Mark Zuckerberg, CEO of Meta Platforms, announced the creation of a high-level product suite focused on Meta products for turbocharging with generative AI.
AI teams from across the company will come together to focus on ways to implement generative AI for a more enjoyable experience.
Generative AI high level product suite
New AI for Meta High level product range It will develop AI characters that can help users of Meta Products in various ways. Examples given by Zuckerberg include the following.
- AI chat experiences in WhatsApp and Messenger.
- AI photo filters and ad formats for Instagram.
- AI video and multimedia experiences.
In the year 2022, Meta AI foot Make a video. This AI system allows users to create videos from a text message. You can read a file Research paper For more information and sign up to receive notifications about future tool releases.
These new features will give users a more personalized experience and new ways to express themselves. This will allow creators to produce better content faster to earn more income with it Facebook And Instagram Income programs.
They will also allow Meta to compete with social media platforms that already integrate AI.
Racing artificial intelligence features on social media
Snapchat has released its own AI persona, My AI, which allows users to be asked for recipe suggestions, gift ideas, and content inspiration. It runs from OpenAI’s ChatGPT.
While the feature is limited to Snapchat+ subscribers, for now it will be available to all users after an initial person test.
Instead of just turning their face into an animal or adding glowing layers, users can see themselves as they were in their teens or get a preview of what they might look like with a charming makeover.
YouTube’s new CEO, Neil Mohan, posted a message about priorities for 2023. In addition to new monetization opportunities, he noted that with the capabilities of generative AI, creators will be able to elevate their storytelling and production value.
LLaMA for researchers from Meta AI
In related news, Meta announced the release of LLaMA – A large Meta AI model for language in an effort to democratize access for researchers. It will be available for non-commercial license, specifically for research use cases.
Importantly, LLaMA will not require researchers to have the large number of resources normally needed to train and run large language models. Researchers can choose from 7B, 13B, 33B, and 65B parameter models, the smallest of which is trained on 1 trillion tokens.
This allows researchers to test new ways to reduce or eliminate the risks of bias, hallucinations, and toxic responses that users experience when interacting with AI.
They also called on everyone in the AI community—academic researchers, civil society, policymakers, and industry—to develop guidelines for responsible AI and big language models to ensure a better future.
Featured image: Sergei Elagin/Shutterstock