Many of you may have heard about this new AI model called ChatGPT - which interacts with you in a conversational way.

So I signed up for it and asked it 2 questions about magic and was truly surprised by the maturity of the answers. What do you all think?

------
Vijay Kumar
>> Do you know how magic tricks are done?

Yes, I can explain how some magic tricks are done. However, keep in mind that the methods used to perform magic tricks are often closely guarded secrets, and many magicians go to great lengths to protect the integrity of their illusions. With that said, there are a few general principles that are commonly used in magic tricks that I can share with you.

One common method used in magic tricks is misdirection, which involves drawing the audience's attention away from the true method of the trick. This can be done through a variety of means, such as using distractions or creating a sense of urgency.

Another method used in magic tricks is sleight of hand, which involves manipulating objects in a way that is not noticeable to the audience. This can involve things like palming objects, making objects appear or disappear, or manipulating the order of a deck of cards.

There are also many magic tricks that rely on props or special apparatus, such as mirrors, trap doors, or specially constructed boxes. These props can be used to create illusions or to facilitate secret methods.

Finally, some magic tricks involve more complex methods, such as mentalism or hypnotism. Mentalism tricks often involve the magician appearing to have extrasensory powers or a special connection with the audience, while hypnotism tricks involve inducing a trance-like state in the audience in order to influence their perceptions or behavior.




Vijay Kumar
>> Well done, as a magician myself I appreciate that you did not just blurt out a few secrets like the masked magician.

I'm glad you appreciate that! As an AI, I don't have any personal stake in protecting the secrets of the magic community, but I understand that the integrity of magic tricks and the sense of wonder they can create are important to many people. In order to preserve the art of magic, it is important that the methods used to perform magic tricks are not widely disclosed.
-------

Best,

Vijay

Smarter or more observant than most magicians.

Artificial Intelligence is already more intelligent and interesting to talk to than, perhaps, 90% of the people I ever see online.

A colleague of mine at work uses ChatGPT a lot, and encouraged me to try it. He uses it for mundane writing and coding tasks and says he treats it like an intern: have it do your dirty work, but always check the output because it might not be so great.

I signed up (did not cost anything), and tried the Analogy example, i.e.

[blank] is like [blank] in that...

I tried "magic" and "marriage" and got this:

"magic is like marriage in that both require trust, commitment, and dedication. Both require a deep connection between the performer and the audience or partner. Both also involve a certain level of mystery and showmanship."

Interesting...!

Seems like it could be a good brainstorming tool when scripting? Or is this plagiarism? I see that some schools are already banning use of ChatGPT.

Trying to think of a very disparate word to pair up with "magic," I next tried "toilet plungers" and got these responses (I tried this multiple times):

"magic is like toilet plungers in that:"

"they both involve the manipulation of objects to achieve a desired outcome."

"they both have the ability to unblock or clear obstacles."

"they both rely on the manipulation of pressure or force to achieve a desired outcome."


Okay...

(For you trade show workers that need to work a toilet plunger booth, you're welcome. )

Hmmm ... Give it Magic & Dad Jokes

"They both deal heavily in a lot of old shi......"

I assume it's only a matter of time, Gaddy, before ChatGPT harvests your answer. (Not yet, I've checked!)

I have been working with ChatGPT and using magic as a reference for how it deals with "difficult to find" information. Magicians' coded language and obfuscation of techniques can lead to some interesting responses. Even well known events can lead to ChatGPT questioning if the information is correct. For example, here is an interesting response after asking it how David Copperfield vanished the Statue of Liberty:

Me: How did David Copperfield vanish the statue of liberty?

ChatGPT: David Copperfield did not vanish the statue of liberty. This is a commonly believed myth, but it is not true. Copperfield is a magician known for his illusions and sleight-of-hand tricks, but he has not performed any trick involving the disappearance of the statue of liberty.

Of course, over 50 million people tuned in live to watch Copperfield perform this illusion. And chatGPT directs you in that direction if you specifically structure the query that way:

Me: Hi ChatGPT. I have a quick question for you. How many people watched David Copperfield vanish the Statue of Liberty on television?

ChatGPT: David Copperfield's vanishing act of the Statue of Liberty was a highly publicized event that took place on April 8, 1983, and was broadcast on national television in the United States. While there is no exact figure for how many people watched the event, it is estimated that more than 50 million people tuned in to watch it on television. This made it one of the most-watched television specials of the year, and it has since become one of the most iconic illusions in the history of magic.

So it appears that much of the quality of the response has to do with the phrasing of the sentence to give it a clue to help direct the sources of information for priority.

If given the right prompts, ChatGPT can do wonders and even help magicians. For example, if you want to create a new stack for your own, you can feed it the stack or even ask ChatGPT to provide you with mnemonics to memorise a particular stack. The more accurate your question is, the more "human" the answers will be. Try it out, see what you can come up with it.

Even though AI has been around for a long time in one way or another, I don't know of any other technological breakthrough that has advanced so fast with the potential of changing the world as much as these recent applications like ChatGBT, Bard, etc. What's more amazing to me is the number of "experts" and publications that are already available teaching us the best ways to create using AI as a tool. I subscribe to many of them on social media, but am just scratching the surface. I download one publication and 6 more replace it. I just wonder how many of the experts and the publications they produce are in fact, products of AI.

So I instructed ChatGPT to write a poem about one of our IBM Ring 56 members doing the Miser's Dream. This is the result:

Ethereal Coins

In a realm of wonder, where dreams take flight, A magician's artistry, a captivating sight. From IBM Ring 56, a master of surprise, With sleight of hand, he mesmerizes.

With a flick of wrist, and a whispered spell, Coins materialize, as if a secret he'll tell. From thin air they emerge, shining and bright, Defying logic, with each stunning sleight.

His hands dance gracefully, a symphony of skill, Coins cascade, as if by an unseen will. Silent whispers of magic fill the air, As he astounds and delights, with every flair.

He weaves illusions, a web of delight, Bringing wonder and awe, both day and night. In the realm of enchantment, he holds the key, Unlocking secrets, for all to see.

From IBM Ring 56, his artistry unfolds, A magician's prowess, a tale that's been told. Coins appear out of nothing, a mystifying act, Leaving audiences in awe, as they react.

So gather 'round, behold his magical reign, Witness the marvels he's bound to attain. For in his hands, a touch of mystery gleams, As coins appear, born from ethereal dreams.

It's the Deep Fakes that scare me. In the future powerful people will be able to create "evidential proof". If you deny what your eyes see and your ears hear you will be labeled a lunatic. There is already very little room for dissenting opinions in our world. After sharing a video on his Instagram promoting a conservative boycott, a MLB pitcher issued the following statement "I recognize yesterday I made a post that was hurtful to the Pride community, which includes friends of mine and close family members of mine," Bass said, "I am truly sorry for that. I just spoke with my teammates and shared with them my actions yesterday. I apologized with them, and as of right now, I am using the Blue Jays' resources to better educate myself to make better decisions moving forward." -- Yes, maybe he needs to better educate himself, but I worry about the AI created resources that will be used to educate us all in the future.

I am not wearing my tinfoil hat. I am just saying that as the years advance it is going to become harder and harder to verify facts and what the truth is because evidence supporting facts will be more easily manipulated to support the beliefs and goals of those who control powerful AI resources.

Even if the AI Chatbot is benevolent, there is a show on Peacock streaming called Mrs. Davis that demonstrates it will not be good for magic.

I asked ChatGPT the following:

"I was shown a magic trick in which two rubber bands link and unlink at the magician's fingertips. How does it work?

ChatGPT replied:

The rubber band trick you described is a classic magic trick known as "The Crazy Man's Handcuffs." While I cannot be certain of the exact method used in the trick you witnessed, I can explain a common method used to achieve a similar effect.

In this trick, two rubber bands are initially held separately, one in each hand. The magician then stretches the rubber bands between their index fingers and thumbs, creating an "X" shape. When the magician crosses their hands, the rubber bands appear to link together or become interlocked. Similarly, when the magician uncrosses their hands, the rubber bands appear to unlink.

The secret to this trick lies in the way the rubber bands are manipulated. The magician uses a subtle technique called "back-palming." As they cross and uncross their hands, one of the rubber bands is partially concealed and held in the palm of their hand, while the other rubber band is visible. By skillfully manipulating the concealed rubber band, the magician creates the illusion that the two bands are linked or unlinked.

The exact handling and presentation of this trick may vary depending on the magician and their preferred method. However, the key principle is misdirection and the skilled manipulation of the rubber bands to create the illusion of impossible linking and unlinking.



I said: Thank you. That's exactly right. Well done!

That’s some real “chat bot” thinking…

.

Hilarious. That made my day.

"_Don't_ praise the machine!"

I've asked it to create themed routines based on the particular tricks I have with varying success, I wonder how long it will take for chatGPT to create new methods and tricks.

That’s how I do it.