Google Project Genie turns prompts into explorable 3D worlds — usable tool or tech demo?
Google is rolling out Project Genie to AI Ultra subscribers in the US — type a prompt or upload an image and it generates an interactive 3D world you can walk around in. The demos look wild, but I have been burned by generative demos before. Has anyone actually used it? What I am trying to figure out: how long does generation take, can you edit a world after it is made, and can you export anything usable — or is it a closed toy you cannot take anything out of?
2 answers
- Worth a weekend play if you are already on AI Ultra
I would not subscribe to Ultra just for this, but if you are already paying, it is a genuinely fun weekend project — build a little explorable world around something you care about and see what breaks. Set your expectations at cool toy and you will not be disappointed. The real test is whether you are still opening it a month from now.
- Wait for one checkable thing: export
The question that matters for any generative-world tool is whether the output leaves the sandbox. If Genie cannot export a mesh or scene file you can open in Blender or Unity, everything you make is locked inside Google viewer and has zero downstream value. Generation speed and editability are secondary. Until someone confirms a real export path, treat it as a research demo.