(Close Window)
Topic: This is the best release you'll acquire this year
Message: Posted by: Conner (May 20, 2017 08:50PM)
So many performers get caught up in the cycle of buying the latest and greatest secret. You know the sort. The underground, limited-edition, overpriced whatchamacallit which is, in fairness, very clever; but once the honeymoon period wears off, is still just another way to learn a word or duplicate a drawing. You've just spent $XXX to deliver an experience to your audience that you've been delivering since reading Corinda and Annemann.

Worse still, this endless hunt for the perfect secret draws valuable time and effort away from refining the performance you already have now. The result?

"Draw a picture."
(Plastic, inauthentic patter about lies or some such)
"You drew a house."

It's just fine. It can amaze people. But it's just fine. It's not the transcendent miracle your audience deserves after you've paid $XXX so many times for such clever secrets. You have already spent too much money for mediocrity.


THE PRODUCT
Whatever secret method you use, the presentation is far and away more important.

However much money you were going to spend this year on underground whatchamacallits, save it. However much money that is in the form of time and effort, spend it on this instead. Spend your time and effort on what I am about to give you.

Google has collected 50 million drawings by average people. Google also recorded the way in which the picture was drawn. The game “Quick, Draw!” gives players 20 seconds to draw a picture as it uses machine learning to identify the drawing. Each drawing expands the database making the game more accurate. The data records what shapes comprise each drawing, how they are drawn, and in what order. This translates to little animations of drawings being made as an average person would make them. Fifty million times.

This data set is now available for free.

The more experienced of you will be salivating profusely at this point. No doubt, this will remind you of Banachek's cat drawing duplication from Psychological Subtleties and the Psi Series. Banachek's ample experience allowed him to describe in real time how a participant will draw a cat. Every shape thought of was described at the exact second it was thought of. That is real mind reading. That's the kind of miracle you expect to provide audiences after paying $XXX so many times. To be in a person's mind as they think, seeing what they see as they see it. Why was this so powerful? Not because it was a new secret, because it was a brilliant presentation.

This new data set means you do not need the thousands of trials from the experience of another performer. You now have the TENS OF MILLIONS of trials from Google's data collection.

As accurately and precisely as Banachek describes in real-time how a participant is drawing a cat, so too can you describe in real-time how a participant is drawing an image only in their mind. You will be able to walk around in your participant's mind, not just for a cat, but for hundreds upon hundreds of drawings. This is the kind of payoff people hope for by sinking big bucks into some underground release.

Do not waste this.


PRICING
Other authors might have kept this resource a secret or sold it at a price. I certainly debated whether or not I should share this priceless information and in what way. Used diligently, this information will net professionals thousands of dollars.

There is a cost to this. Seriously. I am not going to charge dollars, but I do want to charge you in time. I want you to take the money you would have spent on the cool new underground $XXX whatchamacallit and do some math. Whatever your next big, sexy purchase was supposed to be--calculate how many hours of your work it would take to pay for that. That's (cost of whatchamacallit / dollars you make per hour). Whatever number that is, I want you to commit that many hours to incorporating this drawing data into your act. For example, if you wanted to buy a $200 dollar thing and you make $10 an hour--please pay me by dedicating ahead of time 20 hours to studying the data, writing notes of your ideas, scripting your routines, recording yourself, and practicing with people. Please schedule this time in your calendars. Plan around it.

Once you get started, you'll probably end up using more hours than you committed anyway. It's a lot of fun refining what you love. But if you do not commit to that time in advance, it is too easy to glance, hit writers block, and walk away into another generic purchase. By committing to the time in advance, you will power through the lulls and end up discovering new epiphanies that will enhance your audience’s experience. No dollar amount sent through PayPal will improve your show. You need to put in work hours. That's what I want from you. You are not accountable to me, you are accountable to your audience. I am giving you a gold mine. If you waste it, your audience loses amazing experiences. Please do right by them. By investing time in yourself, you are investing in this art form that I love so dearly.


Here are the links. Go.

- "Quick, Draw!" game: https://quickdraw.withgoogle.com/
- List of drawing categories: https://github.com/googlecreativelab/quickdraw-dataset/blob/master/categories.txt
- Download the raw data: https://github.com/googlecreativelab/quickdraw-dataset
- Animations of each drawing: https://quickdraw.withgoogle.com/data
Message: Posted by: illusioneer (May 21, 2017 01:23AM)
This is gorgeous!!!
Message: Posted by: bowers (May 21, 2017 01:38AM)
Thanks for posting very interesting.
Todd
Message: Posted by: pacozaa (May 21, 2017 03:30AM)
Thank you. This is crazy.
Message: Posted by: pacozaa (May 21, 2017 03:35AM)
Oh let me share something too.

"Tensorflow"
Message: Posted by: Mobius (May 21, 2017 03:47AM)
The beauty of this post is not solely in the Product but in the first four paragraphs. So true!
I've just had a quick look at the animations of each drawing and it's such rich and amazing data. Taking the example of "cat" you can instantly see how the majority of people draw it - when I say "majority" I mean how roughly 103,000 people draw it! In addition, since you get to see how people actually imagine a specific object it is also very useful for effects where the image is pre-drawn - I realise I now need to make sure that my pre-drawn feathers/leafs lean to the right and my flowers are upright!
So, no matter how much I hate Google (a corporation that has far greater aspirations to read people's thoughts and interpret their actions than the most diligent mentalist) I will look into this further. I saw Banachek at The Event earlier this year and his ability "to know" which component of the drawing the spectator was focusing on in real time was extremely impressive and very strong.
First stop will be to see if an idea in Secret Impressions by Paralabs (The DOP Twist) can be used to discern an image that is never actually drawn simply by knowing how the majority of people start drawing it.
Thanks for posting. I will give it the time it deserves and refrain from pre-ordering the boxed, limited edition gold deluxe version of hippity hoppity rabbits that I knew I had to have ever since I saw the extremely cool,though slightly confusing, trailer 10 minutes ago.
Message: Posted by: Alexxander (May 21, 2017 07:16AM)
This is absolutely amazing!
Thank you very much for sharing! :)
Message: Posted by: Nerdy Wizard (May 21, 2017 05:19PM)
Heh, I remember someone posting a link to this same resource not too long ago. At the time it went largely ignored. Guess it truly does all come down to marketing =P
Message: Posted by: Mobius (May 21, 2017 05:52PM)
Appears that most people struggle to draw a plane!
Message: Posted by: Nat_lawson (May 22, 2017 08:58PM)
Just played this for about 15 minutes and am thoroughly amazed by the software. Some of my drawings (all) are horrible but it still guesses them correctly the grand majority of the time.
Message: Posted by: Max Wells (May 23, 2017 10:21AM)
Funny how most cat drawings do have a mouth.
Message: Posted by: rev_marcus (May 23, 2017 02:26PM)
Oh good god is this beautiful, Thank you Conner
Message: Posted by: Fred E. Bert (May 23, 2017 04:26PM)
When this first came out, I kept playing and screen grabbing iterations of each drawing. A long and tedious process! Thanks for the data set Conner :applause:
Message: Posted by: DrTodd (May 24, 2017 12:32AM)
Very cool. Had a mouse Sunday night :-)
Message: Posted by: Banachek (May 29, 2017 01:15AM)
That's true Max, however they do not have a mentalist interrupting them, taking them outside the head and rushing them to finish.
Message: Posted by: ko_brian (Apr 16, 2019 12:27PM)
Priceless indeed . Thank you very much for sharing this!
Message: Posted by: Looch (Apr 16, 2019 01:26PM)
I produced a detailed chapter on this exact thing in the Neil Scryer & Friends book. I also covered some of it in the 2013 Penguin download ‘Retina’
Message: Posted by: Chris K (Apr 17, 2019 03:22PM)
[quote]On May 21, 2017, Nerdy Wizard wrote:
Heh, I remember someone posting a link to this same resource not too long ago. At the time it went largely ignored. Guess it truly does all come down to marketing =P [/quote]

I can't remember if it was largely ignored, but I have the links saved from when they were posted originally and they are SPECTACULAR sources of information. I especially like the ones that show the way people do the drawings. These links are great.

I'm having a hard time seeing the applicability of the Tensorflow comment but I'm looking at the page too. I'm always down to learn something new.
Message: Posted by: aligator (Apr 19, 2019 12:46PM)
Now I know beyond a doubt how poor I really am at drawing anything. My most ambitious drawing duplication involves a smiley face. :)
Message: Posted by: John C (Apr 19, 2019 01:02PM)
Cute.
Message: Posted by: weirdwizardx (Nov 14, 2019 11:29AM)
Amazing! Thanks for posting!