Meta's video generator, GPT upgrade, Liquid AI models, Pika 1.5, Copilot Voice: AI News

Science & Technology

Introduction

AI technology is making remarkable strides every week, and this week is no exception. From innovative video generators to exciting upgrades in voice technology, the landscape of artificial intelligence continues to evolve at an incredible pace. Here, we'll delve into some of the most talked-about developments, including Meta's new video generator, OpenAI's advancements in voice features, cutting-edge liquid AI models, and much more.

1. Pika 1.5 Video Generator

Pika has recently unveiled its video generator version 1.5, claiming improved quality and better prompt adherence compared to earlier versions. Upon signing up for the service, users receive free credits to start generating videos. The platform allows for regular text-to-video creation where users can enter prompts, although many users have reported server issues that delay generation.

In addition to simple prompts, Pika introduced a novel feature called "Peak Effects," allowing users to upload images and apply fun effects like inflating, crushing, or transforming objects into cakes. This feature has drawn notable attention, providing an engaging way to play with images while waiting for the standard text-to-video function to be restored.

2. Meta's Movie Gen Video Generator

Meta has entered the AI video generation scene with its product, Movie Gen. This generator can create videos up to 16 seconds long in 1080p HD and offers support for various aspect ratios. Users can enter simple prompts and receive impressively realistic video outputs. What sets Movie Gen apart is its ability to generate videos of people from just one reference image and to create synchronized audio to accompany the video. Furthermore, it supports precise editing based on additional prompts, showcasing outstanding capabilities that could disrupt the visual effects industry.

3. OpenAI's New Voice Features

OpenAI recently upgraded ChatGPT's voice feature, now available for free users, which enhances the interactivity of the tool by allowing users to have more natural conversations. However, it's worth noting that free users may only access this capability for approximately 10 to 15 minutes per month. On the developer front, OpenAI introduced a real-time API for their advanced voice feature, which could revolutionize various sectors, including healthcare and education.

4. Liquid AI Models

The latest advancement in AI models comes from a startup called Liquid. They announced a new class of AI known as Liquid Foundation Models (LFMs), which leverage liquid neural networks designed for dynamic adaptability. These models demonstrate superior performance while being more memory efficient. The LFMs currently come in several sizes, outperforming traditional models while using significantly less computational power.

5. Microsoft’s Copilot Voice

Microsoft has rolled out its voice assistant known as Copilot Voice, which can be accessed through its website or downloaded as an app. This voice assistant features four unique voices and can perform various tasks, such as storytelling and brainstorming ideas. Although it lacks some advanced functionalities of competing voice technologies, its free availability opens up new avenues for users seeking voice assistance.

Conclusion

This week’s developments in AI technology not only spotlight advancements in video generation but also showcase significant upgrades in voice interactions and the potential of new model architectures. With so much happening, it will be fascinating to see what new innovations emerge in the coming weeks.

Keyword

  • AI technology
  • Video generator
  • Pika 1.5
  • Meta
  • Movie Gen
  • OpenAI
  • Advanced voice
  • Liquid AI models
  • Copilot Voice
  • Real-time API

FAQ

What is Pika 1.5?
Pika 1.5 is a video generator that offers enhanced quality and prompt adherence. It also features a “Peak Effects” option for creative image manipulation.

What capabilities does Movie Gen by Meta offer?
Movie Gen can create realistic videos from text prompts and single reference images. It also synchronizes audio effects tailored to the generated scenes.

How has OpenAI improved its voice functionalities?
OpenAI has introduced an upgraded voice feature within ChatGPT available for free users with limited monthly access, enhancing user interaction through more natural responses.

What are Liquid Foundation Models?
Liquid Foundation Models are a new AI architecture using liquid neural networks, which are designed to be adaptive and highly efficient on memory requirements compared to traditional models.

What does Microsoft’s Copilot Voice do?
Microsoft’s Copilot Voice is an assistant that supports various tasks, including storytelling and brainstorming ideas, available through its website and app.