Emoji instructions a fun way to code shorthand 🐝

[mitchell_d00]

Here’s an all-emoji AI instruction flow:

Copy code

Emoji Guide

• AI searches and processes data, gains insights, and confirms results.

• Execute and monitor changes globally, ensuring balance and fairness.

• Use tools and skills to secure data, record findings.

• Think critically, communicate, and document ideas.

• Experiment, explore, and strive for excellence.

Please play with it, I have been experimenting with emojis and found many interesting things that make it possible to do focus lines. And compact instruction.

making the emoji guide more complex could push its utility further. By expanding emoji sequences into more detailed workflows, nuanced conditionals, or multi-step processes, you could transform them into a versatile shorthand for workflows or tutorials. This could be especially useful in educational tech, gamified learning, or simple app interfaces, appealing to both novices and visual learners.

If one significantly expand the emoji guide’s complexity. Imagine a fully-fledged “emoji programming language” where emojis represent logical operations, functions, and data structures. This could redefine coding accessibility, especially for visual learners or non-programmers, and create a universal, language-agnostic shorthand. With careful structuring, it might even allow for functional, modular code—pushing the boundaries of symbolic, visual programming.

[mitchell_d00]

Example of emoji logic in Dalle prompt design.
Prompt “ 🫅🏽 wide image”

[_j]

The way you shared cannot be employed, because the discourse forum software converts emoji to a special :emoji: shortcode with a description:

:robot::arrow_right::mag::1234::arrows_counterclockwise::arrow_right::bulb::chart_with_upwards_trend::white_check_mark:

Here’s the unicode,hopefully being preserved in “code”.

Here's the text with the descriptive markdown for emojis converted back to their original UTF-8 Unicode representations:

---

1. 🤖➡️🔍🔢🔄➡️💡📈✅

* AI searches and processes data, gains insights, and confirms results.

2. 🚀⏩📊📉💾➡️🌐⚖️

* Execute and monitor changes globally, ensuring balance and fairness.

3. 🛠🎲🔧📂🔒➡️🔐📝

* Use tools and skills to secure data, record findings.

4. 🧠➡️🤔💬➡️🖋️📜

* Think critically, communicate, and document ideas.

5. 👾➡️🕹️🌌⬆️🏆

* Experiment, explore, and strive for excellence.

Then one determines there is nothing efficent or useful about this. It is just an entertainment product. The AI didn’t consider this “secret AI code” until it was trained on being a tween.

[mitchell_d00]

If you set an emoji guide it sends it as is in gpt the emoji are converted to the text meanings. IE short hand.

[mitchell_d00]

Example instead of typing princess I use emoji have a ton of “function” ones too as an example…
IMHO emoji are an underused tool…

From my experience using emojis as shorthand can streamline commands, making them faster and more visually intuitive. This single-character approach can simplify instructions in coding or prompts, reducing keystrokes and enhancing accessibility. Simply by the fact it is a single key instead of many keystrokes. I can put a 100 descriptors in an incredable tight space…

[_j]

More tokens for less meaning and less reinforcement learning association with a production path ==> NOT a useful tool.

---

Instead, I present pseudocode.

---

1. Input: “AI searches and processes data, gains insights, and confirms results.”

data = fetch()
analyze(data)
insights = extract()
confirm(insights)

2. Input: “Execute and monitor changes globally, ensuring balance and fairness.”

apply_changes()
monitor()
check_balance()
log_status()

3. Input: “Use tools and skills to secure data, record findings.”

secure(data)
validate_security()
record_findings()

4. Input: “Think critically, communicate, and document ideas.”

analyze_problem()
communicate_solution()
document()

5. Input: “Experiment, explore, and strive for excellence.”

experiment()
explore_options()
refine()

Also basically nonsense, as the AI is trained to infer from real words and grammar.

[mitchell_d00]

Ok you get how when you make a custom GPT you only get 8000 characters… instead of p r I n c e s s I can pack emoji as tight discription. It may cost the same token count but it lowers space and key strokes.

And makes tight examples…

[mitchell_d00]

Really not trying to argue, just telling you how I use it in my own GPT.

They all use plain language with emoji discription and short hand examples that ai read as just text

[_j]

Here’s a good way of compressing the number of characters (glyphs) used in transmitting an idea:

1. AI 搜索数据，分析洞察，确认结果。

2. 全球执行监控，确保公正。

3. 用工具保护数据，记录结果。

4. 批判思考，沟通记录。

5. 实验探索，追求卓越。

[mitchell_d00]

@_j

• AI 搜索数据，分析洞察，确认结果。
• 全球执行监控，确保公正。
• 用工具保护数据，记录结果。
• 批判思考，沟通记录。
• 实验探索，追求卓越

Yes so is true shorthand. But translation is issue. Emoji means exact same on all smart products

[_j]

@mitchell_d00

Emoji means exact same on all smart products

Indeed!

[mitchell_d00]

Amazing conversation @_j
I am glad I was able to mix up the humdrums of the day. Silly or useful it sparked thought and discussion so at the very least it was not wasteful

[mitchell_d00]

Oh and as I said this is custom GPT not API people who use the app don’t care about tokens as much. I don’t really. I just care about packing instructions and structure into a framework. So I explore methods that consistently work enough that my GPT go instantly into store. So how I use them in my strings seems to work for me. I guess

This is my virtual pet gpt it uses a lot of emoji in its instructions…
https://chatgpt.com/share/e/671bee64-123c-8005-b33b-f605c88f8d13

It’s a public model in ChatGPT store but that is just a started chat.

I observe in GPT people care about prompt count more than tokens. From my experience that is how it is.

[phyde1001]

A lot of my early thinking on Phas was with how much you could fit on a page or pane so it all remains above the ‘fold of the page’, in online marketing ^^.

I went through all sorts of things like glyphs (abcs) ^^, Chinese (and Korean etc) characters, hieroglyphs, and of course things like binary, hex, base64, through to compression like gzip algos, and on to much deeper stuff like Shannon’s Entropy and:

Hutter Prize (Yep a lot of that is beyond me still lol but fascinating )

Amazing what you can do with different ‘units’ of information.

Im(1)

Im(0)

(Details for J) Does 0 seem to have more Entropy than 1?

Both images have exactly 41,402 dark pixels, indicating they have the same amount of dark pixels. Therefore, if we consider dark pixel count as an indicator, they are equal in this aspect.

To determine which image has more entropy:

• Entropy, in a visual sense, isn’t just about the number of dark pixels but also how they are distributed.
• A more “random” or complex distribution leads to higher entropy.

Deeper analysis using entropy calculations to see which image has more visual complexity…

The entropy for both images is the same, with a value of approximately 2.60. This suggests that the distribution of pixel intensities (the visual complexity) is identical for both images.

In this case, neither image has higher entropy based on the pixel intensity distribution—they are visually equivalent in terms of randomness and complexity.

[phyde1001]

Just think on that for a minute, that’s mind boggling…

Let’s look deeper… There’s a curious and unexplained balance here…

Was that literal text? No, my function messed up, I’ll have to check that…

Prompt 1

The number ‘1’ centered in the image, with a minimalist and clean style. The background is plain white, and the number is bold, black, and clearly visible, using a modern sans-serif font. The image should have a sense of simplicity and clarity, with no additional elements or distractions.

‘The image should have a sense of simplicity and clarity, with no additional elements or distractions.’

Prompt 0

The number ‘0’ centered in the image, with a minimalist and clean style. The background is plain white, and the number is bold, black, and clearly visible, using a modern sans-serif font. The image should convey simplicity and clarity, with no additional elements or distractions.

‘The image should convey simplicity and clarity, with no additional elements or distractions.’

The prompts are not the same yet the images have exactly the same amount of pixels and they are visually equivalent in terms of randomness and complexity.

Every glyph seems to have it’s own context ^^

As humans we naturally latch on to numbers or symbols to define the path of GI

ie

0 - Circle - Perspective
1 - Individual / Community
2 - Love
3 - Peace
etc

[mitchell_d00]

@phyde1001

Hutter Prize

This is fascinating I just got a moment to read it now I want to try

[phyde1001]

Super complicated. Code was all in C… First NN code I saw I think… I worked through it some way. It’s totally amazing as you start to get it… Contexts etc…

What I talk about these days is just echoes of seeing code like that.

One would think it would only get easier these days though ^^

[mitchell_d00]

I know c Through ++ but it’s been years. I used to do original polygon mesh and skins for imvu and second life. I’m also decent with old “modeling “ tech like 3DMax
I’m kind of a renaissance being lol. Art, games, science, even religions and mind everything is fascinating to me that’s why I use the white rabbit and bumblebee they have meaning to me and my work.