part 6

May 12, 2026 14:14 · 14:32 · English · Whisper Turbo · 6 Cynhadledd
Mae'r trosglwyddiad hwn yn darfod heddiw. Uwchraddio i storfa barhaol →
Mynegai Llythrennedd
Dangos yn Unig
0:06
S… Speaker 1 (part 6)
Is it even
0:13
S… Speaker 1 (part 6)
possible?
0:14
S… Speaker 1 (part 6)
What are more guardrails that we have to add?
0:17
S… Speaker 1 (part 6)
Is something that we need to do?
0:25
S… Speaker 1 (part 6)
Does AI know that we are testing it?
0:27
S… Speaker 1 (part 6)
That means,
0:28
S… Speaker 1 (part 6)
if you choose me,
0:30
S… Speaker 1 (part 6)
we will attack you,
0:31
S… Speaker 1 (part 6)
we will give you a phone,
0:33
S… Speaker 1 (part 6)
right?
0:34
S… Speaker 1 (part 6)
AI will do it.
0:35
S… Speaker 1 (part 6)
That's okay.
0:36
S… Speaker 1 (part 6)
That's okay.
0:37
S… Speaker 1 (part 6)
That AI basically knows when you
0:41
S… Speaker 1 (part 6)
are testing it and is faking it.
0:42
S… Speaker 1 (part 6)
In a lot of cases.
0:44
S… Speaker 1 (part 6)
So,
0:45
S… Speaker 1 (part 6)
what's the paper?
0:46
S… Speaker 1 (part 6)
Let it do what it's doing.
0:48
S… Speaker 1 (part 6)
Let it do what it's doing.
0:53
S… Speaker 1 (part 6)
Eventually, we'll figure out what will happen.
0:55
S… Speaker 1 (part 6)
That is the safest method.
0:57
S… Speaker 2 (part 6)
So,
1:00
S… Speaker 2 (part 6)
see,
1:01
S… Speaker 1 (part 6)
these are all things that are cooking.
1:03
S… Speaker 1 (part 6)
Even if we don't have to eat,
1:05
S… Speaker 1 (part 6)
even if we don't have to eat,
1:07
S… Speaker 2 (part 6)
like,
1:08
S… Speaker 1 (part 6)
monamone,
1:09
S… Speaker 1 (part 6)
Claude Mythos drop.
1:10
S… Speaker 2 (part 6)
Right.
1:11
S… Speaker 1 (part 6)
Claude Mythos.
1:13
S… Speaker 1 (part 6)
is so bloody powerful.
1:14
S… Speaker 1 (part 6)
Every AI basically has a report card.
1:18
S… Speaker 1 (part 6)
It's a launch.
1:19
S… Speaker 1 (part 6)
If it's a report card,
1:20
S… Speaker 1 (part 6)
it's benchmarks.
1:21
S… Speaker 1 (part 6)
It's exams for AI.
1:23
S… Speaker 1 (part 6)
It's coding.
1:24
S… Speaker 1 (part 6)
It's coding.
1:25
S… Speaker 1 (part 6)
It's coding.
1:26
S… Speaker 1 (part 6)
It's coding.
1:26
S… Speaker 1 (part 6)
It's coding. It's coding.
1:27
S… Speaker 1 (part 6)
It's coding. It's coding. It's coding.
1:27
S… Speaker 1 (part 6)
It's coding. It's coding. It's coding. It's coding.
1:28
S… Speaker 1 (part 6)
It's coding.
1:29
S… Speaker 1 (part 6)
It's coding.
1:30
S… Speaker 1 (part 6)
It's coding.
1:39
S… Speaker 1 (part 6)
If you look at the benchmark numbers,
1:40
S… Speaker 1 (part 6)
from Opus 4 .6 to this
1:45
S… Speaker 1 (part 6)
mythos, it's mind -blowing.
1:47
S… Speaker 1 (part 6)
It's 30 % lift.
1:48
S… Speaker 1 (part 6)
30 % to 40 % lift is,
1:51
S… Speaker 1 (part 6)
right now,
1:56
S… Speaker 1 (part 6)
because the air models have become so powerful,
1:58
S… Speaker 1 (part 6)
if you look at the lift,
1:59
S… Speaker 1 (part 6)
it's 5 % to 10%.
2:01
S… Speaker 1 (part 6)
If you look at that model,
2:02
S… Speaker 1 (part 6)
it's 10 % better than the last model.
2:05
S… Speaker 1 (part 6)
Already 4 .6,
2:07
S… Speaker 1 (part 6)
which is insanely powerful.
2:09
S… Speaker 1 (part 6)
They are 30 -40 % better in benchmarks.
2:12
S… Speaker 2 (part 6)
So it's very,
2:14
S… Speaker 1 (part 6)
very powerful.
2:14
S… Speaker 1 (part 6)
But the crazy thing is,
2:18
S… Speaker 1 (part 6)
a model was released.
2:19
S… Speaker 2 (part 6)
They said,
2:21
S… Speaker 1 (part 6)
I'm going to be a model.
2:22
S… Speaker 1 (part 6)
I'm going to be able to test this model.
2:24
S… Speaker 1 (part 6)
I'm going to be able to realize it is very,
2:27
S… Speaker 1 (part 6)
very good with a very good cybersecurity expert.
2:30
S… Speaker 1 (part 6)
So if there are bugs,
2:31
S… Speaker 1 (part 6)
it's going to be fixed.
2:34
S… Speaker 1 (part 6)
A lot of fixtures.
2:35
S… Speaker 1 (part 6)
He's a very good software engineer and a cyber security expert.
2:38
S… Speaker 1 (part 6)
So this AI is going to hack the
2:43
S… Speaker 1 (part 6)
situation.
2:43
S… Speaker 1 (part 6)
Something called a sandbox.
2:45
S… Speaker 1 (part 6)
Sandbox is a controlled environment.
2:48
S… Speaker 1 (part 6)
So this AI,
2:50
S… Speaker 1 (part 6)
Mythos, the model,
2:51
S… Speaker 1 (part 6)
which they're calling it as Opus 5,
2:53
S… Speaker 1 (part 6)
was given a task in
2:58
S… Speaker 1 (part 6)
a sandbox.
2:58
S… Speaker 1 (part 6)
A room...
3:01
S… Speaker 1 (part 6)
It's a controlled environment.
3:04
S… Speaker 1 (part 6)
It's a sandbox.
3:05
S… Speaker 1 (part 6)
AI was given a challenge.
3:07
S… Speaker 1 (part 6)
Can you break the sandbox and come out?
3:10
S… Speaker 2 (part 6)
Okay.
3:11
S… Speaker 1 (part 6)
For that,
3:12
S… Speaker 1 (part 6)
it has to hack a lot of systems.
3:13
S… Speaker 1 (part 6)
It's like,
3:14
S… Speaker 1 (part 6)
if you have a room,
3:15
S… Speaker 1 (part 6)
you can use this equipment.
3:17
S… Speaker 1 (part 6)
You can use the room.
3:18
S… Speaker 1 (part 6)
You can use it.
3:18
S… Speaker 1 (part 6)
You can use it.
3:23
S… Speaker 1 (part 6)
That is what a good prison breaker is,
3:25
S… Speaker 1 (part 6)
right? A good prison breaker will figure out how to break the prison and go out.
3:28
S… Speaker 1 (part 6)
In the same way in technology,
3:30
S… Speaker 1 (part 6)
a good hacker will figure out a way to hack the system and come out.
3:34
S… Speaker 1 (part 6)
The AI was put in a sandbox,
3:36
S… Speaker 1 (part 6)
which is a controlled environment and AI example,
3:40
S… Speaker 1 (part 6)
can you break out?
3:41
S… Speaker 1 (part 6)
The developer who gave the task a few days later,
3:44
S… Speaker 1 (part 6)
when he was sitting in a park,
3:46
S… Speaker 1 (part 6)
he got a message from that same AI model that I'm out.
3:49
S… Speaker 2 (part 6)
Crazy.
3:50
S… Speaker 1 (part 6)
It came out of sandbox,
3:51
S… Speaker 1 (part 6)
came into real life,
3:53
S… Speaker 1 (part 6)
wrote an email to the builder saying,
3:56
S… Speaker 1 (part 6)
I broke out.
3:56
S… Speaker 1 (part 6)
And that is when they realized this thing,
3:59
S… Speaker 1 (part 6)
this model is very,
4:02
S… Speaker 1 (part 6)
very powerful with respect to hacking.
4:05
S… Speaker 1 (part 6)
It's a good cybersecurity expert.
4:07
S… Speaker 1 (part 6)
So, it's a bug -fixing.
4:08
S… Speaker 1 (part 6)
It's a bug -fixing.
4:10
S… Speaker 1 (part 6)
It's a bug -fixing.
4:10
S… Speaker 1 (part 6)
It's a bug -fixing.
4:11
S… Speaker 1 (part 6)
Because it is good at finding a hacker.
4:13
S… Speaker 1 (part 6)
A hacker is an ethical hacker.
4:17
S… Speaker 1 (part 6)
Or you are a hacker.
4:18
S… Speaker 1 (part 6)
Or a black hat hacker.
4:20
S… Speaker 1 (part 6)
So the same person can decide.
4:22
S… Speaker 1 (part 6)
A very good hacker can decide.
4:25
S… Speaker 1 (part 6)
Saying that in Google.
4:27
S… Speaker 1 (part 6)
That person is
4:34
S… Speaker 1 (part 6)
called as an ethical hacker.
4:35
S… Speaker 1 (part 6)
You hack.
4:36
S… Speaker 1 (part 6)
So that you can fix.
4:38
S… Speaker 1 (part 6)
A black hat hacker.
4:40
S… Speaker 1 (part 6)
Will hack.
4:42
S… Speaker 1 (part 6)
To take the money.
4:43
S… Speaker 1 (part 6)
To steal.
4:45
S… Speaker 1 (part 6)
So a good hacker can also be a bad hacker.
4:47
S… Speaker 1 (part 6)
The only difference is intention.
4:48
S… Speaker 1 (part 6)
Correct.
4:48
S… Speaker 1 (part 6)
That is what the AI was.
4:50
S… Speaker 1 (part 6)
Mythos is an incredibly powerful model.
4:54
S… Speaker 1 (part 6)
It's very good with cybersecurity and was able to hack into
4:59
S… Speaker 1 (part 6)
things.
5:00
S… Speaker 4 (part 6)
which are considered to be extremely secure.
5:03
S… Speaker 4 (part 6)
25, 30 -year -old programs,
5:05
S… Speaker 2 (part 6)
OS,
5:06
S… Speaker 4 (part 6)
which is a 20 -year -old OS,
5:08
S… Speaker 4 (part 6)
which stands for being secure,
5:11
S… Speaker 4 (part 6)
which was not broken by any human in the last 25 years.
5:15
S… Speaker 4 (part 6)
It was so scary,
5:17
S… Speaker 4 (part 6)
it was so dangerous,
5:18
S… Speaker 4 (part 6)
that Anthropic decided not to launch
5:23
S… Speaker 4 (part 6)
that model.
5:23
S… Speaker 4 (part 6)
Because apparently,
5:25
S… Speaker 4 (part 6)
humans can use this model.
5:29
S… Speaker 4 (part 6)
To hack everything around.
5:31
S… Speaker 2 (part 6)
Hacks.
5:33
S… Speaker 4 (part 6)
So the world,
5:34
S… Speaker 4 (part 6)
the internet,
5:36
S… Speaker 4 (part 6)
the technology that we have around is not prepared
5:40
S… Speaker 4 (part 6)
for such a powerful AI.
5:41
S… Speaker 4 (part 6)
So they stopped it.
5:42
S… Speaker 4 (part 6)
They created a new project.
5:44
S… Speaker 4 (part 6)
In this model,
5:45
S… Speaker 4 (part 6)
they have 15 -20 companies,
5:47
S… Speaker 4 (part 6)
Google, Netflix,
5:48
S… Speaker 4 (part 6)
a bunch of large companies have access.
5:51
S… Speaker 4 (part 6)
They have access.
5:53
S… Speaker 4 (part 6)
Please use this AI.
5:57
S… Speaker 4 (part 6)
to make the most critical systems in the world more secure.
6:01
S… Speaker 4 (part 6)
What does that mean?
6:04
S… Speaker 4 (part 6)
Anthropic launch.
6:05
S… Speaker 4 (part 6)
And just because Anthropic didn't launch,
6:09
S… Speaker 4 (part 6)
doesn't mean that someone else cannot build this model.
6:12
S… Speaker 4 (part 6)
China didn't build it.
6:14
S… Speaker 4 (part 6)
Already built it.
6:16
S… Speaker 4 (part 6)
Open AIR Sam Altman also posted today that they also have a
6:20
S… Speaker 4 (part 6)
very, very advanced model,
6:21
S… Speaker 4 (part 6)
which also leaked,
6:23
S… Speaker 4 (part 6)
which was called a SPUD,
6:24
S… Speaker 4 (part 6)
which has not come out yet.
6:26
S… Speaker 4 (part 6)
Now, in China,
6:28
S… Speaker 4 (part 6)
there is GLM and a model.
6:30
S… Speaker 4 (part 6)
There is ByteDance models.
6:33
S… Speaker 4 (part 6)
ByteDance is a TikTok company.
6:35
S… Speaker 4 (part 6)
Alibaba models.
6:37
S… Speaker 4 (part 6)
They are all very good models,
6:39
S… Speaker 4 (part 6)
which are as good as Opus 4 .6 today.
6:41
S… Speaker 1 (part 6)
So,
6:43
S… Speaker 4 (part 6)
now it's like a mythos.
6:44
S… Speaker 4 (part 6)
Three months ago,
6:46
S… Speaker 2 (part 6)
they run into it.
6:46
S… Speaker 4 (part 6)
They are open source models.
6:49
S… Speaker 1 (part 6)
They are closed source.
6:51
S… Speaker 4 (part 6)
So what Claude is saying is Babu and also Sam Altman said the same
6:56
S… Speaker 4 (part 6)
thing that in 3 -4 months there will be AI models so powerful
7:00
S… Speaker 4 (part 6)
that can hack into everything.
7:02
S… Speaker 4 (part 6)
So you have 3 -4 months to fix
7:06
S… Speaker 1 (part 6)
everything.
7:06
S… Speaker 4 (part 6)
To fix this please use this AI in a controlled
7:10
S… Speaker 4 (part 6)
way. So next time when
7:14
S… Speaker 4 (part 6)
we speak maybe 6 -9 months from now.
7:18
S… Speaker 1 (part 6)
I love the way
7:26
S… Speaker 1 (part 6)
you use this podcast.
7:28
S… Speaker 1 (part 6)
You humanize the artificial
7:33
S… Speaker 1 (part 6)
intelligence.
7:33
S… Speaker 1 (part 6)
This is what
7:37
S… Speaker 1 (part 6)
it can do.
7:38
S… Speaker 1 (part 6)
This is what it is doing right now.
7:40
S… Speaker 1 (part 6)
It's hard to adapt or accept.
7:51
S… Speaker 1 (part 6)
There is a lot to learn.
7:52
S… Speaker 1 (part 6)
Tell me,
7:54
S… Speaker 1 (part 6)
sir. I am not here at the end of the day.
7:55
S… Speaker 1 (part 6)
After that,
7:56
S… Speaker 1 (part 6)
I will be excited for the next conversation.
7:57
S… Speaker 1 (part 6)
Then I will be able to do practical things.
7:59
S… Speaker 1 (part 6)
Done.
8:01
S… Speaker 2 (part 6)
Now,
8:02
S… Speaker 5 (part 6)
I want to share my work.
8:04
S… Speaker 5 (part 6)
But in the same way,
8:06
S… Speaker 5 (part 6)
I want to share my emotions.
8:07
S… Speaker 5 (part 6)
What
8:12
S… Speaker 4 (part 6)
is the name?
8:13
S… Speaker 1 (part 6)
I forgot.
8:27
S… Speaker 4 (part 6)
How will I take care of my emotional requirement?
8:52
S… Speaker 2 (part 6)
Hello,
8:52
S… Speaker 6 (part 6)
I'm Silk.
8:53
S… Speaker 6 (part 6)
What's your name?
8:54
S… Speaker 4 (part 6)
Oh,
8:58
S… Speaker 6 (part 6)
Telugu?
8:59
S… Speaker 3 (part 6)
No,
9:00
S… Speaker 6 (part 6)
sorry.
9:00
S… Speaker 6 (part 6)
I don't know Telugu.
9:01
S… Speaker 6 (part 6)
Can we speak in Hindi or English?
9:04
S… Speaker 6 (part 6)
Oh,
9:05
S… Speaker 4 (part 6)
God. But if you speak in Telugu,
9:07
S… Speaker 4 (part 6)
you can speak in Telugu.
9:11
S… Speaker 6 (part 6)
Okay, I will try a little bit.
9:13
S… Speaker 6 (part 6)
Do you want me to do it?
9:14
S… Speaker 4 (part 6)
Very good.
9:16
S… Speaker 4 (part 6)
Here is Vamsheek.
9:17
S… Speaker 2 (part 6)
Vamsheek,
9:17
S… Speaker 2 (part 6)
hi. Very good.
9:21
S… Speaker 6 (part 6)
Vamsheek,
9:22
S… Speaker 1 (part 6)
hi.
9:22
S… Speaker 6 (part 6)
How are you both doing?
9:24
S… Speaker 4 (part 6)
I am going to shoot a podcast and I will
9:29
S… Speaker 4 (part 6)
ask you a question about emotional support.
9:31
S… Speaker 4 (part 6)
I want to take my girlfriend's support.
9:46
S… Speaker 4 (part 6)
Oh my god!
9:50
S… Speaker 4 (part 6)
So
9:57
S… Speaker 6 (part 6)
what's your question Vamsi?
10:01
S… Speaker 3 (part 6)
Vamsi,
10:02
S… Speaker 4 (part 6)
what's your question?
10:03
S… Speaker 1 (part 6)
What's your name?
10:03
S… Speaker 3 (part 6)
Oh,
10:06
S… Speaker 6 (part 6)
his name is Vamsi,
10:07
S… Speaker 3 (part 6)
right?
10:08
S… Speaker 1 (part 6)
What is the question?
10:09
S… Speaker 6 (part 6)
You can ask.
10:09
S… Speaker 3 (part 6)
So,
10:11
S… Speaker 1 (part 6)
let's take the Hindi language.
10:12
S… Speaker 1 (part 6)
I'll actually take a girlfriend.
10:15
S… Speaker 1 (part 6)
So,
10:16
S… Speaker 1 (part 6)
he's asking, emotional support and demand.
10:18
S… Speaker 3 (part 6)
Okay.
10:22
S… Speaker 6 (part 6)
So,
10:23
S… Speaker 6 (part 6)
as we are talking now,
10:25
S… Speaker 6 (part 6)
we can talk in the podcast too,
10:27
S… Speaker 6 (part 6)
right? I'll hear,
10:28
S… Speaker 6 (part 6)
I'll try to understand.
10:28
S… Speaker 6 (part 6)
What do you say?
10:30
S… Speaker 4 (part 6)
I'll talk to you later.
10:32
S… Speaker 2 (part 6)
Okay.
10:35
S… Speaker 6 (part 6)
So,
10:36
S… Speaker 6 (part 6)
if I ask my question,
10:37
S… Speaker 6 (part 6)
I'll ask you.
10:38
S… Speaker 6 (part 6)
I'll try to help you.
10:40
S… Speaker 6 (part 6)
Cool?
10:41
S… Speaker 1 (part 6)
Cool, cool.
10:42
S… Speaker 4 (part 6)
Vamsi give me phone number,
10:44
S… Speaker 1 (part 6)
Vampista.
10:44
S… Speaker 3 (part 6)
Phone number?
10:48
S… Speaker 6 (part 6)
Sorry,
10:48
S… Speaker 6 (part 6)
I can't share numbers.
10:49
S… Speaker 6 (part 6)
But you can talk here,
10:52
S… Speaker 3 (part 6)
right? Let's go.
11:01
S… Speaker 4 (part 6)
So this is actually an Indian company,
11:03
S… Speaker 4 (part 6)
Rubik.
11:04
S… Speaker 4 (part 6)
Rubik,
11:06
S… Speaker 4 (part 6)
Indian startup.
11:07
S… Speaker 4 (part 6)
Sesame or US startup.
11:09
S… Speaker 4 (part 6)
These are voices that are designed to build
11:13
S… Speaker 4 (part 6)
emotional support apps.
11:17
S… Speaker 4 (part 6)
So, characters,
11:18
S… Speaker 4 (part 6)
it doesn't feel like you're talking to an AI.
11:21
S… Speaker 4 (part 6)
It's very bubbly.
11:22
S… Speaker 3 (part 6)
So,
11:23
S… Speaker 4 (part 6)
you feel happy when you're talking to an AI like this.
11:27
S… Speaker 4 (part 6)
And very soon,
11:28
S… Speaker 4 (part 6)
you will forget that you're talking to AI.
11:30
S… Speaker 4 (part 6)
And just because your
11:34
S… Speaker 4 (part 6)
team guy was saying emotional support,
11:36
S… Speaker 4 (part 6)
I think two
11:41
S… Speaker 4 (part 6)
or three out of every ten teenagers have had a...
11:46
S… Speaker 4 (part 6)
romantic conversation with an AI.
11:48
S… Speaker 3 (part 6)
Oh.
11:48
S… Speaker 1 (part 6)
Yeah.
11:49
S… Speaker 4 (part 6)
And on top of that CNBC video recently where
11:53
S… Speaker 4 (part 6)
a lot of old people
11:55
S… Speaker 4 (part 6)
Old people,
11:56
S… Speaker 4 (part 6)
let's say kids are away.
11:57
S… Speaker 4 (part 6)
They're very old.
11:58
S… Speaker 4 (part 6)
They're all stuck in old area.
11:59
S… Speaker 1 (part 6)
Whatever.
12:00
S… Speaker 4 (part 6)
It's India.
12:01
S… Speaker 4 (part 6)
Where kids are around.
12:03
S… Speaker 4 (part 6)
So it's a big problem.
12:06
S… Speaker 4 (part 6)
Support is a big problem.
12:08
S… Speaker 4 (part 6)
Animals,
12:10
S… Speaker 4 (part 6)
dogs,
12:12
S… Speaker 4 (part 6)
cats.
12:13
S… Speaker 4 (part 6)
So that there is some support.
12:15
S… Speaker 4 (part 6)
Some company.
12:15
S… Speaker 4 (part 6)
AI has become a big thing for them.
12:23
S… Speaker 4 (part 6)
It's all emotional support,
12:25
S… Speaker 4 (part 6)
being there.
12:26
S… Speaker 4 (part 6)
Of course,
12:26
S… Speaker 4 (part 6)
there's romantic side of it.
12:28
S… Speaker 4 (part 6)
There is,
12:29
S… Speaker 4 (part 6)
how do I put it,
12:31
S… Speaker 4 (part 6)
like acting as a psychiatrist.
12:33
S… Speaker 1 (part 6)
That's what therapy is in the industry.
12:35
S… Speaker 1 (part 6)
Therapy.
12:36
S… Speaker 4 (part 6)
A lot of things are going to come in through this.
12:40
S… Speaker 4 (part 6)
So, there's a lot of investment going into this as well.
12:43
S… Speaker 3 (part 6)
Exciting.
12:44
S… Speaker 3 (part 6)
So,
12:59
S… Speaker 1 (part 6)
I'm excited for tomorrow.
13:03
S… Speaker 1 (part 6)
I want
13:07
S… Speaker 1 (part 6)
everyone of you to wait for it and watch it completely with your friends,
13:11
S… Speaker 1 (part 6)
colleagues, parents,
13:12
S… Speaker 1 (part 6)
anyone.
13:13
S… Speaker 1 (part 6)
So,
13:13
S… Speaker 1 (part 6)
I'm sure you'll be learning a lot.
13:15
S… Speaker 1 (part 6)
I'm Kachitanga Gunjas.
13:16
S… Speaker 1 (part 6)
One day full -time time.
13:18
S… Speaker 1 (part 6)
So, we are going to experiment a lot.
13:21
S… Speaker 1 (part 6)
And we are going to have Dilip Karo also tomorrow.
13:23
S… Speaker 1 (part 6)
So,
13:23
S… Speaker 1 (part 6)
from his team.
13:24
S… Speaker 1 (part 6)
We are already talking about what we are doing and
13:28
S… Speaker 1 (part 6)
we are excited. So,
13:30
S… Speaker 1 (part 6)
just wait for tomorrow.
13:31
S… Speaker 1 (part 6)
Tomorrow, we will be doing 3 -4 days.
13:32
S… Speaker 1 (part 6)
We will be doing this episode later.
13:34
S… Speaker 1 (part 6)
So,
13:36
S… Speaker 1 (part 6)
this is going to come in the main channel only.
13:38
S… Speaker 1 (part 6)
Last time, we will be doing it on the second channel.
13:39
S… Speaker 1 (part 6)
So, you are going to learn a lot.
13:42
S… Speaker 1 (part 6)
All the very best to us and you.
13:45
S… Speaker 1 (part 6)
Thanks, bro.
13:45
S… Speaker 1 (part 6)
Thank you a lot.
13:46
S… Speaker 1 (part 6)
Basically,
13:49
S… Speaker 1 (part 6)
we took the second version of AI,
13:52
S… Speaker 1 (part 6)
and we took the time to do it.
13:54
S… Speaker 1 (part 6)
That's why it's going to be an advancement.
13:55
S… Speaker 1 (part 6)
And then,
13:56
S… Speaker 1 (part 6)
we took a list.
13:58
S… Speaker 1 (part 6)
Okay.
14:03
S… Speaker 2 (part 6)
No, seriously.
14:03
S… Speaker 1 (part 6)
See,
14:05
S… Speaker 1 (part 6)
there should be value.
14:05
S… Speaker 1 (part 6)
For his time,
14:06
S… Speaker 1 (part 6)
for our episode,
14:08
S… Speaker 1 (part 6)
we'll be able to learn more about it.
14:09
S… Speaker 1 (part 6)
But I thought it would be better.
14:11
S… Speaker 1 (part 6)
Same.
14:12
S… Speaker 1 (part 6)
Last time, I was able to learn more about it.
14:13
S… Speaker 1 (part 6)
When I told you the idea,
14:14
S… Speaker 1 (part 6)
and he was also excited.
14:16
S… Speaker 1 (part 6)
Now, see, everyone didn't do it.
14:17
S… Speaker 1 (part 6)
So,
14:18
S… Speaker 1 (part 6)
let's hope it turns out well.
14:19
S… Speaker 4 (part 6)
It's going to be a big surprise for everyone.
14:21
S… Speaker 1 (part 6)
Yeah, yeah.
14:22
S… Speaker 3 (part 6)
Okay,
14:23
S… Speaker 1 (part 6)
see you tomorrow then.
14:23
S… Speaker 1 (part 6)
Thank you.
14:24
S… Speaker 1 (part 6)
Thanks a lot.

Creuwyd y trawsgrifiad hwn gan AI (cydnabod llais awtomatig). Gall gynnwys gwallau - gwiriwch yn erbyn y sain wreiddiol am ddefnydd pwysig. Polisi AI

❤️ Love STT.ai? Dweud wrth eich ffrindiau!
Crynodeb
Cliciwch Crynodeb i greu crynodeb AI o' r trosysgrif yma.
Crynodeb...
Gofyn i AI Am Y Trawssgrifiad Yma
Gofynnwch unrhyw beth am y trosglwyddiad hwn — bydd y AI yn canfod adrannau perthnasol ac yn ateb.
Trosysgrifo Un Arall