AI continues to evolve rapidly, and with that, more and more new applications for this technology are emerging. A few years ago, we didn't even consider the possibility of generating large texts or having AI edit photos. In the meantime, we have become familiar with this technology, and many people use it daily.
Now, a new application has been released to the general public: AI-generated video. Fragments that are completely created by artificial intelligence.
What is AI generated video?
Both OpenAI (known for ChatGPT) and Google have made their video generation AI available to the public. OpenAI has 'Sora' and Google has 'VEO', platforms where you can generate videos based on text. The videos are still limited to a few seconds, so you can't yet have a Hollywood film made by AI.
We would love to give you a demo, but the models have not yet been launched in Belgium. However, we can show examples available on the websites of OpenAI and Google. Keep in mind, all the videos in this blog post are completely AI-generated.
OpenAI Sora video model. The input text for this video is: Storyboard: A group of six workers, dressed in dark, worn clothing and carrying tools and backpacks, are seen slowly walking on a desolate, fog-covered beach.
The differences between the two video models
Since these video models were released, they have been improved several times. There are still clear differences between the two models. For example, Google's model, Veo, seems to be better at understanding physics and movement.
People move more naturally, interactions between objects look more realistic, and movement is more convincing.
Google Veo videomodel. The prompt used to get this video: A lone cowboy rides his horse across an open plain at beautiful sunset, soft light, warm colors
What will this technology be used for?
The models are not perfect yet, so the chances of them being used in professional environments at this moment are small. However, we have already seen many short generated clips on YouTube, for example, to give an idea extra impact or to represent something visually in a short format.
These models are being improved quickly, so we expect that they will be used in professional environments in the near future. Think of marketing campaigns, video production for social media, and perhaps even in large film productions.
Curious about a demo of these models?
Discover our recent posts
Hoe AI-tools zoals ChatGPT het leven van ondernemers makkelijker maken
EN 
NL