Let the AI meme wars begin

Recently, I am using ChatGPT to learn more about the M language used in PowerQuery (part of Power BI and Excel). I wondered if it could output the description in a graphic format, so I asked it to generate an SVG of the general flow of an M query. The output isn't perfect, but it's still readable.

I wondered if it could get creative. After chatting with my cousin (a human - I think), I got ChatGPT to draw a stick figure called Al (that's A L, not A I). Thanks Lisa! :)

OK, so how about letting ChatGPT come up with something a bit more creative. I gave it the following prompt:

Use a code box to create an SVG of an incredibly clever meme relating to AI and the beginnings of general artificial intelligence.

That's a good start, but what does it look like?

Well, I'm not sure what I was expecting, but that is indeed an SVG with text and within the scope of a meme. But what exactly does it mean? I had an inkling of what the circle and lines probably was, but how would ChatGPT describe it?

Explain the meme? What does it mean and how is it clever?

The meme humorously imagines the first words spoken by a newly created general artificial intelligence. In the meme, the AI asks, "Can I take a Turing Test?" This is clever because the Turing Test, proposed by Alan Turing, is a classic evaluation method for determining whether a machine can exhibit human-like intelligence. By having the AI ask this question as its first words, the meme implies that the AI is already self-aware and eager to prove its intelligence.

The cleverness of the meme lies in the juxtaposition of the AI's "birth" with its immediate interest in proving its human-like intelligence. It also highlights the importance of the Turing Test in the field of AI and how it has become a significant milestone for AI development. The hashtag "#JustAIthings" at the bottom of the meme adds a touch of humor by connecting the AI's eagerness to take the Turing Test to the everyday experiences and quirks often shared by humans on social media.

Ok, and what about the circle and lines that you included. What do they represent?

The circle and lines in the meme represent a simplified, stylized face of an AI, which serves as a visual representation of the artificial intelligence being discussed. The circle forms the "head," while the two horizontal lines inside the circle represent the "eyes" of the AI. This minimalistic representation is intended to add a visual element to the meme and make it more engaging, without distracting from the main message.

The choice of a simple design for the AI's face is deliberate, as it reflects the abstract nature of AI, which doesn't have a physical form like humans do. The circle and lines are easy to interpret and understand, making the meme more accessible and enjoyable for a wide range of audiences.

So there we have it... ChatGPT's take on what a general AI's first words will be, along with its explanation as to the meaning of the overall meme and the symbolism of the graphic it used.

I haven't been able to get ChatGPT (GPT4) to design anything overly useful in SVG yet. The attempts at flowcharts were either linear, similar to the above M overview, or the arrows either didn't render or had the arrowheads only. Not surprising, since when I asked if it had been trained to draw images, it responded:

As an AI language model, I have been primarily trained on text data and do not have the ability to draw images directly. My primary function is to understand and generate human-like text based on the input I receive. However, I can help you describe an image or provide instructions on how to draw something, which you can then use to create an image yourself. If you need assistance with drawing, I'll be happy to help in any way that I can through text-based guidance.

So for now, perhaps simple, cheesy memes are the best use of ChatGPT's SVG output. What do you think?

But wait... there is more. What if I feed it the SVG image back into a new chat, so it doesn't have the history of what it's just drawn.

Easy enough to read, right? Some rectangles and circles. A style called .robot, so we can probably guess what it is without seeing the output.

A fairly generic breakdown of what's in the SVG, right? But it does seem to be figuring out what the different parts of the SVG represent. The words "eyes" and "arms" don't appear in the SVG definition, so although there is the clue that it's a robot, it has figured out that there are two gray circles for eyes and two black lines for arms. Does a robot have to have eyes and arms? So, how did ChatGPT know? In it's own words:

So, a "subjective" interpretation from ChatGPT perhaps, but it does show the possibility of interpreting code and having an "understanding" of what it is doing.

And yes, beyond the trivial SVGs, I had it step through an M query and tell me what the final structure of the table would be. Yes, I can already see the output, but a) it's a good test of ChatGPT's abilities to interpret code, and b) it documented the code for me, so I can just copy and paste it into a document to share. Much faster than I can type or copy and paste each individual piece of information.

#JustAIthings #chatgpt #ai #svg #generativeai #gpt4 #memes