AI Chatbots: A nightmare of epic proportions

I came across a Twitter ad for an "AI Chatbot" app that's disturbing: it's been downloaded 10M+ times already. These apps represent a new, dangerous covert vector for influence OPs and mind control.

I’ve been wanting to write this post about AI chatbots for some time, and today is the day. Today’s post is not directly about ChatGPT, which you may have recently heard about, but about something else: something alarming that we need to become aware of, be vigilant about.

In case you have been out of the loop, a quick overview: ChatGPT is a new AI model that is making significant strides forward—it is appearing to demonstrate signs of rudimentary “intelligence”, at least in the eyes of some observers.

The recent chatter about ChatGPT makes today’s topic even more immediately relevant and timely, which is what prompted me to finish writing it.

I’ll write more about what ChatGPT is really doing (and what it isn’t) in another future post; but for now, at least be aware of this: recent news headlines claim that the AI that underlies ChatGPT can potentially pass (with mediocre marks; but still pass!) licensing exams for Medical Boards, the Bar Exam, and MBA exams.

AI’s like these are also already able to eat the lunch of marketing copywriters — right now. 2/6/2023 Update: here’s more.

That’s right: “robots” are coming for certain jobs: copywriters, doctors, lawyers, and accountants. Certainly not the jobs we once expected to fall first to automation…but I digress.

While these claims about ChatGPT are a bit sensational and extraordinary, and may not yet be entirely accurate, it does at least serve to illustrate that Artificial Intelligence systems are making exponential strides forward every day.

Human beings are particularly bad at grasping non-linear and exponential growth trends, which is why the coming progress will appear to be shockingly faster than we ever imagined possible.

As an example of how progress in AI accelerates: when voice recognition systems from Google and Microsoft first came onto the public scene about a decade ago, I followed the pace of “learning improvement” by looking at some of the published data from Google.

I was stunned at how fast these systems improved in recognition ability and quality in just the first four months of widespread use.

In case you don’t know how the “training” system works, here’s one insight: if you use “Hey Siri” or “Hey Google” and ask it a question, and the voice-to-text translation is wrong or you get the wrong answer, what do you do?

You either speak your request again, using clearer diction, or you bring up the keyboard (usually in frustration) and correct the words by re-typing them, or you modify the search manually to get better results.

But when you do that, the AI system is designed to notice how you corrected the results and learns from your doing so.

Then the next time you or someone else asks, using voice, for the same thing—the results will be that much better. It has learned from its mistakes simply by watching how you manually corrected it.

In other cases, snippets of voice interactions are sent back to the programmers, and they “tweak the algorithm” by listening to the audio clip themselves and adjust recognition algorithms to improve their accuracy.

Sometimes these systems send audio recordings of your household that you aren’t even aware were being listened to by engineers in some lab in Silicon Valley. Recently, photos were also taken by some roomba robots for “training models” and inappropriately distributed online.

Soon, one of the AI systems that is modelled after ChatGPT may be able to pass the so-called “Turing Test.” In May of 2022, I wrote about the Turing Test and a surprising tie-in that might exist between a new form of the Turing Test that I envisioned—and how that connects to a phenomenon of human consciousness in this substack post.

If you’re into deep thoughts, that post above is one to dive into. Fully exploring that post and all of its implications could take a few days, however: be warned!

I decided to split the topics that I was going to write about today into two different posts, since they expose two different things that are now appearing in various online arenas.

But first, by popular request from readers: a photo from my collection for your enjoyment!

A new dawn is coming with the advancements in AI. Is humanity ready?

( Here’s the song that I was listening to as I captured the above photograph. )

The first topic—which is the main subject of this post—has to do with AI Chatbots (but not the ChatGPT kind.)

The second has to do with a new class of smartphone game that entices players into taking photos of themselves to create digital characters.

Both are highly problematic developments, and hardly anyone stops to really think about why.

Let’s start with a little history lesson before pressing forward with the AI Chatbot topic.

As I have written about before, Edward Bernays is known as the father of advertising, public relations, and propaganda. His work influenced Goebbels, Hitler’s Nazi propagandist.

Born in 1891, Bernays was the nephew of the famous psychologist Sigmund Freud. In a book he published in 1928, long before there was the faintest inkling of computers and the modern “Internet”, Bernays wrote this:

“The conscious and intelligent manipulation of the organized habits and opinions of the masses is an important element in democratic society. Those who manipulate this unseen mechanism of society constitute an invisible government which is the true ruling power of our country. “

—Edward Bernays, father of modern advertising, public relations, and propaganda, in a book published in 1928.

Bernays was talking, back then, about the power of mass media like newspapers, radio and print posters, but his ideas find purchase today in the world of (anti-)social media, which I wrote about here on substack: part 1 and part 2. These posts are very much worth reading, if you haven’t yet.

What I have written about for years, starting in this WordPress post titled (Anti-) Social Media: Pandora’s Box is Opened, is how modern social media has been weaponized almost since the beginning for mass influence operations and PSYOPs worldwide by covert actors; the evidence for which was exposed by Edward Snowden, and which is now being further exposed in the Twitter Files thanks to Elon Musk.

The issue at hand with (anti-)social media platforms has to do with what I termed Memetic Injection. Bernays would know it as “propaganda”, but I coined the phrase Memetic Injection because of the widespread modern use of “memes” to inject ideas into the public consciousness.

I wrote about Memetic Injection in this substack post. We now understand all of these influence operations techniques to be part of 5th Generation Warfare.

Social media platforms have become covert battlegrounds for Psychological Warfare operators from many different sectors (government, private business, and NGOs) and many countries around the world.

The idea of Memetic Injection—the original introduction of novel, behavior-shaping ideas into the public consciousness in order to influence and control the mass population—was once the exclusive realm of the priestly class.

That’s no longer the case. It’s now in the realm of advertisers and the controllers of mass media and (anti-)social media. Who I guess, see themselves as the new priestly class.

With the various forms of media now utilized in mass distribution (text, music, images, video), these tools for viral memetic injection can permit covert operators to quickly induce so-called Mass Formation Psychoses, sometimes via an almost subliminal, hypnosis-like effect.

Or maybe it should be called Hypgnosis, or Hipgnosis. Right, Justin Bieber? Good time to sell out, my friend, now that you’re vaccine injured. Interesting name of a record label to sell to, though, don’t you agree. They don’t really attempt to hide what they’re up to at all…

“Superstar Justin Bieber sold his music rights Hipgnosis Songs Capital, the company announced Tuesday, in a deal reported to be worth over $200 million, making him the latest musician to cash in on his catalog.”

Edward Bernays would be stunned at how effective the tools available to the modern crowd-steering elites have now become. He would have instantly recognized the effects of the COVID-era propaganda operations for what they were: attempts at mass conformity and control.

As the Twitter Files continue to reveal, we can now see some of the “hidden hands” at work behind the scenes that were scheming to “boost” or “deboost” — or outright censor — narratives and voices that ran counter to the plans of the power elite. We can see the hidden hands, exposed, trying to exert control over the masses.

But here’s the thing: to people like me who have been studying these things for decades, it isn’t at all surprising what is now being revealed. We are saying “finally” just when the broader public is only now saying “What? OMG!”

Many of us have been watching it happen, knew exactly what it was, and spoke up loudly the whole time to draw attention to it. Now the rest of the world is being exposed to what we long knew to be happening.

I’ve always known that the covert operators were running operations of all types on platforms like Twitter to shape perceptions, attitudes, and opinions—to steer the body politic in the directions the elites desired, whether for reasons of power or profit or both. Bernays’ work from 1928 has long been studied by those wishing to control the masses.

With all of the bots, trolls, and “algorithms”, however, at least one thing was true: the influence operations could at least be seen with those who were trained to look.

Because the messages were on the surface, even if the operators’ identities and motives were hidden.

That era of visibility, however, is now coming to an end, just as we started to awaken to the rigged game that was being played against us.

That’s where the rest of this post is headed.

Let me take a small detour before I continue. Years ago, I watched a YouTube video about Darren Brown, a famous “mentalist” and entertainer; it was called Toy Story: How to Control a Nation.

In this video, Darren shows a subliminal trick he used to achieve the result he intended: the fact of subconscious UV light perception.

It is an interesting claim that our brains can “see” something that is beyond our conscious visual perception if it is written with UV-light visible “disappearing ink.” That is, we can be subconsciously influenced by writing that we can’t even “see”.

Herein lies the problem with the new “AI Chatbots”—and the problem I’ll describe just begins to scratch the surface of what is coming. They, too, are “invisible ink.”

Unlike the 5th Generation Warfare influence OPS on Twitter, Facebook and the like—in which you can spot the messaging and attempts to influence if you are trained in how and where to look for it—the world of “direct messaging” and private interactions with “AI Chatbots” introduces a new complication.

Perhaps we should call this one “6th Generation Warfare”.

Now the channel of psychological influence is hidden from the public, submerged in the twilight zone of DMs—and it is cleverly hidden inside “Health and Fitness” apps sometimes targeted at lonely people (the ones who are more vulnerable to suggestibility in the first place.)

So, what happens when the same PSYOPs that have been conducted for the last decade out in the (relative) open on Twitter and Facebook are now executed in private chat channels or DMs—operations that are invisible to all but the endpoint targets?

Furthermore, what if the “entity” that is conducting the influence operation is a highly-skilled chatBot that “studies” its interactions with its “victims”, learns from its failures to engage, and then figures out how to gain “confidence” with influenceable people—so that it can ultimately engage in even more effective memetic injection?

Now the problem of PSYOPs on “social media” takes on a whole new dimension, doesn’t it? Just like with “invisible ink”, we won’t be able to see the influence anymore, other than later seeing its effects begin to manifest in the world, with no idea where the origin of influence was.

Unless, of course, the public is allowed to mine the databases of the major social media platforms to look for the “points of origin” of ops in order to uncover the signatures of these kinds of operations. Please Elon: make it so.

Now take this idea of “subsurface messaging” further, to online gaming—which also have “private” chat channels, and then further to the “Metaverse”, which is the 3D photorealistic “virtual world” that was first proposed in the book Snow Crash by Neal Stephenson back in 1992.

In the vision of the MetaVerse that Mark Zuckerberg has presented for Facebook’s future direction, their platform will create entire photo-realistic virtual 3D worlds that you’ll “see” using 3D headsets that they produce and sell to you; and they’ll populate this “world” with a mixture of human and non-human “characters” moving and talking inside an artificially generated “reality”.

Imagine the possibilities for psychological shaping and influence if you can control everything that someone sees and hears in a completely artificially generated “reality”.

Imagine what happens if those “characters” that people interact with are trained ChatGPT like systems, designed to ferret out their individual human personalities and weaknesses and then exploit these to shape their “offline” behavior.

Terrifying, isn’t it.

The development of the MetaVerse will also lead to entirely new problems humanity has never faced—besides reality-distorting influence operations, think of the other problems it will introduce. What if your “avatar” commits a “crime” in virtual reality—was it really a crime? How can you even know who you are really interacting with in a 3D simulated “reality”?

The world is about to get a lot more complicated.

“The price of liberty is eternal vigilance.”

(Thomas Jefferson, whether in spirit or actual quotation is in dispute.)

---

Personal note:

To all of those who thoughtfully responded to my recent posts A New Year’s Miracle and Thank You to all the Angels, I’m forever grateful. Your kindness touched many people and the positive effects continue to ripple forward.

My dream of being able to return to a life of writing and research full time gets that much closer for every new subscriber who converts to paid. I can share with you that I’m now able to relocate to Michigan in the next two months to unite with my fiancé and be a dad to her daughter: thank you, especially, to those special Angels among you who made that dream possible.

Upgrade to a paid subscription here
Donate via buymeacoffee here

A final photo as a way of saying thanks to my loyal readers. I have been asked by some people to make prints available of some of the better photos in an online store; my fiancé and I will endeavor to do that in the coming months.