Sam says that robots make social media feel “fake”

X ACTUSUSIAST and Reddit Sam Altman on Monday: The robots made it impossible to determine whether social media publications are really written by humans, it is it to publish.

The perception came during the reading (and sharing) some posts from R/Claudecode Subreddit, which was constructed by Openai. Openai launched software programming service This takes the human symbol in May.

Recently, this Subreddit is filled with publications of self -announced programming instructions who announced that they have moved to the national constitution that a Reddit user jokingly: “Is it possible to switch to the manuscript without publishing a topic on Reddit?”

Leave this Taman wondering about the number of these jobs from real people. “I went through the strangest reading experience this: I assume that everything is fake/robots, although in this case I know that the manuscript growth is really strong and the direction here is real,” I admit at X.

Then he analyzed a region. “I think there is a set of things that happen: the real people have picked up the dribbles of LLM-Speak, the crowd was drifting very on the Internet in very related ways, and the noise cycle is” very it’s. “Robots).

To dismantle it a little, it accuses people of starting to look like LLMS, although LLMS – led by Openai – was literally invented to imitate human communication, all the way to EM’s impulse. Openai models Certainly trained on RadetAs altman was a member of the Board of Directors until 2022, and it was It was revealed as a large shareholder During the public subscription in the company last year.

It indicates that Fandoms, led by highly social media users, always tend to act in strange ways. Many groups can turn into hate tools if those frustrations that breathe for their brothers exceed.

Altman also throws excavations in incentives when social media sites and creators depend on sharing money. fair enough.

But then Altman admits that one of the reasons why he believes that supporters to appear in this subredit may be robots because Openai was also “Astroturfed”. This usually includes leaflets by the people or robots that the opponent pays, or by some third -class contractor, which gives the competitor a reasonable denial.

We have no evidence of stroturfing (although possible). But we saw how Openai subreddits Running the company After he released GPT 5.0. Instead of praise waves of believers on the new model, many angry functions have been voted. People moved to Reddit and X to complain about everything from GPT to how to burn it through credits Without completion of the tasks.

A day later RoughnessAltman Reddit verb Ask me anything A session on R/GPT in which he confessed to issues and changes to her. GPT Subreddit was not completely recovered from her previous love level, with Users are still spreading regularly Throughout their hatred of changes with GPT 5.0. Are they human? Or is it, as it seems that the trust means, somehow?

Altman holds, “The net effect is somehow Ai Twitter/AI Reddit feels very fake in a way that was not really a year or two years ago.”

If this is true, then who is the error? GPT led to good writing models, so LLMS became plague not only for social media (which has always been a robot problem) however To schooland JournalismAnd Courts.

Although we do not know the number of Reddit leaflets written by robots, or are fictional accounts by humans who use LLMS, it is likely to be a large number. Imperva Data Security Company mentioned that More than half of the internet movement in 2024 It was inhuman, due to LLMS. X private Bot Groke says: “The fine numbers are not general, but estimates 2024 indicate hundreds of millions of robots on X.”

Many defendants suggested this The lament of Taman It was his first invasion in marketing Planetary social media platform from Openai. In April, The Verge reported that such a project to be done on X and Facebook was in the early stages. This product may be or not. Altman may have hidden motives for the suggestion that social media is very fake these days.

But aside the motives, if Openai is planning a social network, what are the possibilities that will be a robot -free area? And funny enough, if that reverse and banned human beings do, it is likely that the results will not be different. LLMS is still hallucinogenic facts, but when researchers at the University of Amsterdam built a social network that consists of entirely of robots, they found that robots soon Zommers and echo rooms formed for themselvesalso.