a Unicycle B) Godzilla taking a math test #) Godzilla fixing a printer iv) Godzilla roasting a marshmallow with his breath V) My corporate logo with Godzilla attacking superimposed. None of the above is a joke.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 6 child comments
    •
  • [–] 5 points 3 months ago (3 children)

    AI is unintentionally hilarious. The biggest problem is that the humor rarely translates to others, so it's sort of just a joke for you to enjoy and then throw away.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 12 points 3 months ago (2 children)

    And we're destroying the environment for it

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 3 months ago (1 child)

    I wonder how it compares to meat consumption, personal vehicles, air conditioning, cheap manufacturing in poorer countries, fast fashion, the internet and its infrastructure...

    I'm glad to see people talking about the environment, but I have no idea how AI usage compares to other things that are destroying the environment.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 2 points 3 months ago

    In my case, I use a locally hosted stable diffusion setup, so it's not even running 99.9% of the time. Turn it on, ask for the thing, turn it off again. Fans are spinning for a minute. A single session of any AAA game is worse.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 9 points 3 months ago

    I AM FUCKING TIRED OF THIS SHIT YOU WANT THE REAL REASON WE CREATED AI? Rick and Morty needed to run for millions of years but the actors, artists and creators would all be dead. So we created AI to immortalize them so they would make Rick and Morty for a million years.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 14 points 3 months ago (4 children)

    i'll take a shot.

    1. education: these are the pyramids of giza. this is what they used to look like.

    2. real eatate: generate an image of how the development of somethjng would look

    3. business: generate a map with detaled renederings of all their locations.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 22 points 3 months ago

    I would prefer to have human generated images in all those cases, education needs to be accurate or it's not worth it, I don't want the AI interpreting what the pyramids used to look like I want a render based on archaeologists analysis. A new development will have a particular look, in order for the image to be useful it has to be an accurate representation of the building, an AI will just put something generic there.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 3 points 3 months ago* (last edited 3 months ago)

    There's also a big push internally to add AI features to their products. I don't think this was some shadowy push to discredit satellite imagery (to help Israel I guess?) - the more mundane explanation that this was a poorly conceived feature thought up by some team internally to meet business demands unfortunately rings far more likely to me.

    Edit: I want to make it clear that I'm not defending Google or the team that made this feature here. This is awful, and they should have seen it coming. But falling into conspiracies and looking for the man behind the curtain just leads to overlooking the mundane people who just follow orders or fail to consider the impact and ethics of their actions. We can't reach these people (or consider these lessons for ourselves) if we talk past them (especially if they dismiss criticism as coming from it off touch conspiratists).

  • permalink
  • fedilink
  • source
  • parent