Don't worry, AI will tell us the truth

In a recent interaction with an acquaintance, he disclosed the disturbing news that the neighbours had ‘gone down the rabbit hole’ of conspiracy theories. He then went on to explain how the stupid neighbours believe that the UK is being overrun by immigrants, and is in trouble, and that the world is being manipulated by an elite set of psychopaths!

“Right…”, I said slowly, not knowing where to go with this.

“But there is trouble in the UK… didn’t you see all the riots and stabbings?” I offered.

A blank and ever so slightly confused face stared back at me.

It was then that I realised just how big the gap is between those who are unwittingly programmed by the establishment and those who do their own research beyond mainstream media (AKA ‘conspiracy theorists’).

Similarly, a family member who’s always, as far as I can remember, backed the Republican Party, may not have loved Trump but acknowledged he was the man for the hour, has suddenly jumped ship and hates Trump with a passion and thinks Harris is the woman for the job! Even sending me ridiculous videos suggesting Trump is a tottering old fool, completely incompetent, who needs to be locked up – with seemingly no recollection whatsoever that I actually back Trump (if pushed to back any of them).

It's the Twilight Zone my friends!

Popular Science, in a post on September 12, highlighted a study in the journal Science called “Durably reducing conspiracy beliefs through dialogues with AI”. Both the Popular Science article and the research article in question are very telling and maybe join some dots regarding my recent interactions of people who seem to be in the Twilight Zone.

Here’s how Mack Degeurin, writing in Popular Science, opens his piece:

In 2024, online conspiracy theories can feel almost impossible to avoid. Podcasters, prominent public figures, and leading political figures have breathed oxygen into once fringe ideas of collusion and deception. People are listening. Nationwide, nearly half of adults surveyed by the polling firm YouGov said they believe there is a secret group of people that control world events. Nearly a third (29%) believe voting machines were manipulated to alter votes in the 2020 presidential election. A surprising amount of Americans think the Earth is flat. Anyone who’s spent time trying to refute those claims to a true believer knows how challenging of a task that can be. But what if a Chat GPT-like large language model could do some of that headache-inducing heavy lifting? 

The ’once fringe ideas’ is a good set up to couch a belief in a ruling class controlling world events and even the manipulation of voting machines as crazy ‘conspiracy theories’. Throwing in the Flat Earth theory is a transparent propaganda strategy to put reasonable theories and completely implausible ones in the same basket. The stupid reader (they seem to be proliferating) will assume that a ‘world controlling elite’ is as crazy as the idea of a flat earth. The editor’s summary of the Science paper adds his bit by saying that “Beliefs in conspiracies that a US election was stolen incited an attempted insurrection on 6 January 2021. Another conspiracy alleging that Germany’s COVID-19 restrictions were motivated by nefarious intentions…”

Here's the intro to the paper:

Widespread belief in unsubstantiated conspiracy theories is a major source of public concern and a focus of scholarly research. Despite often being quite implausible, many such conspiracies are widely believed. Prominent psychological theories propose that many people want to adopt conspiracy theories (to satisfy underlying psychic “needs” or motivations), and thus, believers cannot be convinced to abandon these unfounded and implausible beliefs using facts and counterevidence. Here, we question this conventional wisdom and ask whether it may be possible to talk people out of the conspiratorial “rabbit hole” with sufficiently compelling evidence.

So, if you don’t believe what the government is telling you then you are trying ‘to satisfy an underlying psychic need or motivation’ – nothing to do with grasping the truth. And what are we to do with the growing number of these crazy people (thanks to the ‘podcasters, prominent public figures, and leading political figures’ - how sad they missed the Substack crew!) who won’t abandon their unfounded and implausible beliefs? Well, use AI that knows all the facts and counterevidence to slam-dunk all the misinformation, disinformation, malinformation (and just information) of course!

And how can we trust that the AI isn’t going to lead us up the garden path? Well, the researchers at the Massachusetts Institute of Technology, Cornell, and American University (so they must be beyond reproach!) used a chatbot (obviously impervious to any false premises) to change people’s minds about ‘conspiracy theories’.  According to one of the researchers, David Rand, “We see that the AI overwhelmingly was providing non-con conspiratorial explanations for these seemingly conspiratorial events and encouraging people to engage in critical thinking and providing counter-evidence. This is really exciting,” he added. “It seemed like it worked and it worked quite broadly.”  

Here's a summary of the results of the study:

The treatment reduced participants’ belief in their chosen conspiracy theory by 20% on average. This effect persisted undiminished for at least 2 months; was consistently observed across a wide range of conspiracy theories, from classic conspiracies involving the assassination of John F. Kennedy, aliens, and the illuminati, to those pertaining to topical events such as COVID-19 and the 2020 US presidential election; and occurred even for participants whose conspiracy beliefs were deeply entrenched and important to their identities. Notably, the AI did not reduce belief in true conspiracies. Furthermore, when a professional fact-checker evaluated a sample of 128 claims made by the AI, 99.2% were true, 0.8% were misleading, and none were false. The debunking also spilled over to reduce beliefs in unrelated conspiracies, indicating a general decrease in conspiratorial worldview, and increased intentions to rebut other conspiracy believers.

One has to wonder what the AI ‘thought’ were true conspiracies!

Now you have to understand what data this sort of AI chatbot is trained on. Has it been trained on all of the sanctioned information ‘out there’, including all the ‘fact checkers’ ‘facts’? Yes, of course it has. If the starting assumption is that any anti-COVID-19 narrative is a conspiracy theory, then of course the AI is going to come up with any and every counter-argument (posing as ‘critical thinking’) to convince the experimental group they are as crazy as people believing the earth is flat.

And here’s the rub - the whole planet is that experimental group now.  Don’t trust the government or Big Pharma, or Big Food, or Big Anything? Did your super fit cycling mate suddenly die after his second jab and you wondered if it could have been the ‘vaccine’? Are you questioning your local council as to how excluding you from the local forest is somehow saving the planet? Then you are trying to fulfill a psychological need by adopting conspiracy theories and if you don’t stop it you’ll not only kill your grandmother but you’ll be responsible for the planet burning up! No wonder these AI wonderkids are working so hard to stop your conspiratorial fantasies – you’re damn dangerous and you need to be stopped.

But fear not. As long as we are plugged into the stream of AI generated propaganda truth telling, then Big Brother reason will win at the end of the day and you will see that two plus two did equal five all along.