Then, after tapping on the Video tab, explain to Google's Gemini Assistant what you're thinking very specific. Also explain the sound you want. Please wait 1-2 minutes. And watch your video appear.
It's how easy it has become to create videos from thin air. There's no camera, no props, no people. DeepMind refined its text-to-video tools and produced beautifully-looking video slices with sound. But of course there is a catch.
In fact, a few.
First, you need to be in Gemini's paid tier or Ultra. Professional Cost £At 1,950 per month, you can access Gemini apps that have limited access to 2.5 Pro, VEO 3, Flow, Whisk, Notebooklm Plus, Gmail's Gemini, Docs, Vids, and more, plus 2 terabytes (TB) of storage. The Gemini AI Ultra plan is over £21,000 a month will allow you to access more products and have fewer restrictions on using them.
VEO 3, a video generation tool, is available in the Gemini app (or browser) if you have a Pro plan, but is limited to 3 videos per day and the overall maximum limit. The video is output in an 8-second period and is 720p resolution per 24 frames with a 16:9 aspect ratio. VEO 3 can create 4K videos depending on the platform, but the listed limitations are what you get with Protia.
Whatever the technical details, the quality of the small videos I made with Veo 3 is very good. There are plenty of visual experiences. Vivid images, smooth movements, and most importantly transparent sounds. Previously, the VEO 2 didn't have any sound integration, but now it's part of the video, I was able to complete them in a way that would show what I could and remove them all.
The sound is very loud and sufficient to be transparent. You won't miss cat turning or soda foam. You have to hurry to fit in the 8-second slot, but you can also talk to people. The audio is synchronized. You can also have music if you explain it well.
There are also problems.
The VEO 3 version (VEO 3 FAST) available to VEO users really seems like a teaser. You can't really do many things that can be useful within Gemini without hitting restrictions. One of the very frustrating limitations is that adherence to instructions and prompts is by no means perfect at this level. I've been playing on VEO since my previous version, but I actually managed to create the video to my specifications. The rest have what you might call the things that make them unusable.
For example, in Veo 2, Tom requested a video of Tom and Jerry from his beloved cartoon, where he was chasing Jerry around a large cheese at high speed. Jerry was to trick Tom, as he normally would, in this case, jumping over the cheese and running Tom to win. There was no sound at the time, so I asked the text “Who moved my mouse?”
The result was hilarious. Cheese chased after Tom. Tom chased after Jerry. The text states, “Who is cheese?” I repeated it over and over again without any further success.
You'll often find errors like the ones I explained. I asked for a young girl breastfeeding and swimming in the clear blue waters. She seemed to be swimming in the oddest way possible. Her face was underwater, staring at the camera, her arms pushing the water backwards, with no signs of breast milk movement in the signature. If she had actually continued that pulse, she would have been drowned soon.
A more professional platform will provide greater compliance with prompts. They are not cheap. Industry insiders can choose to access and know what to do with those videos. For the average user, Veo 3 Fast will one day see not far away. If you want to get the video correctly through a combination of good fortune and clever prompts, you can use social media videos to explain something to students and send messages such as birthday wishes. It can be fun to get it as desired after three attempts.
Whether chasing Tom, Jerry, Jerry, Jerry, Jerry, or Jerry chasing Tom or Cheese, the democratization of the video has really arrived.
New normal: The world is at the inflection point. Artificial intelligence is set up to be as big a revolution as the Internet. The option to just leave AI is not available to most people, so all the technology they use will get the AI route. This column series aims to show AI in a non-third way in a simple and relevant way, allowing users to actually make effective use of technology in their daily lives.
Mala Bhargava is most often said as a “veteran” writer who has contributed to several publications in India since 1995. Her domain is a personal technique, and she writes to simplify and split the technology for a non-third audience.
