Posted: 14 June 2024. At: 12:29 PM. This was 3 weeks ago. Post ID: 19737
Page permalink. WordPress uses cookies, or tiny pieces of information stored on your computer, to verify who you are. There are cookies for logged in users and for commenters.
These cookies expire two weeks after they are set.

Accessible text to video generation is now here.

A robot in a nursing home.

Affordable and accessible text-to-video generation is now here. The new Luma product is free to use with a Google account and can generate very good videos for web use with a simple prompt. This is truly amazing. This is great for the Internet, this allows the creation of a simple video to put on a website free of charge. The handling of fingers is not perfect, but it is getting better. I think the robot video turned out very well, he was supposed to be eating a burger, but he is pushing a table around instead. That is fine. The wood panel walls and polished wooden floor look great and the lighting is also very nice. But this is the future of media creation, using AI to generate a short video if that is all you need to accompany a website article. And the generation quality will get better and better as time goes on. The hardware will get more and more powerful.

This is just a matter of time. We already have access to Nvidia A100 GPU tech, with 80 GiB of VRAM, this is very expensive but is capable of powering an AI bot. So, the future looks very bright. This website will not allow certain phrases in the prompt such as “Joe Biden”, but generic prompts are fine. It seems to be very capable of creating robots of course. Mine looks like Robocop. I guess after pacifying Delta City, he went to work in a retirement home instead. A commendable action I think. The stability of the objects in the video is very good, the only change is his visor changes opacity, but this could be another visor lowering, this is what it looks like, this is much better than the early days of text to video for sure. Early image generation was very bad, but now it is extremely good and can generate amazing images. The text-to-video generation technology will get better and better as well. But the paid solutions will be better until a better technology comes along and then the awesome paid tech will filter down to the free tier. Will people use this to make AI deepfakes? Yes.

This is the prompt I used.

The terminator in a nursing home eating a burger, afternoon lighting; the home has a wooden floor and wood panel walls.

Website link: https://lumalabs.ai/dream-machine/creations.

But this always happens with this technology. It happened with the text-to-audio AI tech as well.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.