We already have search results providing useless ChatGPT results. I bake my own bread for my son due to certain food allergies. We have some leftover almond f

We already have search results providing useless ChatGPT results. I bake my own bread for my son due to certain food allergies. We have some leftover almond flour, so I googled up some bread recipes using almond flour this morning. Web searching in 2023 is almost impossible for food now. My results were the shovelware of autogenerated content farm recipes. One example I found was a quickbread (In my country that's a baking powder based cake, sometimes with yeast as a flavoring rather than leavening agent, being marketed as a bread). This is a legit looking recipe and would probably work as an almond-flour-flavored quickbread. However, it was provided with ChatGPT flavor text surrounding it. Statistically, most bread recipes have instructions and answer questions about yeast and yeast problems. Therefore ChatGPT decided this recipe also needs an extensive wrapper of yeast-issue flavor text. Of course any actual sentient being that's not the output of mere statistical randomness would have enough cognitive function to instantly identify this recipe as a yeastless quickbread and would not include the flavor text about yeast issues. In a similar manner, ChatGPT "knows" that bread recipes always start with preheating the oven. However, this clickbait recipe was ostensibly for a bread machine. So the recipe and flavor text was a bizarre randomized non-cognitive mixture of oven vs bread machine instructions. "Preheat your bread machine to 375" and so forth. I've had similar experiences with ChatGPT generated "technical" content. ChatGPT is great at imitating hot take formulaic blogpost clickbait type content, but with zero cognitive ability backing it up the content tends toward meaningless word salad. I think as per the Sokal Affair from the last century, some people get VERY mad when its pointed out their favorite, most emotionally triggering content sources have all the cognitive intellectual basis of a random number generator. The sad part, is literacy is so weak in most countries that ChatGPT can "outwrite" a large segment of the mostly illiterate population. I find it pretty easy to identify autogenerated filler content, both online and social media, because there tends to be zero cognitive function behind the well written article. Its like the article has a three dozen well written and stylish lines from Pulitzer Prize winning articles, but because they were selected based on randomness plus being found nearby each other, the higher level cognitive functioning is completely absent, which is jarring when juxtaposed with well written individual sentences and phrases. ChatGPT is, essentially, a pollution source, and should be regulated and limited as such. I never would have guessed what kills the internet is "Super Eliza" flooding out real content with infinite random filler garbage content.