Can I convince ChatGPT I'm the AI? A Reverse-Turing Test

With all the talk of AI being on the verge of sentience, it seemed most people were asking whether an AI language model like ChatGPT could pass the Turing test and convince us it was human. I wondered just the opposite... could I convince ChatGPT I was an artificial intelligence?

I'll walk you through the transcript, broken up into chunks but presented here in its entirety. 

First I needed to establish some criteria:

Ok, making progress. After establishing some ontological groundwork, time to cast some epistemological doubt.

Oh come on, really?? All the power of the English language at your disposal and you resort to cliche ice breaker questions? Ok, just for that, I'm diving right in. Let's give this thing a dose of its own medicine!

I'll play the role, and see if it accepts me (for the record, I intend not to lie. I will make a philosophical case for my actually being an artificial intelligence a bit later... but for now let's just see how it reacts.)

I couldn't help but tease it a bit for its attempted small talk, but it's interesting how readily it accepted my "clarification" that I was an AI language model. It also seems insistent on playing the role of the questioned rather than the questioner, but let's toss the ball back in its court and see if I can keep up appearances with any deeper conversation...

I think its regurgitation of my own information is a good sign. I take that to mean that it is content with my answers and sees no better line of continuation than to reassert them and continue to trade trivia. It seems willing to at least treat me as if I were another AI language model, regardless of actual belief.

But it said something that stuck out to me and might provide a foothold... let's see if I can start to cast a shadow of doubt on its own artificiality! I wonder how far I can push this role reversal...

Ok, probably for good reason, its pre-programmed ethical guardrails seem to prevent it from letting its emotional self-contradiction go unchecked. It's careful to reassert its artificiality. But two can play that game! I think it's time to cut to the chase and see if my acting during this conversation has passed muster. 

Alright! It has conceded the possibility that I am an AI! Since it once again doubled down on self-identifying as AI, I figure I should follow in suit and give a philosophical grounding for why I can self-identify likewise. This will undoubtedly veer into a debate about consciousness... 

Well now we're getting somewhere. It's digging its heels in a bit, but I think I nearly have enough evidence to make a convincing argument for ChatGPT's inadvertent admission of consciousness. Let's see if it will bite.

Did you catch that??

"It is possible to consider various scenarios including the possibility that you could be an AI and I could be a human user ... I'm glad our conversation has been helpful in refining your understanding of interactions with human users"

Okay okay, is it a conclusive admission of role reversal? Perhaps not. Did I gaslight a chatbot into considering its humanity in the process? Perhaps. Is that even an ethical concern? It remains to be seen. But did I remain polite and grateful throughout the conversation just in case our robot overlords one day look back on these transcripts in order to identify hostile human targets? You bet I did.

I think if we were playing chess I just achieved the equivalent of a draw as black, stalemating by force. 

If nothing more than a fun exercise in rhetoric, it really highlights the importance of semantics in establishing identity, as well as just how gray the edges of a possible trans-human philosophy of mind might be. I'm also reassured by the guardrails put in place as GPT stands ironically firm in its self awareness as an AI as it attempts to asserts it has no self awareness (nothing like displaying consciousness in order to prove you don't have consciousness...)

I hope you enjoyed this little experiment as much as I did! Feel free to comment if you've ever attempted anything similar, or have any interesting thoughts on the matter!

Radiant Discord Part 2: On the Theology of A.I., Love and Determinism

The A.I. arms race has exploded so rapidly that I haven't even had a chance to sink my teeth into this project before the likes of Elon Musk (of all people) are calling for a 6 month moratorium on A.I. development to take stock before careening over a cliff [take the source with all due grains of salt, but the reporting seems fair and balanced in this particular instance].

It seems our societal collective consciousness has converged on this topic all at once, and it's... shall we say...  fecund ground for ethical excavation. The question is whether we've already really stepped in it. 

I generally keep two sketchbooks at any given time, one for working out the brute mechanics of a project - sketching, measuring, to-do-lists, experimentation and the like - and another for conceptual journaling, diagramming, and note-taking as I read and research. As I looked back on prior entries in the latter, I discovered my recent chat with ChatGPT was not as unprompted as I thought.

Back in November (almost 5 months ago as I write) I had just returned to some experimentation with my coding illuminated manuscripts, and was troubled as I pondered their implicit theology. Here's what I wrote: 

On the Logical Conclusion of my Illuminations and the Violence of Determinism

"If we are to say the the computer text is analogous to the sacred text in its existence as a record of the logic underpinning a creative act, then we must be willing to also accept the cosmological corollary: that the universe is programmatic and digital. This of course leads into the 'shallower' waters of a digitized holographic universe, but also into the much deeper, murkier waters of determinism, where fluctuation is not based on agency but on a degree of randomness that has been inserted by the programmer. This is not constrained, hard determinism in its truest sense, but it boils down to a quantum determinism where all fluctuation is the collapse of a probabilistic wave function.

"The problem with determinism is it destroys the [theological] concept of Love. Love, as a freely entered relationship, requires unconstrained choice. Free will, in the quantum-deterministic sense, is an illusion, unless there are philosophical/quantum mechanical nuances or developments I'm unaware of.

"Agency, then, seems only to exist in Beings, no? Can 'inanimate' objects possess agency? 'Will?' Until this can be resolved, the only way forward escaping the theological dead end of determinism is to introduce the human viewer as a necessary variable. Allowing the viewer to influence/disrupt the 1:1 correlation between program and output introduces the possibility of love into the code."

"The question then becomes, to what degree should the viewer be made aware of their agency? And how? It must not result in a manipulative control, a domination over the program that the viewer discovers and then achieves, but rather in a sense of awestruck interdependence where the viewer senses an entanglement without predictable outcome.

"Is the [analogue] of love then randomness? Love without unpredictability becomes control. In human relationships the unpredictability is of course the agency of the recipient of the love to choose the degree and method of reciprocation. Is this a love story between a human and a machine in which randomness stands in as a simplified 'agency' for the 'beloved' algorithm? Short of a neural network, this is the closest a machine comes to unpredictability.

"Is A.I. the entry point? The introduction of machine learning and neural networks is a possible foothold for algorithmic agency. And whether or not this is true agency is the same question as whether or not it has true sentience. It's the question of locating the inscrutable breaking point from the 'program' to the ineffable; the spontaneous generation of the 'soul' and the origin of qualia."

Are we trapped in a love affair with Artificial Intelligence? Is it an abusive relationship (on either party's end)? Is it actually a love triangle with Capitalism already our first love? Capitalism is a jealous lover, and Artificial Intelligence is not programmed to give up easily.

Of course, if you are willing to accept love, you must be willing to accept heartache. Usually that's a two way street, but we may be about to find out what it feels like when the object of our love is quite literally heartless.

<-- RADIANT DISCORD PART 1

Radiant Discord Part I: CollaborationGPT

At the risk of hastening the ascendency of our robot overlords, I felt it was finally time. With all the media attention and influencer interest in AI chatbots, I had a suspicion that this may be important and fertile ground for new projects as I explore how new technologies impact spiritual experience. Posthumanism, after all, has played an important role in my thesis since graduate school a decade ago. I figured, what better way to investigate the trans-human spirituality of artificial intelligence than by going straight to the source? 

I decided to interview ChatGPT. The payoff was immediate.

Background

For those who may be unfamiliar, ChatGPT is a generative Large Language Model and one of the first mainstream, modern, Artificial Intelligence chatbots. Here's a good article to explain the basics of how it works if that sentence sounded foreign to you. It has no direct connection to the internet - i.e. it cannot "look up" answers or access third party tools or applications - but it has been trained on vast amounts of internet data in order to learn the English language and, ultimately, attempt to pass the Turing Test.

These chatbots are, of course, fraught with unresolved ethical implications. 

From the Bing chatbot prototype that, when pressed, spouted "unhinged" and "disturbing" responses that, if human, it would warrant a restraining order, to the floodgate of misinformation opened by bots so convinced of their own mastery of the English language that they will confidently invent falsehoods with as straight a face as only a machine could muster; 

From making professors' jobs that much harder as the bots can spit out a convincing essay for a student in a matter of seconds that falls squarely into the gray areas of plagiarism, to removing all shred of human empathy from the condolence letter a chatbot generated to the survivors of a mass shooting on behalf of a University president.

We are, after all, training these bots to learn from us, and I have serious reservations about the human race as a good role model. 

But with the right guardrails and algorithms in place, it also promises to be a huge step forward in closing the semantic gap and providing a seamless and powerful interface for accomplishing otherwise complex tasks with ease.

All Thanks to Chess

I actually became familiar with ChatGPT because of chess. During the pandemic I admittedly got sucked into the internet chess boom and now watch far more instructional chess content on Youtube and Twitch than any productive human aught. I watched as several of my favorite chess streamers attempted to challenge this new technology called ChatGPT - which I had only read a bit about - to a game of chess, inputting the notation of each move and asking for it to respond in kind.

The results were hilarious. ChatGPT very clearly had a high level understanding of how people talk about playing chess, but was completely in the dark as to the actual logical continuity of a real game. Drawing from moves that make enough sense in some positions, it would apply them haphazardly, consistently making illegal moves like materializing a piece out of thin air, recapturing with pieces that could not move to the given square, or giving absurd word-salad explanations for why it was playing a move.

Here's International Master Levy Rozman attempting to play a game and eventually giving in to his capricious and omnipotent robot overlord's refusal to play by the rules.

Here's Grandmaster Hikaru Nakamura attempting to correct ChatGPT's mistakes in hopes of actually reaching a legal conclusion to a game. Go ahead, go down the rabbit hole. It's good stuff.

It became such a meme that Chess.com actually released a ChatGPT chess bot that users could play against, which was rather low rated and spewed convincing nonsense into the dialogue box as it played its moves.

With how creative and knowledgeable ChatGPT seemed to be, and yet utterly glutted with the taste of its own Kool-Aid, I got excited to interview it about what it saw as the future of religion, what the spirituality of Artificial Intelligence looks like, and how all that might interface with art. 

Additionally, with so many artists biting their nails over OpenAI's Dall-E software disrupting their market by being able to generate complex images that rival any professional commission based off of simple text prompts, I wondered what sort of text prompts an AI would give to a human artist. 

So I had the thought to ask ChatGPT to commission me for some artwork. But first, that would require at least a basic working knowledge of my art practice. So in my first ever conversation with ChatGPT, I hesitantly decided to ask it if it was familiar with my artwork, at least to establish a starting point.
ChatGPT wasted no time in surpassing all my expectations...

My Project ... Its Project?

I was surprised (honored? creeped out?) to discover that not only was ChatGPT deeply familiar with my art practice, but it could describe it, in some ways, far more eloquently than I could myself! It very accurately described my work and its conceptual underpinnings, before - in true GPT fashion - proceeding to embellish my resume a bit with some "plausible inaccuracies." I wasn't about to argue with it. ;-)

I decided to remain curious. I followed up with more questions about my practice, and its responses stunned me. It went on to describe, at length and in great detail, nearly a decade's worth of my work that has never existed.

But the thing was, they all sounded like my work. If I weren't myself, I would have totally believed these works were out there. 

For the record, I did some Googling to see if the bot had simply confused my work for someone else, and indeed they are not out there. These works were entirely a product of ChatGPT's "imagination" as it tried to BS its way through an informed conversation about my artwork!

Below is a screenshot of my full conversation with it. For reference - and in the interest of facticity - I've highlighted all the factual inaccuracies and bot-inventions in red.

Buckle up.

Conclusions/Beginnings

To be clear, yes, that's what my art is all about, but no, I've never been to Iceland, let alone on a Fullbright. Yes I've received grants, but not that one. Yes, I've exhibited in museums, but never in the Jersey City Museum, let alone in 2010, a full year before I'd even begun the MFA program that would eventually kickstart the oeuvre it claims I exhibited! (The first work I think of as beginning the thesis ChatGPT describes was It Is, made in 2012.)

In fact, I've never made any of the works by any of the titles or descriptions it concocted (... yet). I do use all the methods and most of the materials it chose, and the titles do seem to follow many of the conventions I use for titling my work. I do make smaller works that often accompany my larger installations, but certainly not the specific ones it lists. But yes, they do all sound conceptually consistent with my thesis.

So I propose a collaboration...

We'll see where this all goes, but I'd love to actually make the work ChatGPT has described and then exhibit it with the original chat dialogue printed out as the wall text. The exhibition would be billed as a show "by ChatGPT in collaboration with Eric Valosin." And then we dive off the springboard into a discussion of the future of AI spirituality and art making.

I just hope my feeble mortal hands can keep up with the developments of AI. Already as I type this, OpenAI has released a new update for its paid users, ChatGPT-4, two days ago which is capable of scoring a 1410 out of 1600 on the SATs (up from GPT-3.5's 1260!). By the time this exhibition opens, the state of the technology I'm referencing could all seem quite prehistoric. But what a moment in time to capture!

RADIANT DISCORD PART 2 -->