AI Horizons: Meta and Gemini Unveil Future

On December 6th, the company previously recognized as Facebook unveiled a series of AI-centric enhancements for its social media and messaging platforms.

In a blog post, the company stated, "We're exploring over 20 innovative approaches through which generative AI can enhance your interactions on Facebook, Instagram, Messenger, and WhatsApp." This commitment to advancement includes a comprehensive upgrade to Meta AI, the company's chatbot, and virtual assistant. The improvements focus on augmenting its helpfulness, featuring more intricate responses and more precise summaries of search outcomes.

Today, we’re sharing updates to our core AI experiences and new capabilities you can discover across our family of apps. Read all about it here: https://t.co/9CjalsAVKE pic.twitter.com/uQ4mUwJCxi

— Meta (@Meta) December 6, 2023

Meta's Capability in Recognizing AI-Generated Images

Meta AI's capabilities extend to the identification of AI-generated imagery, with upgrades to Meta's Images and Reels platforms. Reels recommendations are seamlessly integrated into Meta AI chats to present visual information effectively. The company also launched a standalone AI image generator website called "Imagine with Meta AI."

Training the AI model involved using 1.1 billion publicly accessible Facebook and Instagram images. This model can generate unique images based on written prompts. Facebook users will experience AI working behind the scenes to aid in content creation and sharing, with integration into the Marketplace for buyer and seller assistance.

To enhance transparency and traceability, Meta plans to implement invisible watermarking in the Meta AI image experience in the coming weeks. This feature aims to identify images created with AI without being visible to users.

Despite the technical vagueness of the announcement, it did not disclose the specifics of how the AI model or the company would utilize user personal data. Concurrently, the Thai government raised concerns about Facebook's failure to safeguard users from scams, contemplating a potential ban on the platform.

Google Unveils its New AI Model, Gemini

In a separate development on December 6th, Google, the tech giant, introduced its highly anticipated AI model called Gemini. Google asserts that Gemini surpasses competitors like OpenAI's ChatGPT powered by GPT-4. Gemini distinguishes itself by being multimodal, capable of handling text, images, audio, and video, surpassing the capabilities of its counterparts.

Gemini is available in three versions or "sizes" tailored to specific uses and markets. It is seamlessly integrated into Google's AI assistant, Bard, and is accessible on its flagship smartphone. Google plans to deploy Gemini across various products and services, including Chrome, Ads, and Search, in the upcoming year.

You might also like: Crypto Rollercoaster: 24-Hour Market Whirlwind