Imagine if two ais could chat with each other in a language, no human could understand. RIGHT. Now go hide under the covers.
If you have called customer service in the past year, you have probably discussed with an AI. In fact, the first manifestations of powerful models of large languages showed how these AIS could easily deceive human appellants. There are now so many AI chatbots that manage customer service that two of them are required to compose, and now, if they do, they can do it in their own special sound language.
The developers of the Elevenlabs 2025 hackathon recently demonstrated GibberLink. Here’s how it works, according to a demonstration they provided on YouTube.
Two agents of Elevenlabs AI (we called them the best Synthesis startup of speech) call each other a hotel reservation. When they realize that they are both AI assistants, they switch to a high -speed audio communication called GGWAVE. According to an article on Reddit, Ggwave is “a communication protocol that allows data transmission via sound waves”.
In the video, the audio tones that replace the words spoken sound a bit like old -fashioned fashion handle protocols.
It is difficult to say if Ggwave and GibberLink are faster than speech, but the developers claim that GGWAVE is cheaper because it is no longer based on the GPU to interpret the discourse and can rather count on the CPU less strong intensity of resources.
The group shared its code on Github in case someone wanted to try to build this communication protocol for their own discussion chatbots on AI.
As these were Elevenlabs AI agents, nothing indicates that Gibberlink would work with Chatgpt or Google Gemini, although I will soon try to try similar Ggwave with these AI chatbots and others Generative.
To watch
What do they say?!
A pair of artificial intelligence assistants “speaking” their unintelligible language resembles a disaster recipe. Who knows what these chatbots could do? After finishing reserving this hotel room, what happened if they decided to empty the user’s bank account, then use the funds to buy another computer to add a third “voice” GGWAVE to the mix?
In the end, this is a cool technological demonstration that does not have much goal beyond proving that it can be done. However, he managed to make people a little nervous.
