Google's latest artificial intelligence model, Gemini, recently made headlines with a tech demo that showcased its advanced interaction capabilities. The six-minute video, released on December 7 and attracting over 2.1 million views on YouTube, depicted Gemini seamlessly interacting with a human operator.
The AI model was shown analyzing a duck drawing, interpreting hand gestures, and even creating a game called “Guess the Country” from an image of the world map. The demo initially received widespread acclaim for its seemingly real-time interaction and human-like understanding.
However, the admiration soon turned to skepticism as critics began to question the authenticity of the demo. Oriol Vinyals, an executive at Google DeepMind, clarified that while the interactions shown in the video were real, they had been edited for brevity.
Contrary to what the video suggested, Gemini's interactions were text-based rather than voiced and took significantly longer than portrayed. Google acknowledged this editing in a disclaimer on the YouTube video, stating, "For the purposes of this demo, latency has been reduced and Gemini outputs have been shortened for brevity."
NEW – Google's Stock Sinks Following Reports That Some of Their Gemini AI Demo Was Faked
"So that video demo of Gemini that everyone was kind of going crazy over...It was edited, it was cut to look like it was faster and more capable than it actually is...So I've been pressing… pic.twitter.com/9f7UmdLOlA — Chief Nerd (@TheChiefNerd) December 8, 2023
Criticism and Internal Concerns
This revelation sparked a flood of criticism on social media, with some accusing Google of misrepresenting Gemini’s capabilities.
Software developer Nelly R Q expressed disappointment in an X (formerly Twitter) post, accusing Google of lying about the AI demo. Another software engineer, Chief Nerd, criticized the editing, suggesting it made Gemini appear faster and more capable than it is in reality.
Even within Google, there are reports of discontent about how the video was presented. A Google employee, speaking to Bloomberg, expressed concerns that the video painted an unrealistic picture of Gemini, highlighting the ease with which AI tools can be made to seem more advanced than they are.
Interestingly, Google CEO Sundar Pichai did not mention the finetuning that went into the demo in his post about the release. Despite the controversy, the initial release of the Gemini tech demo was met with enthusiasm and awe.
Data analyst Armand Domalewski praised the AI’s human-like understanding and inference capabilities.
Seeing some qs on what Gemini *is* (beyond the zodiac :). Best way to understand Gemini’s underlying amazing capabilities is to see them in action, take a look ⬇️ pic.twitter.com/OiCZSsOnCc — Sundar Pichai (@sundarpichai) December 6, 2023