[BY]
Dmytro Kremeznyi
[Category]
AI
[DATE]
Dec 11, 2023
Google has just dropped a bombshell by unveiling its long-anticipated AI project, Gemini. This formidable AI model, residing within the Bard chatbot, stands as Google's latest innovation and...
Google has just dropped a bombshell by unveiling its long-anticipated AI project, Gemini. This formidable AI model, residing within the Bard chatbot, stands as Google's latest innovation and response to the rising ChatGPT wave.
Described as Google's "most capable" AI to date, Gemini sets itself apart by being trained on not just text but also video, images, and audio. This unique multimodal approach means Gemini is not your average AI—it's a multitasking maestro poised to reshape the way we interact with conversational AI.
With an initial launch in Bard for English users, Gemini's rollout spans across more than 170 countries. Developers can soon harness its power through Google Cloud's API, set for availability from December 13. A more compact version, optimized for Pixel 8 smartphones, already started, while future plans include integration into Google's search, ads, and Chrome.
<blockquote class="twitter-tweet" data-theme="dark"><p lang="en" dir="ltr">Gemini is our most flexible model yet, efficiently built to run on everything from mobile phones to data centers with three sizes: <br>- Gemini Ultra: largest model for highly-complex tasks <br>- Gemini Pro: best for scaling across a range of tasks<br>- Gemini Nano: for on-device tasks <a href="https://t.co/P68aLkoP4G">pic.twitter.com/P68aLkoP4G</a></p>— Google (@Google) <a href="https://twitter.com/Google/status/1732421911555031470?ref_src=twsrc%5Etfw">December 6, 2023</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Gemini's capabilities are categorized into Ultra, Nano, and Pro versions. Each version caters to different needs, from powering chatbots and generative searches to scientific research and coding assistance.
According to Google, Gemini's AI prowess extends beyond text-centric models like ChatGPT. Its inherent ability to handle multimodal data empowers Bard with improved reasoning, planning skills, and content summarization—considered Bard's most significant upgrade to date.
In demonstrations, Gemini showcased its visual acumen, effortlessly handling video responses and aiding scientific research with graph and equation interpretations. Notably, Gemini Pro, the initial model, outperformed ChatGPT's predecessor across AI smartness benchmarks.
While specifics about Gemini's inner workings remain a secret, its development isn't a hushed affair. The cost and effort involved in training such a large-scale AI model speak volumes about Google's commitment to AI dominance.
With Gemini's release, Google aims to set new benchmarks and regain its leadership position in the AI arena. As the company strategically revamps its AI efforts, Gemini stands tall as a testament to Google's ambition, merging its AI research arms to create this powerhouse AI model.
Content