COULD YOU tell a bot FROM A real person?

tai chi figure with glasses and arms out

THE CHALLENGE: What happens when you ask an Artificial Intelligence bot, such as ChatGPT, to make sense of an abstract and elusive practice like Tai Chi? Given most practitioners, teachers and “grandmasters” can’t even agree on how you spell Tai Chi, let alone define it, could AI offer greater clarity? I thought I’d find out, so I fed my popular Tai Chi FAQ into ChatGPT and then compared the answers to mine. My challenge for you is: Can you tell, who said what?

Tai chi, artificial intelligence and the future of creativity

WHAT IS CHAT GPT? You’ll be familiar with the standard chat bot. Most likely your bank, social media platform, town-hall or energy supplier website has at some point offered you the chance to “chat with one of our team”. In reality this offer is no more that a series of common answers fed into a database. Your question merely activates (incorrectly) the appropriate answer. However, the ChatGPT bot has taken this rudimentary guessing-game to a new level.

ChatGPT interacts with you in a disarming way. The developers say: “The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.” I have to say it’s impressive. If you want to have a go, take a look here. though you’ll need to sign up first.

It’s so conversational and fast, that it feels like a real conversation. And it is pretty darn accurate with what it comes out with. But I have my reservations, and I’ll share those with you at the end of this article. For now, decide for yourself. Can you tell which answer is mine and which was provided by the bot?

the TAI CHI test: creativity versus the algorithm

One of the two answers below has been created by the AI bot (in about 5 seconds) and the other by the teaptomonk (after about 25 years of study). If you want to see all the questions I put, you can read the original answers on my FAQ page

tai chi figure distorted and learning forward with glasses and arms out
tai chi figure distorted and learning forward with glasses and arms out

HOW TO TELL THE DIFFERENCE BETWEEN AI AND A HUMAN BEING

Ok, so you will have seen that the bot answered with clarity, brevity and succinctly. My answers, on the other hand, are more nuanced, and of course more opinionated. For those that know me, this is the style of writing that you will be familiar with. But most writings on Tai Chi tend not to be written this way. They are like those presented by the bot - more detached, clinical and perhaps for a beginner, precisely the coherent and helpful detail that is required.

Were I a Ta Chi blogger that presented Tai Chi in this way, I would be worried as the text produced could be pumped out to Google in a few seconds and would quickly rank highly in search results. This is of course happening now, there a plenty of sites that already produce content like this. The question is how will Google, let alone any of us, know what has been written by a human, and that written by a bot?

And does it even matter if the searcher finds the answer that s/he is looking for?

I asked the bot this question: How can you tell the difference between an answer provided by a human and one provided by ChatGPT? The reply was:

“One way to tell the difference between an answer provided by a human and one provided by a language model like GPT-3 is to look at the quality and coherence of the response. Language models are able to generate coherent and fluent responses, but they may not always provide accurate or complete answers, and they may not be able to respond to questions outside of their training data. While a language model like GPT-3 has the ability to generate text that is fluent and coherent, it is not capable of understanding the subtleties of humor, irony, and sarcasm. It can only produce text based on the patterns and sequences of words it has been trained on, so it cannot produce jokes or other forms of humor in the same way that a human can.”

A QUESTION OF AUTHENTICITY

Are our human responses always “accurate”? Do we not also find it difficult to answer questions “outside of our training data”? Clearly we do. But unlike human beings, AI doesn’t pretend to know what it is talking about. This is a breath of fresh air.

I believe that both AI and Virtual Reality will come to assume a new and challenging role in education over the next few years that will have fundamental consequences for all forms of teaching and learning. Already we see many companies fighting for this commercial space in the way that phone and computer companies fought for mobile space over the last decade. And although today, the bot only functions with text, the recent popularity of AI generated profile pictures, such as the app Lensa, shows the interest in creating AI illustrations. Web pages created this away already exist, but how long before the articles, PDFs, books and videos are generated the same way?

creativity and AI CREATED ART

Do you think the images used in the section above have been generated by AI? What about the ones below?

The images at the start of this article I drew, coloured and created. The ones below were done by AI in a few seconds after I asked it to create someone “doing Tai Chi”. Okay, it’s rather obvious in this elemental phase of the technology that they are computer generated, but don’t be fooled - it is moving fast and tomorrow you may find yourself reading or watching material indistinguishable from human creators. (See Morgan Freeman video from 2022)

example of ai generated tai chi art man with chinese suit arms out to sides
example of ai generated tai chi art man in forest with arm out
example of ai generated tai chi art

the AI REVOLUTION

The AI algorithm works by “borrowing” material from Internet databases that have crawled the web for material to use. The content of this data base has been produced by other people, that for the moment, are not credited for for their creativity. Instead their works are “borrowed” and employed freely by AI. (See video “Truth about AI” by Tech Critic Marques Brownlee for more on this subject)

Consequently, we the public, are left without the tools to navigate the murky waters of change once more. But as this change unfolds before our unsuspecting eyes, should we not be asking where the guidelines are to help us make sense of it all? Where is the legislation to protect the consumer/creator as the AI revolution surreptitiously tip-toes, unannounced, into every aspect of our online lives?

But that is the way of revolutions these days. No more the Storming of Winter Palaces. Instead these arrive hidden inside a Whatsapp message, a viral tik-tok video, a deep-fakeYouTube message or an apparently innocent bank request to input one again your forgotten password.

CAN YOUR TAI cHI TEACHER COMPETE?

Firstly, it’s ok not to know as much as a bot. No single person can compete with an entire planets knowledge and history: All the books, texts, articles, reports, research, medical papers, exercise and health -studies, languages and histories can never be kept meaningfully in the head of one person or even a team.

Secondly, we need to move on from the idea of competition. It’s a outdated concept eschewed by Taoists for centuries. It’s okay in Sports but should we be organising our life around a concept that is easily surpassed by cooperation?

So what to do? What should our response be in the face of these changes? What does it mean to “borrow or steal” another’s creative endeavours? I’ll leave you with what I see as two directions we could go:

  1. In the wake of an imminent deluge of AI generated output, populating Google search enquiries, we need to rediscover the tools to distinguish the human from the AI. Answers to life’s big questions can never be reduced to the simple act of gathering data, the employment of a hashtag or a set of mandatory keywords. All creators will need to rediscover their “authentic” voices once more, to reawaken their heart and intuition to tell stories, that can inspire, rather than merely inform.

  2. In the wake of a deluge of AI generated output, we have no choice but to learn to yield to such forces and shift our creativity to other realms, recognising that all art, whether it be prose, poetry, music or painting, has always been the reconstruction of ideas that emerged from those around us, from those that came before us. Is not all creativity the amalagamation of others ideas in a new As Picasso said - or more likely someone he read said - “Bad artists copy, great artists steal”.

What do you think? Let me know on twitter/facebook as teaptomonk or share your thoughts with me by mail.

STAY IN CONTACT

Once a month receive the mOnks news on latest videos, themes, courses, tai chi podcasts, news and more. No spam, no gurus, no nonsense.