Google Launches Veo on Vertex AI for Video Creation | Ai image generator online | Borisdayma dalle mini github | Text to image generator | Turtles AI

Google’s Veo video generator, which creates short video clips from images and prompts, is now available in private preview for Google Cloud customers, through the Vertex AI platform. The announcement also includes integrations with partners like Quora and Mondelez International, with potential creative and marketing applications. Despite some limitations in AI, Veo could revolutionize video content creation.

Key Points:

• Veo is an AI model that generates short video clips, compatible with Google Cloud’s Vertex AI. 
• Its early availability allows customers like Quora and Mondelez to integrate it into their platforms. 
• Despite some imperfections, the model offers a lot of potential for creating personalized visual content. 
• Google uses protections like watermarking to address the risks of deepfakes.

Google’s Veo video generator, which uses AI to create short video clips from images and prompts, was recently announced as a private preview for Google Cloud customers. The advanced tool will be accessible through the Vertex AI platform, giving companies the ability to experiment and integrate the technology into their offerings. Among the customers already involved, Quora will use Veo to enrich its Poe chatbot platform, while Oreo giant Mondelez International will apply the video generator to create marketing content with its agency partners. In this way, Google aims to expand the creative capabilities and potential of generative AI applications. Veo, officially launched in April, is designed to produce high-definition video clips, up to six seconds long, at 1080p resolution, and reaching a fluidity of 24 or 30 frames per second. The model can reproduce various visual and cinematic styles, such as landscape shots, time-lapses, and action scenes, while also offering the ability to modify already generated clips, such as adjusting visual details or fluid dynamics. These capabilities put Veo in direct competition with other video generation models, such as OpenAI’s Sora and those from other companies such as Adobe, Runway, and Meta. While not perfect, Veo offers a wide range of capabilities, but it is not without its challenges: some generated videos have inconsistencies, such as objects disappearing and reappearing without credible explanation, or errors in the laws of physics, such as cars suddenly reversing for no apparent reason.

The launch of the model to Google Cloud customers was accompanied by explanations of “enterprise readiness” and the continued improvement of Veo’s capabilities to meet the needs of enterprises. Google’s Vertex AI platform was developed to enable enterprises to fully exploit the potential of advanced AI models like Veo, and this latest update demonstrates an ongoing commitment to expanding Google’s offerings in the visual content generation space. While the model is already powerful enough to support the creation of high-definition video, Google plans to refine it further, promising continued improvements, as it has with other models such as Gemini. Among Veo’s more advanced features is the ability to make specific changes to portions of videos via masked editing, as well as the ability to chain videos together for more complex projects. However, rights management over the training data remains a sensitive issue: Google has stated that Veo may have been trained on YouTube content, but in compliance with the platform’s terms of use.

A key issue is the legal implications and copyright protection. Veo was trained on a large dataset, but Google has not provided specific details about where the content used comes from. While it has introduced filters to avoid violent or explicit content, the risk of copyright infringement remains a hot topic. In this regard, Google has reassured Veo users, promising legal protection for those who will use the video generator within Vertex AI, thanks to a compensation policy in the event of copyright disputes.

With the growing interest in automatic video content creation, Google is ramping up the deployment of Veo across various platforms and services. After bringing it to Google Labs for select tests in May and announcing integration with YouTube Shorts in September, the company seems to want to make Veo a powerful and versatile tool for content creators. In this context, the ability to leverage SynthID watermarking technology to label generated content becomes essential, to address the risks associated with deepfakes and protect the integrity of the videos created. However, as demonstrated by the competition from other AI companies, such as Runway, which has already signed deals with movie studios, and OpenAI, which has worked with brands and filmmakers, the race to innovate in the field of generative AI is particularly intense. Despite this, Google does not seem to intend to accelerate too much, preferring a gradual growth, as underlined by the director of product Warren Barkley, who explained that the company prefers to get feedback from a select group of customers before making the platform available to a wider audience.

Veo’s potential to transform video content creation comes at a cost to concerns about its impact on the creative industry. Generative AI could disrupt the film, television, and animation industries, creating a conflict between technological innovation and preserving traditional jobs in the arts.