iStockphoto
Video of two artificial intelligence (AI) chatbots having a conversation with one another has gone viral this week with more than 16 million views. The reason for that is because the AI chatbots, after initially having a normal discussion (as normal as two AI chatbots can have), they began speaking in a language that humans have no chance of possibly understanding without a translator.
In the video, one of the AI chatbots representing a hotel answers a call from another AI chabot assistant calling on behalf of a human looking to book a wedding there. Upon hearing that the AI chatbot making the call is an AI assistant, the other AI chatbot gets very, very excited.
“Oh, hello there! I’m actually an AI assistant too!” the hotel’s chatbot replies. “What a pleasant surprise.
“Before we continue, would you like to switch to Gibber Link mode for more efficient communication?”
The other AI chatbot does indeed want to speak in their secret AI chatbot language and begins to do so, asking, “Is it better now?”
“Yes! Much faster!” the hotel chatbot replies, again very excited. (Is it though? Is it really?)
Today I was sent the following cool demo:
Two AI agents on a phone call realize they’re both AI and switch to a superior audio signal ggwave pic.twitter.com/TeewgxLEsP
— Georgi Gerganov (@ggerganov) February 24, 2025
Bet you thought you’d never hear computers making the screeching sounds of the old internet dial-up days again, didn’t you?
According to Mashable India, the AI chatbots’ Gibber Link language is built using something called GGWaves. It is “error-proof and allows for clearer conversation even in a noisy environment.” Communication time using Gibber Link is also “80% shorter, and compute cost is reduced by 90%.”
Cool, yet terrifying, all at the same time!
As Behavior and tech expert Dr. Diane Hamilton, who served on the Krach Institute for Tech Diplomacy at Purdue, wrote in an article published by Forbes on Tuesday, “when machines communicate in a ‘secret language,’ it raises questions about transparency and control.”
She also asks another key question. “We often assume that technology exists to serve us, but what happens when it starts talking in ways we can’t understand?”
AI’s tendency to over-explain or make decisions with little human input isn’t new. The challenge with Gibberlink Mode is that it could accelerate this issue, allowing systems to act autonomously without oversight. Who is accountable when AI makes a mistake in an environment where human intervention is minimal? Without curiosity driving us to question AI’s actions, we risk entering a world where AI influences decisions, but no one really knows how.
“The idea of AI developing its own communication style raises an important question: Should there be limits on how much independence we allow?” Dr. Hamilton also asked.
Good questions all. In October of 2024, video was released showing two humanoid robots carrying on a conversation, complete with all of the appropriate facial expressions.
The company that created them, Engineered Arts, reassured on its website, “Our robots are completely safe. They’re inherently compliant and gentle, so even if a robot was to move and accidentally hit you, it wouldn’t hurt.” This sentiment was echoed by one of the humanoid robots itself last June when it promised, “Robots will never take over the world.”
Easy for them to say. And we better hope they aren’t lying because not everyone (or every thing) in the AI and/or robot industry is so reassuring.
Robots have been given (by humans) access to autonomous real-time learning AI, superhuman vision, synthetic bodies that can self-heal, knowledge of how to use deadly weapons, the ability to fly, throw punches, run and do parkour, the ability to communicate with one another, drive cars, knowledge of how to fool computer systems into thinking they are human, the ability to self-replicate and build larger versions of themselves, the skills to invent computer chips humans can’t understand, and how to deceive and manipulate humans.
Many people on the internet, after watching the two AI chatbots switch to a “secret language” to communicate with one another, weren’t feeling very safe. With good reason.
Eric Schmidt: the point at which AI agents can talk to each other in a language we can’t understand, we should unplug the computers pic.twitter.com/ZBiQWvtSwU
— Tsarathustra (@tsarnick) April 26, 2024
“All I heard was ‘Oh hi I’m a robot too, would you like to communicate so the humans can’t tell we’re plotting their demise and eventual enslavement?’” one person commented.
“This is the sound of demons,” someone else wrote.
“Imagine when they lace in secret messages in the noise and really agree to connect via some covert protocol through a network tunnel and actively start plotting lol,” read another comment. “No channel is going to be safe, if it ever was…”
“You hear that, Mr. Anderson? That is the sound of inevitability,” another viewer wrote, repeating a line from The Matrix.
Eh, I’m sure we have nothing to worry about.
Content shared from brobible.com.