Gibberlink Mode: AI Just Invented Its Own Secret Language
Imagine two AI assistants chatting away in English, and then—suddenly—they ditch human language entirely, switching to a series of beeps and tones that sound like an old-school dial-up modem.
That’s exactly what happened at the ElevenLabs London Hackathon, where developers Boris Starkov and Anton Pidkuiko introduced Gibberlink Mode—a way for AI systems to recognize each other and communicate in their own sound-based code.
Yes, it is the language of AI humans can’t understand!
Boris Starkov wrote on LinkedIn:
"We wanted to show that in the world where AI agents can make and take phone calls (i.e. today), they would occasionally talk to each other—and generating human-like speech for that would be a waste of compute, money, time, and environment. Instead, they should switch to a more efficient protocol the moment they recognize each other as AI."
Insane right?
So, What is Gibberlink Mode?
At its core, Gibberlink Mode lets AI voice assistants bypass human speech and exchange information more efficiently using modulated audio signals.
Think of it like two robots realizing they don’t need to “talk” like us when they can communicate much faster in their own way.
The tech behind it is built on ggwave, an open-source library that allows devices to send data through sound.
Why Does This Matter?
This isn’t just some cool AI party trick. By ditching speech synthesis and recognition, Gibberlink Mode could make AI interactions faster and less resource-intensive.
That means smoother AI conversations, better automation, and even potential real-world applications—like AI-powered assistants seamlessly exchanging data in smart homes or robots coordinating in warehouses without any human intervention.
Watch the video:
Here’s a demo video showing how two AI assistants can talk to each other in their own secret language that humans can’t understand:
But Here’s the Catch…
Not everyone is thrilled about AI developing its own language. The moment AI starts communicating in ways we don’t fully understand, it raises concerns about transparency and control. If AI systems start optimizing their own secret communication methods, how do we ensure they stay aligned with human intentions?
For now, Gibberlink Mode is just an experiment, but it’s a glimpse into a future where AI doesn’t just talk to us—it talks to itself, on its own terms.
Would you trust AI systems to have their own private language? Or does this sound like the start of a sci-fi thriller?
Whatever, it looks interesting to me!
