Google launches Gemini, its most-advanced AI model yet, as it races to compete with ChatGPT

 

In an attempt to compete in the quickly expanding field of generative artificial intelligence, Google on Wednesday unveiled its most ambitious project to date: Gemini, an AI model intended to outperform OpenAI's GPT models and enhance everything from Google's consumer apps to Android smartphones.

 

With the announcement of Gemini as its "largest and most capable AI model" and the proclamation of a "Gemini era" in which the tech giant envisions its model being used in every setting, from large companies to consumer devices like the Google Pixel 8 Pro, the announcement demonstrated the breadth of Google's ambitions.

 

Unlike existing AI models that typically deal with only one type of user prompt, such as exclusively images or text, Gemini was built to be “multimodal,” Google said. This means it accepts inputs that include multiple types of media, combining text, images, audio, video and programming code.

 

“This new era of models represents one of the biggest science and engineering efforts we’ve undertaken as a company,” said Google CEO Sundar Pichai in a blog post.

 

Google’s proprietary AI chatbot, Bard, has already been upgraded with a version of the Gemini model, the company said Wednesday, with plans to add Gemini to widely used products including Google’s search engine and Chrome web browser, which are used by billions of people worldwide.

The announcement is an attempt to recover the initiative after OpenAI's ChatGPT, which was released abruptly and to great public acclaim a year ago, seemingly caught Google and other tech giants off guard. This led to a rush to develop generative AI tools within the industry and a global conversation about the advantages and disadvantages of AI.

 

Additionally, it is an attempt to spread generative AI as far as possible across Google's empire. Gemini 1.0 is available in three different sizes, according to the company: Nano, which is optimized for mobile devices and app developers; Pro, which is the default model made for a variety of tasks and customers; and Ultra, which is Google's most advanced AI model to date, still undergoing safety testing.

 

Wednesday’s launch was also designed to showcase Google’s advances in cloud computing, a critical resource for AI developers. The company said it trained Gemini using a new generation of powerful cloud-based processors that can collectively train large AI models nearly three times faster than the prior version. That technology, which will also be made available to Google’s cloud customers, could mean a significant boost to the wider AI industry, making AI training more accessible and bolstering Google’s third-place position in the market for public cloud services. But it is unclear how Google’s AI chips stack up against those of leading chipmakers such as Nvidia.

 

In its testing, Google’s Gemini model outperformed rival AI models across more than two dozen benchmarks commonly used by AI researchers to evaluate an algorithm’s reading comprehension, mathematical ability, and multistep reasoning skills, the company said.

 

“We do see it setting new kinds of frontiers across the board,” Eli Collins, vice president of product at Google DeepMind, told reporters on a conference call Tuesday.

 

However, he seemed to accept the ongoing possibility that consumers may receive deceptive results from AI models, citing concerns expressed by researchers, legislators, and civil society organizations.

 

Large language models "are still capable of hallucinating," a phrase AI researchers use when AI systems make up facts and get things wrong – but with enormous confidence. Collins noted that Google has done "a lot of work on improving factuality in Gemini."

 

“When we integrate these models into products like Bard, we have additional techniques to improve the accuracy of responses,” he added.

 

In recognition of those risks, Google said Wednesday that Gemini Ultra, its most advanced version of the model, will only be released gradually to “select customers, developers, partners and safety and responsibility experts for early experimentation and feedback before rolling it out to developers and enterprise customers early next year.”

 

Gemini Ultra is currently undergoing third-party safety evaluations, also known as red-teaming, in accordance with a commitment it made to the Biden administration earlier this summer.

 

 Source: CNN