In a strategic move to challenge OpenAI’s stronghold in the artificial intelligence arena, Google has introduced a substantial update to its leading AI model, Gemini. Announced on Wednesday, the update features the launch of customizable “Gems,” signaling a bold step in the company’s push to diversify its AI offerings.
This new feature, Gems, brings the flexibility to create a tailored array of AI assistants based on the Gemini model. These assistants, designed to perform a variety of tasks from coding assistance to providing career guidance, echo the modular capabilities of OpenAI’s GPT feature in ChatGPT. As of now, Gems are accessible to users of Gemini Advanced, Business, and Enterprise tiers, providing them with a potent new tool for productivity and creativity.
Google introduced two new Gemini AI features: “Gems” for customizing tasks like learning, writing, and coding, and “Imagen 3” for creating hyper-realistic images. Free users can try Imagen 3, but Advanced users get it first. #AI #Gemini #GoogleAI #Tech #GoogleGemini pic.twitter.com/vee5TN7CWw
— Kshitiz Jangra (@kshitieeee) August 29, 2024
“With Gems, you can assemble a team of specialized AI experts to tackle complex projects, generate innovative ideas for events, or craft the ideal social media caption,” Google stated in their announcement. “Your Gem can also retain a comprehensive set of instructions, streamlining repetitive, intricate, or time-consuming tasks.”
For those who might struggle with the exact language required to configure their Gems, Gemini includes an AI-driven rewrite feature that refines prompts, ensuring the customized assistants perform as desired. These customized outputs can be easily shared on social media, or integrated into Google Docs and Gmail, enhancing collaborative capabilities across platforms.
However, Google has clarified that, for now, these Gems are intended for personal use. A representative from Google explained, “You can share conversations you’ve had with Gems by creating a public link, but the shared chats cannot be continued by others with whom you share the link.” This marks a distinct difference from OpenAI’s GPTs, which allow more extensive sharing and collaboration on customized functionalities.
Over the coming days, start creating and chatting with Gems: customizable versions of Gemini that act as topic experts. 🤝
We’re also launching premade Gems for different scenarios – including Learning coach to break down complex topics and Coding partner to level up your skills… pic.twitter.com/2Dk8NxtTCE
— Google DeepMind (@GoogleDeepMind) August 28, 2024
Additionally, the integration of the Imagen 3 image generator marks another significant enhancement to Gemini’s capabilities. Google remains cautious about its deployment, especially after a previous release produced controversial images, leading to the tool’s temporary removal. “We conduct thorough internal and external testing and work with independent experts to ensure continuous improvement,” a Google representative noted. “Our Prohibited Use Policy strictly prohibits responses that violate our guidelines.”
Imagen 3, initially unveiled in May, has now been launched following rigorous testing. This tool positions itself against established players like OpenAI’s Dall-E, Midjourney, and Flux, which is incorporated into Elon Musk’s Grok chatbot. Google has touted Imagen 3’s prowess, stating, “Imagen 3 establishes a new benchmark for image quality, capable of generating visuals from just a few words. Users can request images in diverse styles—from photorealistic landscapes to textured oil paintings or whimsical claymation scenes.”
While Gemini can create detailed images of animals and objects, it currently does not support the generation of human likenesses. If prompted to do so, the AI responds that this feature is “coming soon to Gemini Advanced.” Google elaborated, “We’ve made significant advancements in generating better user experiences when creating images of people, but we do not support generating photorealistic, identifiable individuals, depictions of minors, or images containing extreme violence, gore, or sexual content.”
🚀 Google’s Gemini AI is getting a major upgrade! 🌟 New features include improved image generation, customizable Gems, and enhanced safeguards. 🔐
#AI #Gemini #GoogleIO pic.twitter.com/ILAs4j0EwX
— Developer Nation Global Community (@devnationworld) August 29, 2024
Google acknowledges the imperfections inherent in generative AI technology. “Not every image Gemini generates will be flawless,” they admitted, “but we are committed to continuously improving based on feedback from our early users. Our goal is to gradually expand access to these capabilities, bringing them to more users and across different languages soon.”
Major Points
- Google launches customizable “Gems” for its AI model, Gemini, enabling personalized AI assistants for various tasks.
- Gemini’s AI-driven rewrite feature enhances prompt customization, supporting better user experience and creativity.
- New Imagen 3 image generator introduces advanced image creation capabilities but restricts the generation of human likenesses.
- Gemini’s updates currently focus on personal use, contrasting OpenAI’s more collaborative AI sharing functionalities.
- Google’s strategic advancements position Gemini as a formidable competitor in the AI industry, enhancing its AI offerings.
Kirk Volo – Reprinted with permission of Whatfinger News
