Why Samsung Is Betting on AI for Smartphone Innovation | Tech News Briefing | WSJ

[Music]

this is your Tech news briefing for

Wednesday March 29th I'm Zoe Thomas for

The Wall Street Journal

Samsung is the world's largest

smartphone maker by shipments but the

smartphone market is in a weird place

sales are down but customers will still

spend on the highest end phones it's

perhaps no wonder then that Samsung is

betting on artificial intelligence to

develop new features for its devices at

this year's mwc formerly World mobile

Congress Samsung EVP Patrick chomay

spoke at the wsj's journal house he

answered questions from our senior

personal Tech columnist Joanna stern

here are highlights from the event

Patrick thank you for joining us what is

Samsung thinking about Ai and is it

really at the top of the list as you're

innovating yes yes definitely it is

called as being a lot of focus on the on

the Voice assistance so we have Bixby

that have been Alexa and others but it's

much more than that for us actually the

biggest application of AI in our device

is in the digital experience in the

camera things like to get the quality of

pictures in low lights that we have I

would encourage you to try what we do is

we combine a series of pictures and with

a neural engine optimize all the noise

the 200 megapixel camera is not just

about having 12 megapixel as a big

number and Detail in low light condition

we can combine four to 16 pixels and

with AI techniques filter out the noise

to get better quality so AI is very very

deeply embeddy enough for that in image

processing in the power management for

example to foresee which app you are

using not using shut them down and

optimize the battery usage it's all over

what about the generative AI space you

mentioned Bixby and I think there's

there's a feeling out there how many

people think and I'm not going to just

we won't focus on Bixby but we'll say

Bixby Siri and Alexa are just the

smartest in the world anyone out there

think they're just super smart all right

how many people out there think chat gbt

and Dolly and stable diffusion are just

incredible

smart

I mean just are at least wowed me too

yeah okay so for years we've sort of

thought okay these voice assistants

they're not so smart no offense to Bixby

now there's this next wave is this the

FI is this really the like realization

that this is what the voice assistants

are going to be and are we going to see

that integration

um I believe we are talking about two

different things the the food I call the

first generation of voice assistants I

was Bixby and others

I think they were conceived on the on a

model which was some kind of brain

assistant human Centric design that

would control all the devices and the

services and the reality is that the

artificial intelligence is much more

distributed but we just said it's in the

the image processing it's in the

applications like Gmail gmaps is in our

own applications so I I think we are

going for a much more distributed vision

and in that Vision the the ah the

generative AI techniques I think they

offer great possibilities but again not

in this kind of centralized way you can

imagine this technology is used in each

and every part of the service area which

rgbt and Bing I've been using Bing a lot

and you are interacting with a computer

you are talking to it in a way that it

is human-like which I think we sort of

all expected you know this is where Siri

and Bixby and Alexa were gonna go

is are you is Samsung looking at using

these Technologies to make that

conversation better on on a phone

absolutely so we as you know we have

great partners and at this third

Microsoft is a partner Microsoft is a

popular Google is a popular which makes

it interesting

so Bixby Bixby is our voice agent

control the phone so it's a very

effective tool to commence things on

your phone not only your phone also we

use it a lot in the connected home come

on your uh Appliance and things like

that so we are very focused on how to

Simply command and Quantum our Samsung

devices by voice uh the the generative

AI techniques is allowing different

things and our expectation is that we

will work without buffers deeply we have

to leverage all the potential you're

able to use Bixby to respond to a call

yeah that's good right tell me how that

works and then I want to talk about the

generative voice part of that so all

that works if you are in this situation

and some Stephanie would call me I would

answer my text I am in the meeting right

now I call you back later my voice will

play back to where in that voice chord

so for now the voice recordings it works

in Korean and English in the Korean

language you can record your own voice

so the The Listener would receive

actually your own voice in English it

will come very soon so we love that

feature it's good but what you're

virtually doing is doing a voice clone

you're making a version of your voice

and then you can type in whatever and

you're going to hear that version of

your voice right yes but it's still

generative AI this is uh fundamental uh

well digital filtering okay we will

capture the tone of your voice and

reproduce that tone so which is quite

different from the full machine learning

generative AI engine so it won't sound

exactly like me it does it does okay so

it will because it will sound like is

that a good idea

that my phone will sound exactly like it

Won't Anyone can type in what I sound

like you decided to only she decided to

but will there be any so say um somebody

decides they want to do a recording of

Tom Cruise on this and their voice will

be Tom Cruise can that is there

restrictions on that but you can do that

today

you can record any boys they are

application on internet to modify the

tune you can do that but this is going

to be on Samsung phones there's more

Samsung phones than probably some of

these uses I think it's based on user

choice I don't see it whatever it's

really I think it's quite fun actually

okay let's shift to talking about VR ar

your partnership Samsung Google and

Qualcomm around headsets and we talked a

little bit about the future but it

sounds like you guys are all three

teaming up to make something what are

you making

hmm yeah that's one today it's a fun

factor of fact maybe you have a picture

you have a phone you can race or they

can say no I don't have a picture yet

but I can tell you what we do the reason

why we announce is we believe it's a

it's a long journey and there is a lot

of work to be done and a characteristic

of our ecosystem is we are open so we

don't do it alone so we are doing a lot

of fundamental work with obviously

Qualcomm and Google on the core

Technologies and finding out the right

experiences but also as we go we are

signaling to a number of ecosystem

players that they should be ready to

develop applications and apps and

services to that ecosystem because we

believe uh when the mixed reality

devices will come in okay we do some

good stuff and we have some good apps

and so does Google and Qualcomm as

technology but the real power comes from

a richer ecosystem there are a lot of

startups and a lot of smaller companies

will innovate if we give them the right

access and if we have enough time

develop on that on that platform and

they also have the confidence that big

companies like Qualcomm and Samsung and

Google are into it for the long term

thank you so much for joining us and

thank you all of you for being here and

that's it for today's Tech news briefing

for more Tech stories head over to our

website

wsj.com I'm Zoe Thomas for The Wall

Street Journal thanks for listening

foreign