Android 17 Is the Biggest Upgrade in Years
522 segments
So, Google just had a keynote, and
they've announced possibly the single
biggest shake-up to the smartphone in
the last 5 years. In one go, they've
unveiled Gemini intelligence, so you can
ask AI to start taking over your phone
and doing things [music] for you.
That's cool. Android 17, which has ended
up far more eventful than the previous
few versions. There's a new Android Auto
that somehow out of nowhere is the
actual highlight of the show, all
finished off with an entirely new
category of laptop that lets you wiggle
your cursor and do magic AI things. Put
it this way, it is making Apple software
look very depressing by comparison. So,
first things first, Gemini intelligence
lets you do five things that you haven't
been able to properly do before. Like
Rambler. So, you know when you're
dictating a message, which I do all the
time, and then you ramble while
dictating that message, which I also do
all the time. Rambler is a new feature
that fixes you. So, let's say that your
wife asks you how much you enjoyed her
dinner, and you say, "It's a bit dry.
Actually, [music] no way. I love it.
I've never had
canned beans like this in my life."
This will no longer type out your entire
inner monologue. It'll smartly realize
what you're trying to say, or at least
that's what Google claims. I'll be the
test of that.
And it also means if you do accidentally
tell your wife her food was dry, you can
then say, "One, buy flowers. Two, buy
chocolates. Three, eat the beans and
nod." [music] And your phone will
realize by itself that you're describing
a list and format that into one. And
then you can say, "Swap the numbers for
emoji." It'll get that. Or even, I
guess, turn this whole message into
Hindi,
so she
can't read it.
This is really useful, and it actually
makes me feel like one of the
Flintstones when I look at this, and
then I look at my iPhone's current
dictation quality. Or another struggle
that we can all relate to, autofill. I
mean, don't get me wrong, the feature
itself is one of humanity's top [music]
five inventions, but every now and
again, you will come across a site or an
app where the rigid way that autofill
has been programmed means it just
doesn't activate. And
>> [music]
>> the pain that you feel then, having to
manually type out each letter of that
ridiculous email address you made when
you were 14 years old, knowing how easy
this could have been. Point being,
Gemini intelligence should be the cure
forever. It taps into your personal
intelligence pocket. Basically,
everything the AI knows about you,
including those random bits you're never
going to remember, like your passport
details. And then it smartly fills out
every form. And it should have a higher
success rate doing so, because unlike
traditional autofill, which is just
[music] interpreting code, so if a site
is not coded well, it might not realize
what data it's meant to fill in where,
Gemini, like you, can read and
understand what's being asked of it.
Seriously. Fantastic.
That sounds like I'm being sarcastic.
I'm actually gassed.
>> I've never been happier.
So it feels like part of Gemini
intelligence is very much focused on
giving people what they want, fixing
common pain points, but ones that have
actually required this new level of
intelligence to solve. But part of this
update is also improving things that you
didn't even know needed improving. Like
for example, you can now build widgets
by literally asking for them. They've
made it look so easy that this is
possibly the first time ever I can
safely say my granddad could be a
programmer. The example they gave is a
custom countdown to a marathon that
[music] you're running, but I think it
would be so cool to make one that can
pop up with activities happening around
you, but based on the time right now and
your specific GPS location. Another
example of something that I'm not sure
people were specifically asking [music]
for is app automations. They showed an
example of you taking a photo of this
flyer about a tour that you wanted to go
on, but then
>> [music]
>> asking Gemini, "Find me a tour like
this, but make sure it can accommodate
six people. And now Gemini is able to
first figure out what you mean, then
take that information to another app
like Expedia. [music]
It looks like it at this point kind of
thinks in the background until it's
finished, and then finally pings you a
notification to see the result. [music]
You click that, that's your booking
page. I do like that it doesn't try and
book for you. That would scare me. Do I
think I would use this?
Probably not for a while, to be honest.
Call me old-fashioned, but I still think
there's a lot of benefit to you [music]
looking at things like this yourself.
You know, you might find a tour that's
better rated [music] or closer to your
hotel or just something that you'd
actually rather do instead. Like the way
I see it, you and these five other
people are going to spend maybe a full
day doing this tour. Do you trust the AI
enough that you don't want to even spend
>> [music]
>> five minutes checking yourself for other
options? I'll pass. I will say this kind
of task doing [music] can become more
useful when you add in that Gemini can
now also pick up the full context of
what's on your screen. So let's say
you're kind of interested in going to
this stand-up comedy night, but you just
[music] cannot be bothered to plan the
logistics around it. Here, you can say
to Gemini, sort my parking space, and
[music] rest somewhat easy knowing that
it can read all of the context that it
needs by itself from your screen without
you needing to spell it out. And this is
probably the most visual example of how
Gemini is evolving into [music] more of
an agent with this update. Like you're
actually seeing it tap and type things
one step at a time completely without
your intervention. I could so imagine
just watching Gemini jaw clenched as it
confidently books me an eighth floor
spot when I can see that there's three
free on the first. But in principle,
this is a level up cuz like last week I
made a schedule of which supplements I
wanted to take at which [music] points
in the day. Yeah, I get it. I'm 30 now.
Magnesium is exciting. Gemini
intelligence, in theory, means that I
could hold the power button while
looking at the schedule and ask Gemini
to turn it into a set of reminders to
make sure that I follow this routine.
That's seriously helpful. Or let's say
you have your shopping list loaded up.
In theory, you could just bring up
Gemini and say "Buy [music] all of
this." And before you know it, you'll be
sitting there with a basket full of
these products ready for you to hit
checkout. They did also show how you
could be browsing an article and then
ask Gemini to make an infographic
summarizing it. Obviously, it worked,
but I wasn't super impressed with the
result. [music] It kind of screams AI
from 2 years ago. Overall, though,
Gemini intelligence feels pretty bang
on. The main caveat is just
availability. Like a lot of this isn't
coming [music] now. It's starting to
roll out this summer. It's not coming
all at once. Each feature is going to be
released when it's ready, they're
saying. [music] And to be honest, even
when it does come, it's coming to
Samsung and Google phones first.
Seemingly also only to the flagship
versions of them, potentially because it
relies on a more powerful version of the
Gemini Nano model, which needs to run on
the device itself. Point being, it could
be a year before your phone gets this,
if at all. But at least Google is being
clear about that. I think they're
specifically trying not to over-promise
and fall into the same trap that Apple
did with Apple intelligence. Thankfully,
the next Android update, Android 17, is
for all. And even if you took Gemini
intelligence aside, I actually think
what they've done with it is incredibly
clever. For example, you must have seen
by this point people posting this screen
reaction type of content. Well, now
Google is adding the ability to create
this screen reaction content instantly.
[music] It'll record your screen, it'll
use your front camera to record you,
it'll keep you in the right position,
and cut away your background. Having all
of that happening at once in real time
is a very efficient way to produce
videos. And on the pro end of the
spectrum, they're also bringing Adobe
Premiere to Android in summer with
templates specifically designed for you
to make YouTube shorts easily.
You can see why this works out for
Google, but it makes sense why they're
focusing on this because the iPhone
basically owns the creator market
currently, [music] and the creator
market has a lot of influence over the
mainstream market. So, Google's trying
to basically get the influencers on
board. And this thinly veiled full
frontal assault on Apple continues over
to Instagram. You know how for years
shooting content on Instagram for
Android has just been less reliable and
lower quality than iPhone. It's one of
the reasons I first switched away to
Apple like 6 years ago. But now, Google
is boldly claiming if you use a premium
Android device, you're about to get at
least as good if not better-looking
shots than on an iPhone. They're saying
that they've optimized the
capture-to-upload pipeline so that your
photos and videos lose as little quality
as possible when posting them. And that
even if you shoot directly within the
Insta app, you'll actually get
ultra-high dynamic range processing,
full working night mode, and built-in
video stabilization. How it's taken
[music] them till 2026 to do this, I
have no idea, but this could be great. I
say could because we've heard this whole
we fixed [music] Instagram sentiment
quite a lot, especially from Samsung
over the years. But I do feel a little
more confident because this is Google
themselves saying that they're fixing
this on a base Android level as opposed
to other companies whose software sits
on top and who probably have more
limited control. And you might know that
Instagram also has their own edits app,
[music] too. Well,
pretty cool, but with Android 17, you'll
now be able to use your phone's native
AI to power features within that [music]
app. So, like one tap to straight-up
enhance photos and videos. I'm hoping
this can actually upgrade the resolution
as opposed to just making things look
artificially bright and sharper. Or, do
you know how phones now have this audio
eraser feature to understand the
different sound sources in a video
you've taken so that you can turn down,
let's say, the wind noise specifically.
Well, you can now use that capability,
but within Instagram's editor. So,
automatically splitting up each audio
track into the individual stems that
made up that track. Whoever's making
these decisions at Google, I think they
got their head in the right place. And
while they were doing all this, they've
also decided to remake all the emoji on
Android. Literally all 4,000 of them
into 3D. But okay, my favorite part of
this new Android, though,
is Pause Point. You can tell that this
year, specifically, Google's really
looked at what else is out there and
thought, "How do we bring this into
Android itself?" Like the new Ramble
feature is very similar to a lot of the
new AI dictation apps [music] that have
been popping up. Create my widget, very
inspired by Nothing's Essential apps
feature. And Pause Point is functionally
identical to an app called One Sec that
I've been paying like $20 a year for.
The idea is you just set up a few apps
that you find distracting, and the next
time you try and open those apps, it
will ask you to stop and [music] breathe
for 10 seconds. Sounds silly, but it's
incredible. That 10 seconds gives you a
chance to decide either you didn't
really need to use the app, and you
close it, that you actually should
instead be doing something [music] more
useful, in which case you can get it to
redirect you to, say, your meditation
app, or three, that you really did need
to use the app in question, and you
proceed. I found by using One Sec that
even [music] just triggering that
thought means you're probably going to
be a bit more purposeful about your use
of this app, instead of just mindlessly
doom scrolling [music] before your
brain's even realized what's happening.
I do feel bad for the One Sec company.
They specifically say on their site,
"We're a small independent [music] team.
We're not affiliated with big
corporations." Which makes me think that
Google didn't exactly buy them or ask
permission to take [music] their
feature. But ultimately, this is still
probably good for users. Like everyone's
going to get the feature. You get it for
free, [music] and because it's now baked
into Android, it's going to feel a a
slicker to use, too. Now, here's what
was really not on my bingo card today.
Android Auto actually being the star of
the entire show. Because, I mean, for
starters, all the new Android and Gemini
intelligence features, if you have them
on your phone, you get them on your car,
too. And if you think about it, the car,
while you're completely occupied, is
where it's most important to have an
assistant that can do things for you.
Like, if you're on your way home, being
able to just say, "Order me my usual
food from DoorDash." And within 5
seconds, your car is on the checkout
page waiting on you to confirm. That
feels very much like the future. And if
Apple doesn't very quickly pull Siri out
of whatever Cupertino basement she's
been chained up in since like 2014, I
think it's going to make the gap between
these assistants too wide to ignore. And
that's just the start. Google Maps is
now about to have immersive 3D
navigation. The idea is that your view
looking at the map will more closely
match your view looking out to the
actual window. You'll be able to see
same buildings, the overpasses, and it
should be able to better highlight which
lane you need to be in. This is a
million miles more intuitive to me than
trying to squint at these things and
then translate them to the road in front
of you in the split second before you
approach a motorway roundabout. And then
they're saying that this new lane
tracking will be particularly accurate
on cars that support Gemini built in.
>> [music]
>> So, again, this is one of those
unsupported models business. But, if you
have a car that gets the support, then
you're going to get the extra benefits
of Gemini actually being trained on your
car's hardware, which means that Google
can use the live feed from your car's
front-facing camera to know for sure
what lane you're currently in. Plus,
also, you can then ask Gemini, "If I go
and buy this TV from Samsung, will it
fit in my trunk?"
>> [music]
>> And it will be able to answer with an
awareness of the car's exact dimensions.
Or, I'm imagining this also means you
could ask things like, "Remind [music]
me, how do I turn on my cruise control?"
And it would be able to reference the
manual and tell you specifically what
the button looks like. The software
should also look a little jazzier.
Google's saying the new Android Auto
will adapt to all the weird and
wonderful shapes of infotainment systems
we've been seeing lately, that they've
updated the way that individual apps
look, and that your car will now get the
same fonts, the same [music] wallpapers,
and the same smoother animations that
the phones are now getting. Plus,
widgets on the side, like a quick dialer
for your most frequent contact or a
button to open your garage door as
you're pulling up. Kind of feels like my
birthday, because then, for the first
time ever, Google is officially
supporting watching videos in your car
at 1080p 60 frames per second, assuming
your infotainment system supports that,
with Dolby Atmos audio. So, if you did
tell your wife her food was dry, at
least now you can get a high-quality
movie night for one out of it. And then,
to stop you trying to watch Fast &
Furious while you
become Fast & Furious, the videos will
minimize and play just audio as soon as
you start driving. What a clever little
feature. It does also need you to have
YouTube Premium, cuz that's what enables
background play. So, again,
you can see what Google gets out of all
of this, but then they surprised us all
by revealing an entirely new type of
laptop, too. Not a Chromebook, not a
Windows, a Google Book. It's not
actually super clear right now what
operating system this is running, but
they've got the tagline "Intelligence is
the new spec", which
I've actually got a great video all
about invented specs like intelligence.
But, basically, this is them saying, "We
won't be competing on actual measurable
specs." The impression I get is that
these Google Books are not about to be a
competitor to a top-end MacBook Pro for
someone who wants to edit and render
IMAX movies on their couch. I mean, come
[music] to think of it, ultra high-end
performance is just not what Google does
with any of their products, really, is
it? But, I still think it will be a
somewhat premium set of machines that
leans heavily on AI, which is why they
have reimagined the cursor. Now you can
wiggle it, which will put it into AI
mode [music] and let you do smart things
like you select a bunch of images and AI
will fuse them together in front of you.
They're saying that every Google book,
doesn't matter which company
manufactures [music] it, is going to
have this multicolored light strip,
which I assume is its way of indicating
that it's doing intelligence. Plus, some
of the new features from Android are
also getting carried over. Like you can
generate your own widgets on Google
book. And this looks like by far the
most seamless way I've seen to access
your phone and open apps from it [music]
directly without needing to pick the
thing up. With probably the coolest
thing being that your Android phone
storage can become accessible on your
Google book
>> [music]
>> as if it were part of the Google book
storage. So, transferring something from
your phone to your Google book would
just be going to one folder and pull
that file into another. Past experience
tells us that there is a broad range for
how terrible or great this could end up
being. But as far as I'm concerned, if
Google can make sure that Google books
get close to the build quality and the
trackpad quality of the MacBook, then
I would genuinely consider a full on
jump from the Apple ecosystem. I mean, I
spend 98% of my time on it on Google
Docs anyways. And so, looking at all of
this Google stuff together, while not
everything is coming soon and not
everything is coming to everyone,
I am really happy that they seem laser
focused on this mission of reducing
boring admin. [music] And I want that
more than ever when I'm traveling, which
is why for the last 2 years I've been
using Selia, [music] our sponsor. And I
just want to say that not having to
think about setting up a new eSIM every
time I go somewhere or
God forbid doing that 45-minute physical
SIM card scavenger hunt before I can
even Google how to get to the hotel is
one of those quiet life upgrades that
you [music] didn't know you needed. Once
you've set the eSIM up once, you just
download the data pack for [music] any
new country when you need it. It's as
simple as downloading an app, and you
can get an exclusive 15% discount
[music] on Saily eSIM data plans using
the code Boss at checkout.
Ask follow-up questions or revisit key timestamps.
This video provides an overview of Google's latest keynote, highlighting significant advancements in AI integration across their ecosystem. Key features include the new Gemini intelligence, which aims to automate tasks, improve dictation, and offer intelligent autofill. Additionally, the video discusses updates to Android 17, such as 'Pause Point' to prevent distractions, enhanced integration with Instagram, and major improvements to Android Auto, including immersive 3D navigation. Finally, Google introduced a new category of 'Google Book' laptops that prioritize AI features and seamless connectivity with Android devices.
Videos recently processed by our community