I was reading an article in the Finnish news today talking about Statistics Finland and how they are trying to combat AI inaccuracy. What they have noticed is that in the last year, the usage of their site has dropped 17% due to people being satisfied with the AI answers they get from Google. But as you can imagine with statisticians, they were less than impressed with the accuracy of the results they tested, which is because the AI isn't pulling the statistics directly, but rather using other sources that might have taken the statistics originally from Statistics Finland or other sites, and then misinterpreted them or have the wrong numbers to begin with.
The solution they are working on for this is using the Model Context Protocol to create an interpreter for the statistics database, making it far easier for an AI to interpret the information based on the prompt. Well, in theory at this point, as it hasn't been created yet. However, this all does serve as a warning that the AI summaries that are keeping people happy and feeling they have the right information, can be pretty inaccurate. But, a searcher would only know that, if they were an expert who already had the right answer. A bit of a conundrum.
Insert smooth segue.
Smoooooth.
I don't know how long it has been there, but I noticed a "Use of AI" tab which is labelled new in the advanced settings after a refresh. Maybe it has been there for a while, and I didn't see the little green flag, but anyway, it is there now, I have seen it now, and I am writing about it now.
Essentially, it is a checkbox where it is possible to add where a posting account has used AI tools. As you can imagine, I don't use any of them except for researching, but that is pretty rare. Sometimes though, I will take a dictionary definition from the Search results or something, and that is AI generated, pulling from various dictionaries. It saves me going to the dictionary site, or typing it all out and adding the phonetics for it, like I used to do.
I am adding "research" to this post so I see what it looks like in the UI later.
What I wonder is whether people will actually use this feature to tag their posts, when it also says, "Rewards are usually meant for original, personal work". Using AI is not original, nor personal. Yet there seems to be people who believe that because they added a prompt themselves to get a result, the result is theirs. I have written many different perspectives of this nonsense, but suffice to say, using AI doesn't make it yours.
But I don't want to get into that now.
Lately, I have been thinking about how much I dislike AI images as cover photos for articles. The images themselves are fine, the quality is good and all, but I just see it as cheap and tacky. I would much rather see a grainy, crappy image that someone has taken during their day, that doesn't quite match the topic written about, than some polished AI image. I dislike it so much that I have considered making my communities "AI cover image free", but haven't done so yet. What I will start looking at is whether people are tagging their images as AI generated, because it is pretty bloody obvious that they are.
My images are all mine. Because obviously, they aren't as good as AI ones.
Again, not because they aren't good quality images, nor because they don't look real, because many do. But I think we are warping our senses by fooling ourselves into believing what we see. We aren't made to judge like this, so our preferences are going to change based on non-existent subjects. Many people don't see a problem with this, I do. There is a mental illness aspect to it, akin to the people who are willing to marry their robot vibrator with an AI assistant voice, because they are meant to be together.
Get real.
I mean that quite literally. I think many of us are spending so much time in false realities that we can no longer differentiate between what is real and what is not. We are therefore setting our expectations based on impossibilities, meaning that our real-world experience can never meet our virtual reality-based desires. A lot of people would talk about unrealistic body expectations based on the images in magazines, but at least back in the not too distant past, those people were living, breathing people, with good genes, diet and exercise regimen. Later, some photoshop too, but they were still recognisable as the person in the photo.
Until recently.
Now, there are so few real images of people being published that haven't run the gauntlet of face filters to the point that the flesh and blood person and the image posted is so far apart, they are unrecognisable. These aren't a touch up of a blemish, these are changing face shape and eye size, and lip fullness. It is pretty disgusting, if you think enough about it. And then you have people who are now so attached to the false images they have surrounded themselves with, that they have surgery to look like a filter. And now, the surgically enhanced faces all look the same.
But I digress.
Or perhaps I progress in the discussion, because while people keep using AI to enhance their articles and make their "work" far better than they can actually do, what they are also doing is resetting their expectations of what they are capable of, but they aren't actually capable at all. It isn't a skill enhancement, it is a filter over the face of skill, and soon, everyone is going to skill the same.
AI isn't going away and I am no luddite. But, like pretty much any of the technology humans have created, rather than using it for creating a better, more humane world, we use it to feed our ego in one way or another instead.
The ego doesn't exist either.
Yet it drives most of what we do.
Taraz
[ Gen1: Hive ]
Be part of the Hive discussion.
And you may be rewarded.
