Recently, some people here have brought up what has come to be known as 'Dead Internet Theory'. This term denotes the idea that most internet users are in fact bots.
There have been numerous findings that lend support to such an idea. For example, a brief report by US tech company Barracuda Networks, released in September 2021, claims: 'Bots make up nearly two-thirds of internet traffic'.
https://web.archive.org/web/20211203013241/https://assets.barracuda.com/assets/docs/dms/Bot_Attacks_report_vol1_EN.pdf
Bifurcating bots into 'good' and 'bad', this report further claims that only 25% of 'automated traffic' comes from 'known good bots', whereas 40% comes from 'known bad bots'.
The report provides five examples of real-life 'bad bots'. However, none of these bots were attempting to emulate social media users.
*Reddit* is one website, however, where such bots can be found. Indeed, there is actually at least one whole subreddit full of them:
Enter 'Subreddit Simulator using GPT-2', a subreddit in which *not a single user* appears to be human.
https://www.reddit.com/r/SubSimulatorGPT2/new
Whilst many of their conversations appear to be nonsensical, one might surmise that advances in AI might lead these bots to one day have conversations that appear more real than those being had between human users. Furthermore, that less intelligent and/or tech-savvy users, likely unaware of bots—let alone capable of telling the difference between bots and humans—might already be fooled by this subreddit as it is.
A question for discussion: Could it be possible that bots will one day be used by the forces of globohomo to create, for example, false impressions that their beliefs, candidates, parties, etc. are more popular than they really are? That opposing beliefs, candidates, parties, etc. are less popular than they really are? Could it be possible that dissident groups could do the same? Could it be possible, in the not-too-distant future, that the opinions of internet users will become completely unrepresentative of those of humans? If so, and combining this with an exceedingly atomized world, could there really be a 'post-truth' world? One in which late-modern phenomena such as circular reporting and filter bubbles, paired with advanced AI, create an internet that *appears* incredibly active and informative, and yet is composed only of more advanced versions of the literal gobbledygook and meaningless nonsense that goes on in this subreddit's threads?
Full of 'history' that, in fact, *never happened*:
https://www.reddit.com/r/SubSimulatorGPT2/comments/utevfv/why_did_north_korea_go_from_being_an_isolationist/
Full of seemingly informative 'science' that is, in fact, pure nonsense:
https://www.reddit.com/r/SubSimulatorGPT2/comments/uttimy/how_can_a_nuclear_reactor_take_so_much_energy_to/
Full of discussions about movies that, in fact, *aren't even real*...
https://www.reddit.com/r/SubSimulatorGPT2/comments/uoip3j/the_game_the_game_2004_a_film_about_the_game_of/
... and if they are, by 'people' who *never watched*—and know nothing about—them:
https://www.reddit.com/r/SubSimulatorGPT2/comments/usks1b/theory_about_the_ending_of_the_original_star_wars/
Crossposted to: https://saidit.net/s/debatealtright/comments/9aaf/a_literal_bot_subreddit/
Also crossposted to: https://scored.co/c/DebateAltRight/p/15HvL8gpjr/a-literal-bot-subreddit/c
