• 0 Posts
  • 29 Comments
Joined 11 months ago
cake
Cake day: September 27th, 2023

help-circle

  • Reading this comment gave me the idea to get Gemini’s “opinion”. Ironically, it turns out Google might’ve actually done better if they did that.

    AI conversation

    Prompt:

    I design ads for Google as a living and was charged with writing an ad for you (Gemini) to be aired during ad breaks at the Olympics .
    My current idea is as follows:
    A father wants to help his daughter write a letter to her idol, Olympic track star Sydney McLaughlin-Levrone. In a voiceover, the dad says his daughter “might even be the world’s No. 1 Sydney fan.” In the add he will ask you, Gemini, to “help my daughter write a letter telling Sydney how inspiring she is and be sure to mention that my daughter plans on breaking her world record one day. (She says sorry, not sorry.)”
    What do you think of this idea?

    Gemini output (emphasis mine):

    Evaluation of the Ad Idea
    The concept of using Gemini to help a young fan write a letter to her idol is heartwarming and relatable. It aligns well with the aspirational and inspiring tone of the Olympics.
    However, there are a few points to consider:
    Potential Issues
    Overemphasis on competition: The line “She says sorry, not sorry” might come across as overly competitive, which could potentially alienate viewers.
    Dependence on AI: While showcasing Gemini’s capabilities, it’s important to maintain a balance. Over-reliance on AI to write the letter might diminish the authenticity and emotional impact of the story.
    Target audience: Consider the target audience for the Olympics. While the story is heartwarming, it might resonate more with a family audience than with core sports fans.
    Suggestions
    Focus on inspiration: Instead of emphasizing competition, focus on the inspiring qualities of Sydney McLaughlin-Levrone and how she motivates the young girl.
    Highlight human connection: Emphasize the bond between the father and daughter and how they are using Gemini as a tool to connect with their idol.
    Visual storytelling: Use strong visuals to capture the emotion and excitement of the story. Show the girl’s admiration for Sydney, the father’s support, and the magic of Gemini helping them craft the perfect letter.
    Overall, the idea has potential, but refining the focus and execution can elevate the ad’s impact.
    Would you like to explore some alternative approaches or focus on developing the concept further?


  • It’s a bit of a weird article indeed. I also fully disagree with the author’s notion that reading a VN is faster than watching the anime equivalent. And I say this as someone who enjoys reading VNs every now and then. If I wanted to optimize the story/spent minute ratio I would just watch an anime or even better, read a manga.

    Overall I hope the article can push a few people who were on the fence to the try reading a VN to finally do so. I doubt it’ll affect anyone who was not very interested already. Likewise it also won’t do much for people who already took the plunge.

    Also I think DDLC isn’t a good starter VN at all. All the meta stuff will be lost on the reader and the time before the twist is a real slog that might easily turn people off VNs forever. I would argue the only reason it worked is because streamers had their chats pressure them into continuing and “normal people” had friends who told them the same.








  • I didn’t say that.

    I expect it to be about as awful as Starfield. However, unlike Starfield (which didn’t sell horrendously by any source I can find, just not great) it has incredible brand recognition behind it. I have no doubts it will sell based on that alone as long as it looks like Skyrim 2 at first glance.

    Edit: right after posting I figured out how to formulate what else I wanted to say but couldn’t find the correct words for: “Sadly profitability and quality don’t always correlate.”








  • My bad, I wasn’t precise enough with what I wanted to say. Of course you can confirm (with astronomically high likelihood) that a screenshot of AI Overview is genuine if you get the same result with the same prompt.

    What you can’t really do is prove the negative. If someone gets an output then replicating their prompt won’t necessarily give you the same output, for a multitude of reasons. e.g. it might take all other things Google knows about you into account, Google might have tweaked something in the last few minutes, the stochasticity of the model is leading to a different output, etc.

    Also funny you bring up image generation, where this actually works too in some cases. For example they used the same prompt with multiple different seeds and if there’s a cluster of very similar output images, you can surmise that an image looking very close to that was in the training set.