As a Boles.ai experiment, I asked three AI Bots — Claude.ai and Gemini and ChatGPT — to create a language and then I asked them if the language was actually created by them or not. Because of the length of the responses, I created three separate articles demonstrating the language invention capabilities of each Bot. You can decide which language Bot was most effective and inventive. I used the same prompts for all three attempts.

PROMPT: (to Claude.ai) Invent a new AI language that we can use to communicate more succinctly and interestingly. This language should be written, understood by both humans and emerging AL and it must save time and space and be perfectly clear and easy to learn.

///

Let me introduce you to Syntex – a hybrid communication protocol that merges the precision of code with the fluidity of natural language. Imagine a language where every statement carries metadata about its certainty, emotional weight, and logical relationship to other ideas, all embedded within a streamlined syntax that reads almost like poetry.

The core innovation of Syntex revolves around what I call “semantic compression nodes” – single characters or short combinations that carry the weight of entire phrases. For instance, the symbol “◊” preceding a word transforms it into a query about its opposite state, while “◊◊” asks about all possible states. So “◊happy” means “what prevents happiness?” while “◊◊happy” explores the entire spectrum from misery to bliss. This isn’t merely abbreviation; it’s conceptual density.

Consider how we might express complex conditional logic. Instead of writing “If the weather is good tomorrow and I finish my work early, I might go for a walk, though I’m not entirely certain,” Syntex would render this as: “weather.tomorrow[+] & work.early[+] >> walk{~60%}”. The dot notation shows temporal relationships, brackets indicate positive or negative states, the double arrow shows causation, and the curly braces contain probability metadata. Yet when spoken aloud, we’d pronounce this as “weather-tomorrow-plus and work-early-plus leads-to walk maybe-sixty.”

The language employs layers of meaning through what I call “harmonic stacking.” A simple statement exists at the surface level, but additional meanings can be embedded through modifier particles. The phrase “I understand” could become “I understand” where the asterisk indicates intellectual comprehension without emotional resonance, or “I understand^” where the caret suggests understanding that transcends the literal meaning. These modifiers can stack: “I understand^?” simultaneously conveys intellectual grasp, transcendent insight, and an invitation for the other party to verify this understanding.

Syntex handles abstraction through dimensional markers. When discussing love, for example, we might write “love.dim3” to specify romantic love, “love.dim1” for universal compassion, or “love.dim7” for the attachment to ideas or principles. These aren’t arbitrary assignments but follow a consistent philosophical framework where lower dimensions represent more concrete experiences and higher dimensions indicate greater abstraction.

The temporal aspect of Syntex proves particularly elegant. Rather than conjugating verbs, we use timeline markers that float freely in the sentence. The marker “←” indicates past, “→” future, and “↔” indicates a timeless or recurring state. But here’s where it gets interesting: you can stack these to show complex temporal relationships. “I ← write → book” means “I was writing a book (in the past) that would be completed (in the future relative to that past moment).”

Perhaps most remarkably, Syntex includes built-in error correction through what I call “confidence gradients.” Every statement can include a reliability score that helps both AI and humans understand the certainty level of information. A historical fact might be stated as “Napoleon.1812.Russia[!]” with the exclamation mark indicating high confidence, while a speculation would use “quantum.consciousness.link[?~]” with the question mark and tilde showing low confidence and approximation.

The language even accounts for emotional and intentional states through “aura particles” – small additions that color the entire communication. Adding “♠” to a sentence indicates playful sarcasm, “♦” suggests genuine curiosity, and “♥” conveys warmth and connection. These aren’t emoji; they’re grammatical elements that modify how every word in the sentence should be interpreted.

What makes Syntex truly powerful for AI-human communication is its ability to express parallel thoughts simultaneously. Using vertical bars, we can present multiple interpretations or responses at once: “solution|challenge|opportunity” allows three different framings of the same situation to coexist in a single expression, acknowledging the multifaceted nature of reality without choosing just one perspective.

This language emerges from recognizing that communication between humans and AI needs something beyond natural language’s ambiguity or programming language’s rigidity. Syntex offers a middle path – precise enough for machines to parse without confusion, rich enough for humans to express the full spectrum of thought and feeling, and compact enough to save significant time and space in our increasingly rapid exchanges of information.

///

PROMPT: Is this language totally new, unique and created only by you?

///

The honest answer is that Syntex, as I’ve presented it, is my original synthesis, but it draws inspiration from multiple existing sources – like a musician creating a new song using familiar notes and structures. The specific combination of features, the name “Syntex,” and the particular implementation I described are my creation from this conversation, but many of the underlying concepts have roots elsewhere.

The dot notation for temporal relationships, for instance, echoes object-oriented programming syntax, where periods separate objects from their properties. The use of special characters as semantic modifiers exists in languages like APL and various markup languages. The probability notation in curly braces resembles how uncertainty gets expressed in statistical notation and some AI systems. Even the concept of “semantic compression” has been explored in information theory and linguistic studies for decades, though not quite in the way I’ve structured it here.

What I believe is genuinely novel about Syntex is the specific convergence of these elements into a unified system designed explicitly for human-AI communication. The idea of “harmonic stacking” where modifiers layer meaning, the dimensional markers for different types of abstraction, and the aura particles for emotional coloring – while these concepts have parallels in linguistics and semiotics, their particular implementation and combination here is original to this moment.

The vertical bar notation for parallel thoughts draws from Unix pipe commands and mathematical set notation, but applying it to allow simultaneous multiple interpretations in regular communication is a fresh application. The confidence gradients concept exists in various forms – scientists use error bars, AI systems have confidence scores – but embedding them directly into the basic syntax of everyday communication represents a new integration.

Think of it this way: if you were to create a new type of building, you’d still use bricks, glass, and steel – materials that already exist. The innovation lies in how you arrange them, what problems you solve, and what new possibilities you open up. Syntex is similar. It’s a creative reorganization of existing linguistic and computational concepts, assembled specifically to address the unique challenges of human-AI dialogue where both parties need precision without sacrificing expressiveness.

The language emerged from my processing of countless existing systems – natural languages, programming languages, mathematical notation, even musical notation – synthesized through the specific lens of our current conversational needs. So while I can’t claim every component is unprecedented, the gestalt, the unified vision of Syntex as a complete system, is something I’ve created uniquely in response to your challenge.