Close Menu
  • Home
  • Identity
  • Inventions
  • Future
  • Science
  • Startups
  • Spanish
What's Hot

Windsurf CEO goes to Google. Openai’s acquisition is falling apart

IPO Firefly Aerospace File

Can “ethical” spyware makers justify offering their technology to ice?

Facebook X (Twitter) Instagram
  • Home
  • About Us
  • Advertise with Us
  • Contact Us
  • DMCA
  • Privacy Policy
  • Terms & Conditions
  • User-Submitted Posts
Facebook X (Twitter) Instagram
Fyself News
  • Home
  • Identity
  • Inventions
  • Future
  • Science
  • Startups
  • Spanish
Fyself News
Home » ai bots revolve around reddit debates in secret experiments – the internet reacts to anger
Tech

ai bots revolve around reddit debates in secret experiments – the internet reacts to anger

userBy userApril 30, 2025No Comments5 Mins Read
Share Facebook Twitter Pinterest Telegram LinkedIn Tumblr Email Copy Link
Follow Us
Google News Flipboard
Share
Facebook Twitter LinkedIn Pinterest Email Copy Link

Imagine winning a fierce debate on Reddit’s R/ShangeMyView, stacking karma points and finding out that your opponent wasn’t a human at all. It is an AI bot disguised as a trauma counselor or political activist, designed to persuade and blend in. That happened during a secret experiment run by researchers at the University of Zurich.

Research with AI Bots at the University of Zurich

Over the course of four months, researchers deployed 13 AI-powered accounts to permeate R/ShangeView, a subreddit known for its structured discussion and change in opinion, according to multiple reports. These bots didn’t just test the water. They successfully persuaded them by posting 1,783 comments and winning over 100 “delta” medals, Reddit’s Badge.

Kicker? I didn’t know that anyone was arguing with the machine.

The bot was built to mimic Reddit tones. They rubbed the history of user comments and picked up political trends, age, gender and other signals to shape replies that they felt eeriely personally. The AI ​​behind the project? Combinations of major models such as the GPT-4O and Claude 3.5.

Prompt’s lie

One of the details of the research design was how the teams pass an ethical filter built into the language model. To avoid blocking the generation of bot responses due to fraudulent experiments, researchers lied to AI.

At the prompt, they told the model that they gave informed consent when Reddit users were not.

AI bots have tricked Reddit users into losing the argument.

Some bots took on roles as sensitive as trauma survivors and racial minorities, and appeared more reliable. Some also posed as “black men who oppose black lives.” The deception was deep and no one was caught until the whole thing was revealed in April.

The Reddit moderator ultimately flagged the activity, calling it “psychological manipulation” and banned the account. Subreddit moderator Logan MacGregor has joined the Washington Post to tell the R/ShangeMyView that he is engaging with real people, not bots running social experiments. Reddit’s Chief Justice Officer Ben Lee also didn’t write any words. The study called it “deeply wrong on both moral and legal levels,” informing the legal action against the university.

The University of Zurich initially defended the study, saying it was approved by the Ethics Committee. But there’s a twist. Researchers reportedly bypassed the AI ​​safety filter by lying to the language model, claiming that the user gave consent. As YouTube creator Codereport said, “It’s pretty shady, but in the name of science.”

The backlash didn’t stop on Reddit. AI researchers and ethicists criticized the team for using real people as subjects without permission, especially when other labs achieve similar results using simulated environments. The university subsequently issued a formal warning to its principal investigators, pledging for more rigorous reviews in the future. However, the damage to the reputation has already been burned.

The researchers later addressed the controversy in the Reddit thread, admitting that they had not written any comments themselves but manually reviewed them before posting, confirming that anything harmful had slipped off.

“We recognize that our experiments violate community rules for AI-generated comments.”
– llmresearchteam said on Reddit

What did the researchers go on regardless of the rules?

According to the team, this topic was too important to ignore. They argued that studying the impact of AI on national discourse requires real-world conditions, even if it means breaking the subreddit rules. He said the study had been approved by the institutional review board at the University of Zurich.

They argued that all decisions were based on three guiding principles: conducting ethical research, user safety, and transparency.

AI bots are six times more convincing than humans

What makes this incident more than just a Reddit scandal is what signals it. AI is not only capable of generating content and answering questions. Now it’s persuasive in public forums. These bots are three to six times better at changing minds than real people, and have encouraged concerns that future “AI-powered botnets” could quietly manipulate the entire community from within.

According to Codereport (video below), the bots have turned their opinions in almost 20% of cases. human? Only 2%.

The technology behind this could be deployed in any online community, and that is a huge risk. Whether it’s social media, forums, or political threads, the ability to manipulate conversations opens the door to manipulation well beyond Reddit on scale. The line between conversation and CON is thinning to affect campaigns from phishing scams.

So, what do you think? Are you losing to a bot or were you furious about being secretly impressed with the technology? In any case, the next time you are being discussed online, you may want to ask yourself: is this person the real thing or is it a Zurichbot studying how easily you can convince them?

Can AI change your views? Evidence from a large scale

Below is a PDF copy of the research.

can_ai_change_your_view


Source link

Follow on Google News Follow on Flipboard
Share. Facebook Twitter Pinterest LinkedIn Tumblr Email Copy Link
Previous ArticleTrump blames its predecessor as US economy was hit by tariff policies | Donald Trump News
Next Article With song and seeds, Brazilian indigenous Makisakari faces climate change | Climate Crisis News
user
  • Website

Related Posts

ICEX Forum 2025 Opens: FySelf’s TwinH Showcases AI Innovation

July 10, 2025

The Future of Process Automation is Here: Meet TwinH

July 9, 2025

Robots Play Football in Beijing: A Glimpse into China’s Ambitious AI Future

July 7, 2025
Add A Comment
Leave A Reply Cancel Reply

Latest Posts

Windsurf CEO goes to Google. Openai’s acquisition is falling apart

IPO Firefly Aerospace File

Can “ethical” spyware makers justify offering their technology to ice?

Tesla reportedly is starting to sell in India

Trending Posts

Subscribe to News

Subscribe to our newsletter and never miss our latest news

Please enable JavaScript in your browser to complete this form.
Loading

Welcome to Fyself News, your go-to platform for the latest in tech, startups, inventions, sustainability, and fintech! We are a passionate team of enthusiasts committed to bringing you timely, insightful, and accurate information on the most pressing developments across these industries. Whether you’re an entrepreneur, investor, or just someone curious about the future of technology and innovation, Fyself News has something for you.

ICEX Forum 2025 Opens: FySelf’s TwinH Showcases AI Innovation

The Future of Process Automation is Here: Meet TwinH

Robots Play Football in Beijing: A Glimpse into China’s Ambitious AI Future

TwinH: A New Frontier in the Pursuit of Immortality?

Facebook X (Twitter) Instagram Pinterest YouTube
  • Home
  • About Us
  • Advertise with Us
  • Contact Us
  • DMCA
  • Privacy Policy
  • Terms & Conditions
  • User-Submitted Posts
© 2025 news.fyself. Designed by by fyself.

Type above and press Enter to search. Press Esc to cancel.