Google has just released its latest artificial intelligence (AI) model, Gemini, which the company claims possesses advanced “reasoning capabilities” that provide more thoughtful responses to complex queries. Developers acknowledge that AI content generators can sometimes produce false information, which they usually refer to as hallucinations.
With Gemini, Google says it underwent extensive testing across 57 subject areas, including mathematics and humanities, to evaluate its problem-solving skills and knowledge base. Sundar Pichai, Google’s CEO, referred to this development as the initiation of a “new era” for AI.
This is not a standalone product but rather a foundational model intended for integration into existing Google tools, including Search and Bard. For now, this will only be accessible in the United States.
This release follows Google’s cautious introduction of its AI chatbot, Bard, earlier in the year, labeled as an “experiment” due to a public demonstration error where it provided an incorrect answer about space.
READ ALSO: Meta’s New AI Translator Can Preserve Your Accent, Tone, and Voice
However, Google is making big claims about Gemini, describing it as the “most capable” model yet, stating it has the potential to surpass human experts in various intelligence tests. Gemini can both recognize and generate text, images, and audio.
While Google has faced challenges in competing with OpenAI’s widely used ChatGPT, the company claims that Gemini’s most potent version outperforms OpenAI’s GPT-4, the driving force behind ChatGPT, on 30 out of 32 commonly used academic benchmarks.
OpenAI plans to release a more powerful version of its software next year, with CEO Sam Altman expressing confidence that the new products will surpass the current ones significantly.
