you are viewing a single comment's thread
view the rest of the comments
[–] 38 points 13 hours ago* (last edited 12 hours ago) (19 children)

That's either AI or the most stupidly-placed sign in the world

*Edit: I agree with the sentiment, it's just in the middle of the bike lane/sidewalk

*Second Edit: as others users pointed out, sign does exist, it's just a weird photo. There's also a bike lane on the other side of the road, separated from the road by a fence and everything.

I retract my statement, it's been a journey.

  • source
  • hideshow 19 child comments
  • [–] [M] 14 points 12 hours ago* (last edited 12 hours ago) (10 children)

    Frankly, the stupid sign placement doesn't surprise me at all. I would almost be more inclined to think it was AI if it weren't in the way of non-car street users!


    Anyway, this community doesn't (currently) have an explicit rule against AI. I might still remove AI slop anyway because people don't like it on the grounds that the distraction it causes violates rule 4 or something like that, but it's got to be more obvious than this. (I have also already seen and considered VonReposti's argument about the blue sign and the dent, BTW.)

  • source
  • parent
  • hideshow 10 child comments
  • [–] 18 points 12 hours ago* (9 children)
  • [–] -1 points 11 hours ago (8 children)

    Doesn't mean the photo in OP was taken irl with a camera and not edited with AI or even straight up generated. I can use AI to make fake images of Tom Cruise. Posting a real photo of Tom Cruise doesn't mean my fake ones are real.

  • source
  • parent
  • hideshow 8 child comments
  • [–] 5 points 10 hours ago (6 children)
  • [–] -1 points 9 hours ago (4 children)

    There are tools.

    If AI could identify the tell-tale mistakes of AI it would not make them in the first place. AI detection tools are bullshit. There is either a watermark (and it's not hard to get rid of them at all) or AI detectors are just guessing... with about as much (or as little) accuracy as the AI generators creating the image.

  • source
  • parent
  • hideshow 4 child comments
  • [–] 4 points 9 hours ago

    No one should ever take anything they don't personally know to be 100% accurate, as truth.

    Everyone lies. Especially on the internet.

    But not utilizing tools that are available, even if they are flawed, is just stupid. They still give you additional data points to form a more educated perspective.

  • source
  • parent
  • [–] 3 points 9 hours ago* (2 children)

    If AI could identify the tell-tale mistakes of AI it would not make them in the first place.

    Afaik this actually isn't necessarily true, but yeah AI detectors aren't reliable and it's concerning that people are trusting the programs known for hallucinating to detect themselves.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 4 points 9 hours ago (1 child)

    Who said I was trusting the result?

    There's a difference between being skeptical, and just being a contrarian. Saying you can't trust a tool because it isn't perfect, just shows you don't understand that no tool is ever perfect. It doesn't need to be perfect, it 's providing an additional data point for you to make a determination. That's why the tool gives a confidence score.

    We know the sign exists in that location, it's on Google Maps. And I've personally seen similar phrasing on these types of signs in real life before. Given those two facts, there's little reason to assume the AI detector is wrong here, even if it's not 100% accurate. It's just one additional data point, and probably more trustworthy than a random online comment saying "it's fake, trust me bro".

    People are stupid, especially random online commenters. How many times have you seen a comment that would look like it could be accurate to most people, but you have actual knowledge in the field and know that it is 100% wrong? It happens to you all the time, you just don't actively notice it because you don't have the personal experience in every topic to see when it happens.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 8 hours ago

    You're correct, and I apologize for the misinterpretation, but the phrasing of your initial comment ("This particular image is real" while citing only the AI detector tool as a seemingly authoritative source) definitely reads as if you are trusting it.

  • source
  • parent
  • [–] 12 points 12 hours ago (4 children)

    Why’s it stupid? This is the UK. That’s not in the bike lane, it’s on the pavement. These signs are all over the UK, especially London, and they’re used to inform road users about all sorts of things. It’s as close as it can be to the road without sticking into it and people can just want around it. You’re not gonna bang your head on it.

  • source
  • parent
  • hideshow 4 child comments
  • [–] 2 points 9 hours ago

    Imagine two people in wheelchairs trying to pass each other on that sidewalk.

    It's probably "acceptable" from an engineer's perspective because the edge of the sign is right up against the ROW line (probably the edge of the sidewalk) but doesn't encroach past it and there's enough sidewalk width between the pole and the street for wheelchairs to get by one at a time, but obstructing part of the sidewalk width like that is hardly ideal. If that amount of width were considered good enough, they would've just built the whole thing that width to begin with and saved themselves a bunch of money.

  • source
  • parent
  • [–] 6 points 13 hours ago

    The pole the blue sign is on is behind the black pole, yet the blue sign is overlapping the digital sign.

    And if it isn't AI then someone gave the digital sign a serious dent on the top.

  • source
  • parent