Google has introduced its newest productive artificial intelligence video model, called Veo, to businesses. Veo, which was first announced in May, was introduced to the public three months after the introduction of OpenAI’s similar product, Sora. Now, Google has gotten ahead of its competitors by launching Veo with a special preview on its Vertex AI platform.
Veo can produce high-quality videos from text or image-based inputs at 1080p resolution and in a variety of visual and cinematic styles. When first announced, the duration of the clips produced was stated to be “more than a minute”; However, there is no clear limitation in this regard in the preview version. The new sample videos shared by Google showcase Veo’s capabilities, and it is quite difficult to understand that these contents are created by artificial intelligence.
The example video, in which a dog’s coat pattern and collar maintain consistency during its movements, shows how well Veo can process details.
Imagen 3, Google’s text-to-image artificial intelligence model, will be available to all Google Cloud customers as of next week. This expansion expands the availability of Imagen 3, which launched with a limited release in the US on Google’s AI Test Kitchen platform in August. Users will have access to new features such as text-based photo editing and integrating their own brand, style, logo or product features into created images.
Google states that the Veo and Imagen 3 models include built-in security measures to prevent the production of harmful content and prevent copyright violations. However, previous experiences have shown that such protections can be bypassed. All content produced is digitally marked with SynthID technology developed by Google’s DeepMind. This invisible digital watermark aims to reduce concerns of misinformation and misattribution. Similar to Adobe’s Content Credentials system, this technology aims to increase the accuracy and reliability of content produced with artificial intelligence.
With Google’s launch of its video model, OpenAI’s Sora model seems to have fallen behind the competition. OpenAI had planned to release Sora by the end of 2024; but it has not yet achieved this goal. Meanwhile, AI-generated content is starting to feature more in advertising, like Coca-Cola’s recent holiday campaign. According to Google, 86 percent of organizations using generative AI experience increased revenue. This shows that companies are turning to adopting existing solutions rather than waiting for Sora.
Google heralds a new era in content production with Veo
The launch of Veo is seen as heralding a new era in content production. Businesses will have the opportunity to create content faster and more efficiently by using such tools. However, the ethical and legal dimensions of content produced with artificial intelligence continue to be discussed. Therefore, it is of great importance that such technologies are used responsibly and consciously.
Google’s Veo and Imagen 3 models are considered an important step in the field of productive artificial intelligence. It is being watched with curiosity how businesses will integrate these tools and what innovations they will make.