Bitget App
Trade smarter
Buy cryptoMarketsTradeFuturesCopyBotsEarn

Google enhances Gemini AI with ‘Gems’ and built-in image generator

GrafaGrafa2024/08/29 02:40
By:Liezl Gambe

Google has rolled out a significant update to its AI model, Gemini, introducing customisable “Gems” and integrating the Imagen 3 image generator.

This update is part of Google's strategy to compete with OpenAI’s dominance in the AI market.

Gems, first announced during Google I/O in May, allow users to create customised AI assistants for various tasks, such as coding, career advice, or brainstorming ideas.

Google describes Gems as a tool that enables users to "create a team of experts" for specific projects, offering support in areas like event planning or crafting social media captions.

These Gems are available to users of Gemini Advanced, Business, and Enterprise tiers, and can remember detailed instructions to streamline repetitive tasks.

Unlike OpenAI's GPT feature, which allows broader sharing, Gems can be shared via links on social media and integrated into Google Docs or Gmail.

For users struggling to articulate the perfect prompt for their Gems, Gemini offers an AI-powered rewrite feature to fine-tune the instructions.

Imagen 3, the newly released image generator, is now directly integrated into Gemini, expanding its capabilities.

Despite its integration, Google has maintained strict safety guardrails, particularly regarding the generation of images involving people.

The initial launch of Imagen 3 was met with issues when it generated problematic images, leading Google to temporarily take the tool offline.

Google emphasized its cautious approach, stating that they conduct extensive testing and collaborate with independent experts to ensure safety and continuous improvement.

The company has implemented a Prohibited Use Policy, banning responses that violate their guidelines, particularly in generating photorealistic images of identifiable people or inappropriate content.

Google touts Imagen 3 as setting "a new standard for image quality," with capabilities to generate images in various styles, such as photorealistic landscapes, textured oil paintings, or whimsical claymation scenes.

However, the ability to generate images of humans is still restricted, with the feature expected to be introduced in Gemini Advanced soon.

Google's spokesperson noted that while progress has been made in generating better images of people, the tool does not support creating photorealistic depictions of identifiable individuals, minors, or excessively violent or sexual content.

As with any generative AI, not every image produced by Gemini will be perfect, and Google plans to continue refining the tool based on user feedback.

Google aims to gradually roll out these features to more users and languages, ensuring a broader reach and improved user experience over time.

0

Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.

PoolX: Stake to earn
CEC, QTLX, GDV and other popular new coins are in hot progress!
Stake now!