The use of artificial intelligence for writing texts is a hot topic. Of course, ChatGPT takes a large part of the main role, but we must not forget that its developer, OpenAI, has been offering GPT-3 for some time. And as ChatGPT answered a question about the difference between the two services, «GPT-3 is a much larger and more advanced language model that uses a deep transformer architecture and extensive pre-training to generate high-quality text.«.

In other words, in a much more human way, at least in theory, the texts generated by GPT-3 can be used in any context, including professional onesand that’s something that those of us who do that typewriting thing because it leaves us with a little bit of a fly in the ear, because the possibility that some day an artificial intelligence can write the same or better than itself is a possibility that despite our regrets, they will emerge on the horizon.

However, it seems that the technology is not yet as advanced as some thought. And yes, when I say they thought I really mean “they tried to sneak in without warning.” And so it is, as Futurism posted earlier this month, CNET has begun publishing articles written by artificial intelligence, probably CPT-3. It was only when it became public that the outlet began identifying such articles as such and announced that it was suspending acceptance of the technology. Yes, a pause that, in addition to a change of plans, suggests that this is a temporary measure, because the truth is that they plan to restore it in the future.

However, nothing seems to have come of the “experiment” (yes, when I say experiment, I really mean “the covert use of artificial intelligence, passing off those texts as original and written by a human being, hence the quotation marks). “So, as we can read in The Verge, more than half of the AI-written articles published by CNET contained errors. So CNET had to make corrections in 41 of the 77 publications written by the artificial intelligence solution that the outlet uses.

From inaccurate information to “non-original” phrases, which is a very nice and very sympathetic way of referring to plagiarism of texts used to train AI, articles written by AI have received as much attention as CNET would have liked. but not for these reasons they would have wanted to. With an important addition: if they had mentioned the origin of these texts from the beginning, the reaction would have been much less critical of the medium. However, at least for now, the human seems to dislike the medium that tries to impersonate the AI ​​that optimizes the text for the search engines. Thank you so much, humanity, for the part that touches me.

And while there are real “machines” on the MuyComputer team that I’m lucky enough to share space with, I guarantee they’re human… and any AI would love to write like them, I assure you.