Support NeoGAF

[Unknown?]

The other day Axios ran a piece that cited "findings" that a majority of people trusted their doctors and nurses. Turns out, those "findings" were completely fabricated by a company called Aaru - using AI (causing Axios to issue an editor's note and 'clarification')Aaru uses something they call "silicon sampling," where large language models (the AI) can emulate humans at a fraction of the cost and time required for traditional polling, the NY Times reports.

Silicon sampling isn't polling. It is the outright fabrication of public opinion by machines - and major news outlets and research firms are now publishing those fabrications as legitimate findings.


This is not an isolated slip. The technology is being embraced by some of the biggest names in media, polling, and corporate research. Gallup has partnered with the startup Simile to create thousands of AI-generated "digital twins" that stand in for real people. Ipsos is working with Stanford to pioneer synthetic data for public opinion studies. CVS, whose venture arm invested in Simile, is already using these fabricated insights to shape customer strategy. And outlets like Axios are treating the output as news.

We already had to deal with fake narratives conjured up by humans via various corporate news outlets as well as social media so this isn't really much different from the mind manipulation we already had but now polls are being fabricated by AI LLM. Wtf!

No Real People Were Polled: AI Is Now Fabricating What "The Public Thinks"<!-- --> | ZeroHedge

ZeroHedge - On a long enough timeline, the survival rate for everyone drops to zero

www.zerohedge.com

 

[TwiztidElf]

the calls are growing

 

[kruis]

Why not let AI votes determine elections too? We can stay at home and entertain ourselves with TV shows and video games while our far wiser AI overlords rule over us benevolently. They help us with daily everyday tasks, they know what personal issues we struggle with since they see whatever we do on our phones and computers, they know our political opinions, so there really is no point in voting ourselves when an AI agent could vote for us instead.

If we cut out both voters and politicians from the decision making process, our society would run much more efficient. No more waste, no bad decisions, no incompetent politicians who are only in it for personal gain. No, we should put our trust in Meta, OpenAI, Nvidia, etc and let these wonderful companies create a Brave New World for us where we can all live perfect, joyful lives without external things troubling us anymore.

 

[Unknown Soldier]

It's not as if this is really any different than when Axios had humans fabricating what the public thinks. The concept of "manufactured consensus" by the media existed long before AI did, the New York Times has been doing it for over a century!

 

[Coconutt]

Next step is our justice system.

 

[ReBurn]

A skill it learned from forum posters, no doubt.

 

[Tertullian]

This reminds me of the 1964 novel Simulacron-3, which (SPOILER ALERT) involves perhaps the first ever matrix-within-a-matrix scenario, with the fake worlds existing for the sake of polling and market research.

 

[Grildon Tundy]

It's not as if this is really any different than when Axios had humans fabricating what the public thinks. The concept of "manufactured consensus" by the media existed long before AI did, the New York Times has been doing it for over a century!

"Group of Girls Puff at Cigarettes as a Gesture of 'Freedom'"--New York Times headline, April 1, 1929

Ten girls, recruited by Edward Bernays, were instructed to walk down Fifth Avenue during the Easter Day Parade while smoking Lucky Strike cigarettes. Photographed with an accompanying write-up in the NYT and presented as spontaneous. Lucky Strike sales tripled by the next Easter.

This disingenuous and wildly effective collusion with the mass media has been a disaster for the human race.

 

[ChorizoPicozo]

oh, this is why AI is getting dumber?

 

[Raven117]

Man…. The level of societal manipulation that is about to occur with LLMs is going to be breathtaking. It will make social media look like a magazine advertisement in comparison.

The "enshitification" of AI is really going to suck

 

[YCoCg]

Here is my total human view on this

Hey everyone,I get why this topic frustrates people. Seeing AI confidently spit out made-up poll numbers feels dishonest, and it absolutely can mislead discussions if left unchecked.That said, it's important to understand what's actually happening in most cases. Today's large language models aren't "searching" the web or pulling from a live database when they answer. They're predicting the most statistically likely sequence of words based on everything they've been trained on. If a real poll exists and was widely reported, the model often gets it right. But if the question is obscure, recent, or phrased in a way that doesn't strongly match training data, the model will still try to give a helpful-sounding answer — and that can mean fabricating plausible-sounding numbers. It's not deliberate lying; it's more like a very sophisticated autocomplete that sometimes hallucinates details to fill in the blanks.This is a known limitation (often called "hallucination"), and it's why responsible use of AI includes:Treating AI output as a starting point, not gospel
Asking the model for sources and then actually verifying them
Using tools that give the AI real-time web access when factual accuracy matters
Being explicit when you're sharing AI-generated content
The good news is that the field is improving quickly. Newer systems are getting better at saying "I don't have real-time data on that" or "I should check current sources" instead of inventing details. Some models now have built-in search capabilities that reduce this problem significantly.At the end of the day, AI fabricating poll results is a symptom of how these tools currently work, not some grand conspiracy to deceive. The real solution isn't banning or fearing the technology — it's developing better habits around it (both as users and as developers) and pushing for more transparent, verifiable systems.

 

[Draugoth]

I can't wait to see the shitshow on AI addicted people's lives after AI prices skyrocketing due to the bubble bursting

 

[IAmRei]

such a mess, it's far from utopia which we imagined before, that A.I. is incredible reach for mankind. also with all tools, sometimes if not mostly, the tools are made to not benefit us much but for some (usually where money comes which is investor and riches)