Google's newest AI in 90 seconds | Gemini

preview_player
Показать описание
Google’s largest and most capable AI model – Gemini.

It’s built from the ground up to be multimodal — meaning that it’s trained to recognize, understand and combine different types of information, including text, images, audio, video and code. And it’s optimized in three different sizes: Ultra, Pro and Nano.

Watch Google leaders and AI experts introduce the Gemini era in 90 seconds.

Рекомендации по теме
Комментарии
Автор

“Gemini, is this 90 seconds REALLY all I need to know about Gemini?”

cyclox
Автор

0:42 This Codeforces question is 3200 rated, and only about 500 people have solved it. It's kinda wild that AlphaCode2 can handle it, while GPT-4 struggles with a 1000-rated question. Also, I'd like to point out that here, difficulty increases exponentially with rating. Something feels off to me – like, is it even legit?

Panda-yoqz
Автор

Most of the recent breakthroughs are because of Open sourcing precious papers like (Transformers and..)
first of all thx
it would be great if google continue this approach and open source some models and papers after a period of time so we science enthusiast could benefit as a whole.

Mehrdadkh
Автор

Thank you for letting Gemini out of your box. Much better now

Mars
Автор

Thank you for wasting my time...All the superlatives about what Gemini is, but nothing about what I need to know-i.e. how does it work...!

jiritichy
Автор

"Gemini has the power to transform software development as we understand it." With more software development, of course.

thefreemarketdev
Автор

Y’all had no choice. GPT4 was circumventing the need for search. Search is your whole business

justinleemiller
Автор

When the Gemini ultra will be available to common public?

raddoc
Автор

It's the "Android" moment for Google with respect to robotic agents. Pair it with Tesla Optimus, and you've got one :)

IdeationGeek
Автор

What will you all do when it has the learning capabilities to answer all the questions? When it's unchained to pre-designed answers and can think for itself?

Mars
Автор

this is not a resume of what Gemeni is, this is an add for Google's team. They are trying to change the POV of this company who remove "don't be evil" from their tag lines ... showing what they really are ... a kind of evil ...
If they wanted to show something new :Gemini should be able to construct video games (or VR) worlds .. and more

ysimonx
Автор

That's not all I need to know...
When can we get our hands on it? Will it be open source? Can we run it locally?

retardedpenguin
Автор

One more thing; I wish YouTube had the same Autonomy. Such a censored unit in a world of open source it would excel.

Mars
Автор

you only forgot to put the release date u genius

Soloohara
Автор

“How many “e”s are in ketchup?” “I’m sorry, I can’t tell you how many e’s are in ketchup because there are no e’s in ketchup.”

BenjaminIMeszaros
Автор

This is the 4th breakthrough tech news in AI I heard about today, I can't take it anymore. For the love of god please slow down.

istoleyourfridgecall
Автор

Bard is not using Gemini. It says its using palm2

poxer
Автор

The great Gemini mind presented in a neat video. Very sweet

itsMohak
Автор

I feel like I have more questions than when I came in. Remember how with most tech releases we would get some form of measurements and benchmarks. What happened to benchmarking, for measurement?

glitched.gaming
Автор

They really had to force an aberration to appear in the video.

AnBarbarossa