Meta creates Super intelligent Lab!

ABC's Mike Dobuski reports on AGI...What is AGI? It's AI that could match the intelligence of a human brain. Plus, Trump Family Phone Plan

Episode Transcript

Speaker 1 (00:00):
A right, let's go over to the Legacy Retirement Group

(00:01):
dot com phone line. Say good morning to ABC's technology reporter.
It is Mike Debuski on what we call Tech Tuesday.
Mike Meta, which is the parent company for Instagram and Facebook.
They're building the super intelligence Lab. That sounds pretty kind
of superhero ish sci fi. Yeah, definitely so.

Speaker 2 (00:22):
Meta last week announced that they're making this major investment
in a company known as Scale AI. And Scale AI
is a startup that essentially is a pix and shovels company, right.
It makes the data that the large language models are
trained on. If you are an Open AI or a
Google and you want really good information to make your
chatbot smarter, Scale AI is the company that you go to.

(00:43):
And Meta is now investing more than fourteen billion dollars
in this company. It gives Meta a forty nine percent
ownership stake in Scale, but it doesn't give them control
over the company. What they get out of this deal
is Alexander Wang, the co founder and CEO of Scale.
He is going to join Meta on a new team

(01:03):
that is kind of designed to create this idea of
super intelligence, which is this term that's been bounced around
in the AI space. For a while now, these companies
Meta included, but also open Ai and Google have been
obsessed with creating this thing known as artificial general intelligence
or AGI. Basically, this is an artificial intelligence that matches

(01:26):
the intelligence of a human brain. Tech leaders have said
that achieving AGI will be this huge transformative moment when
we can effectively hand over much of our lives to
the world of computers, and we are going to live
in this world of abundance, right and this will be
this sort of tipping point when we achieve AGI technology.

(01:46):
Superintelligence basically takes that concept and goes one step further.
It implies that we're creating an intelligence that is beyond
the capabilities of the human brain. But I'm gonna throw
some cold water on all of this. Despite the fact
that these terms been talked about a lot, they're both
kind of loosely defined. The goalposts for what constitutes achieving

(02:06):
AGI quote unquote have moved quite a bit, but even
still major figures in the space are concerned with it
and are focused on it. Sam Altman says we are
going to achieve AGI by the end of Trump's current term,
and Dario Amada, who is the CEO of Anthropic, another
AI company, says that we're going to see major disruptions
from AGI in the next two to three years. Clearly

(02:30):
Meta doesn't want to be left behind, and that seems
like why they're making this major investment and spinning up
a super intelligence lab.

Speaker 1 (02:36):
Michael, this is more of a tree tops maybe philosophical
question for you, you know, and you mentioned this. The
we continue to race to create computers that rival or
surpass the human brain, then what I mean? Then? I mean,
are we just going to I mean I don't even
remember like my wife's phone number now, I mean some

(02:56):
basic things that we are already Like, I feel like
our brains are going to act feet in the next
twenty years because we won't need to use them. So
once we achieve these these supercomputers and the AGI and
the super intelligence labs, then what then what happens to us?

Speaker 2 (03:12):
So if you listen to the sort of leading lights
in the AI space, they'll tell you that it's going
to be great. Right, we can just kind of kick
back and you go on vacation while the computers or
handle kind of the rote repetitive work of everyday life. Right,
They're going to be healthcare administrators, lawyers, politicians, and that's
going to be kind of the realm of artificial intelligence,

(03:34):
and we can kind of just sit back and relax.
But there's reason to be skeptical that like AGI is
possible at all, at least in the sort of timeline
that a lot of these figures are laying out. Recently,
Apple put out a report that kind of cast doubt
on whether these large reasoning models, kind of the cutting

(03:54):
edge large language models that we're seeing in play right now,
are actually as smart as they claim to be. Right.
They found during this report that these models actually kind
of get dummer over time, they make more mistakes as
they go along, and similar reporting from Arizona State University
seems to back those claims up. So before we get

(04:15):
into the sort of what the world would look like
after we achieve AGI, I think it's important to really
reckon with the question of whether AGI is even possible
in the first place.

Speaker 1 (04:25):
Well, I'll tell you right now, all I'm using AI
for is to plug in a couple of ingredients I
have in my fridge and it creates a recipe for
me and think that it's sandy for that.

Speaker 2 (04:33):
That's a good application of it.

Speaker 1 (04:35):
It's pretty slick. And yesterday in New York, the Trump
Organization announced that they're going to get into the mobile
phone industry, that with a new phone too. What's the
story there.

Speaker 2 (04:45):
Yeah, that's kind of the attention getting piece of this
story is the new Trump phone. But in your right.
In New York, the Trump Organization, which is run by
Don Junior and Eric, the president's sons, announced that they're
getting into the telecom space. That they are a long
Trump Mobile, which is a new phone plan, a mobile
service provider, and it's essentially an MVNO, a small mobile

(05:08):
service provider that basically piggybacks off the large telecom companies.
In this case, it's T Mobile. So if you sign
up for Trump Mobile, you're getting the coverage of T
mobiles network and you're going to be paying forty seven
forty five a month to do that. That is, you know,
your rebutting the pieces together, and that is a little
bit more expensive than some of the other players in

(05:29):
this space. Mint Mobile Cricket Wireless both offer kind of
cheaper plans with similar coverage also on T mobiles network,
but the Trump plan promises telehealth services and roadside assistance
some extra bonuses. We don't really know what that's going
to look like in practice because the plan hasn't actually
launched yet, so reason to kind of maybe be a
little wary there. And though they say this will work

(05:51):
with pretty much any phone on T mobiles network, they
are offering their own sort of you know, primary device.
It is called the T one phone. T one Phone
eight thousand and two gold version is actually the full name.
It's a six point eight inch Android device, And it
is kind of hard to put together exactly what this
phone is because we don't have a ton of information

(06:14):
on it. There's a rendering on the website that is
trimmed in gold. It has a big T and an
American flag on the back. It has a three camera array,
but weirdly no little sort of point for the flash,
which is what you would expect on a phone like this,
So it seems to be a computer generated image, but
you know it is, you know, five hundred dollars. You

(06:35):
can put down a one hundred dollars reservation for it
right now. But you know, when it gets into sort
of things like processors or you know, what kind of
the the battery life is going to look like We
just don't know yet. They do say that it will
be built in the United States, which has been you know,
you know, a real priority for the Trump administration and
for the Trump organization. But as of right now, there

(06:56):
are no mobile phones that are built in the United States.
So we're gonna have to see if that five hundred
dollars price tag stays where it is.