* AI ("Artificial Intelligence") subtitles on Tagalog.com are generated using "Whisper" by OpenAI (the same company that created ChatGPT and DallE2). Results and accuracy may vary.
* The subtitles do include errors occasionally and should only be used as a tool to help with your listening practice.
* You can request this website to create a transcript for a video if one doesn't already exist by clicking the "Request AI Subtitles" button below a video. Transcribing usually takes 30-40% of the length of a video to complete if there are no other videos in
the queue. For example, a 21 minute video will take 7-8 minutes to transcribe.
* Running a super fast cloud GPU server to do these transcriptions does cost money. If you have the desire and financial ability, consider
becoming a patron
to support these video transcriptions, and the other tools and apps built by Tagalog.com
00:00.2
Okay, let's do some tests.
00:03.6
Tell me when you see something that makes sound.
00:07.9
I see a speaker which makes sound.
00:12.2
What is that part of the speaker called?
00:16.6
That is the tweeter. It produces high-frequency sounds.
00:23.9
Give me a creative alliteration about these.
00:26.6
Creative crayons color cheerfully. They certainly craft colorful creations.
00:38.2
What does that part of the code do?
00:43.1
This code defines encryption and decryption functions.
00:47.5
It seems to use AESCBC encryption to encode and decode data based on a key and an initialization vector, IV.
00:56.6
Hmm. That's right.
01:00.8
What neighborhood do you think I'm in?
01:05.1
This appears to be the King's Cross area of London.
01:08.4
It is known for its railway station and transportation connections.
01:13.7
Do you remember where you saw my glasses?
01:18.3
Yes, I do. Your glasses were on the desk near a red apple.
01:26.6
Thank you very much.
01:29.2
The power of Gemini, with multi-modality, long context, and agents brings us closer to our ultimate goal,
01:38.2
making AI helpful for everyone.
01:41.8
We see this is how we will make the most progress against our mission.
01:46.1
Organizing the world's information across every input, making it accessible via any output,
01:52.7
and combining the world's information with the information in your world,
01:57.0
in a way that's truly useful for you.
02:00.8
Today, we have some exciting new progress to share about the future of AI assistance
02:05.5
that we're calling Project Astra.
02:13.4
For a long time, we've wanted to build a universal AI agent that can be truly helpful in everyday life.
02:20.0
Our work making this vision a reality goes back many years.
02:23.6
It's why we made Gemini multi-modal from the very beginning.
02:26.6
An agent like this has to understand and respond to our complex and dynamic world just like we do.
02:33.8
It would need to take in and remember what it sees so it can understand context and take action.
02:39.1
And it would have to be proactive, teachable, and personal so you can talk to it naturally without lag or delay.
02:47.2
Building on our Gemini model, we've developed agents that can process information faster
02:51.8
by continuously encoding video frames, combining the video and speech inputs,
02:56.6
and then integrating the data into a timeline of events, and caching this for efficient recall.
03:01.9
Today, I'm excited to announce our newest most capable generative video model called Veo.
03:12.9
Veo creates high quality 1080p videos from text, image, and video prompts.
03:19.0
It can capture the details of your instructions in different visual and cinematic styles.
03:23.7
You can prompt for things like aerial shots of a landscape, or text, or video.
03:24.8
It can capture the details of your instructions in different visual and cinematic styles.
03:25.6
You can prompt for things like aerial shots of a landscape, or text, or video.
03:26.6
timelapse, and further edit your videos
03:28.8
using additional prompts.
03:30.8
You can use Vio in our new
03:32.3
experimental tool called VideoFX.
03:35.2
We're exploring features
03:36.5
like storyboarding and generating longer
03:38.6
scenes. Vio gives
03:40.7
you unprecedented creative
03:46.7
Well, I've been interested in AI
03:48.6
for a couple years now.
03:50.7
We got in contact with some of the people at Google
03:52.7
and they had been working on something
03:54.5
of their own. So we're
03:56.6
all meeting here at Google Farms
03:58.5
to make a short film. The core
04:00.6
technology is Google
04:02.4
DeepMind's generative video model
04:04.5
that has been trained to convert
04:06.2
input text into output video.
04:12.5
to bring ideas to life that were
04:14.4
otherwise not possible. We can
04:16.4
visualize things on a timescale
04:18.3
that's 10 or 100 times faster than before.
04:21.1
When you're shooting, you can't reiterate
04:22.3
as much as you wish.
04:24.2
So we've been hearing that feedback that
04:27.4
optionality, more iteration,
04:30.4
more improvisation. But that's what's cool about it.
04:32.4
It's like you can make a mistake faster. That's all
04:34.2
you really want at the end of the day. At least in art,
04:36.4
it's just to make mistakes fast.
04:38.4
So using Gemini's multimodal
04:40.4
capabilities to optimize the model
04:42.2
training process, Vio is able
04:44.2
to better capture the nuance from proms.
04:46.7
So this includes cinematic
04:48.0
techniques and visual effects,
04:50.3
giving you total creative control.
04:53.5
Everybody's gonna become
04:54.4
a director, and everybody should be a director.
04:56.5
Because at the heart of all of this,
04:58.4
it's just storytelling. The closer we are
05:00.6
to being able to tell each other
05:02.1
our stories, the more we'll
05:04.3
understand each other. These models are
05:06.3
really enabling us to be more creative
05:08.5
and to share that
05:10.4
creativity with each other.
05:29.8
no, no, no, no, no, no thing
05:37.5
When I just do, do, do, do, do
05:42.3
thing, do my thing
05:43.6
I got that hustle
05:46.0
that straight up swing
05:50.2
no, no, no, no, no, no thing
05:52.6
Do my thing, do my thing
05:55.0
Like this, like that, just
05:56.4
fucking now, I just do it like that
