Google just unveiled its most advanced AI yet – an all-powerful brain called Gemini that could fundamentally transform how you use Google products.
Gemini comes in three sizes tailored to different applications. There’s Ultra for complex data center tasks, Nano for on-device smarts, and Pro for general consumer services like search and Pixel phones. All three handle various inputs like text, images, audio, and video.
Google says Gemini underwent rigorous safety testing to prevent harmful outputs, which is challenging with its multimodal design. Two benign inputs combined can potentially produce something dangerous.
The news comes right after reports claimed Google delayed Gemini’s launch over language translation issues. But the tech giant seems to have accelerated its timeline to counter ChatGPT hype, despite Gemini only supporting English initially.
So what does Gemini mean for you? If you have a Pixel 8 Pro, its new on-device Nano version powers features like smart message suggestions in Gboard and audio summarization in Recorder.
The Google Bard chatbot also got a Gemini upgrade for more advanced conversational abilities. And Gemini will spread to Search, Ads, Chrome, and more Google services over 2023.
Google made some bold claims about Gemini’s capabilities too, saying its Ultra version bests state-of-the-art benchmarks and blows past ChatGPT’s foundational GPT-4 in almost every area – text processing, image tasks, etc.
Of course, real-world testing will determine if Gemini truly exceeds the popular ChatGPT. But clearly Google is funneling huge AI muscle into its consumer products to take on OpenAI and Microsoft.
Between augmented on-device intelligence and souped-up web services, Gemini could make interacting with Google feel downright magical compared to today. And that would pose big trouble for Google’s AI rivals. The race to deliver practical consumer AI continues to heat up.
