- Bard is that talkative text interface, a product.
- Gemini is the LLM design that currently backs Bard (but also other Google AI products).
- Gemini "Basic", Advanced and Ultra are different sizes of that design.
This is conjecture, but "Ultra 1.0" probably indicates that they intend to release more models based on the Ultra configuration. Since that's the most commercial of theirs, I wouldn't be surprised if that comes with some stability promises (e.g. Ultra 1.0 is still available when Ultra 3.0 is released, so that if you do your own validation when integrating in your own project, you can expect small-to-no shifts in the underlying model)
Gemini Models gemini.google.com
------------------------------------
Gemini Nano
Gemini Pro -> Gemini (free)
Gemini Ultra -> Gemini Advanced ($20/month)Given that it's google. I would doubt it.
Ask how the original palm models are going.
They just need to curate their data but I wouldn’t be surprised if their pile is as large as OpenAiz
- Bard: Retiring this name - Gemini: model name (honestly less confusing than just calling it "GPT") - Gemini Advanced: More capable gemini model - Gemini Ultra: Most capable gemini model - Gemini 1.0: They version their models together, gemini has hit 1.0 ad is (supposedly) ready for prime time
Gemini is the name of the model and the service.
Gemini Advanced is the service with access to Gemini Ultra.
Pretty hilarious thinking they can rival ChatGPT pricing with a product that doesn't approach it's capabilities.
Bard was infuriatingly bad too, but more on a subjective level. And they correctly changed it, thank god. At least it’s easy to pronounce.
Software engineers have a weird obsession with Latin, Greek gods etc. Sounds smart and epic I guess. Personally I would have preferred “Steve French”.
I don’t understand why people find this so confusing. Are we not computer people?
/s