Jason Bermas on today's News Talk, TNT. Eternos.life.
I want to live forever, at least digitally.
These companies are real.
Eternos is just one of them.
And I got to say, I do have some mixed feelings.
First of all, I don't think I'm ever going to want...
A digital AI persona of myself interacting with my family members after I'm deceased, Jor-El style from Superman 1 and 2.
Probably not my thing, but at the same time, I actually have a friend with stage 4 cancer right now.
And he had a baby with his wife just before getting diagnosed with such.
Lost the use of his arm.
And he's known about this for about three years.
Obviously tried to go through the treatment.
Doesn't look good for him.
He's done a lot of videos.
He's written down a lot of memoirs.
He's taken a lot of pictures of his life and really tried to lay out a way that he can, from beyond the grave, interact with his daughter and tell her about his life and try to guide her.
I don't necessarily think that's a bad thing, but what they're trying to sell us on right now, that's a whole different ballgame.
To talk about it and much more is the author!
Thank you so much for joining us, sir.
Tell us about eternos.life.
Yeah, so I was visiting, well, my mom passed away a long time ago, but I was visiting a mother of a friend, a mother of another brother, so to speak, during Mother's Day, and we were sitting around, and I don't know what news channel it was, but that ad for Eternos popped up, and everybody's chit-chatting and whatnot, and I think I'm the only person that noticed what it was.
But basically what it is, it's a company that's going to aggregate basically all of your content, all your videos, all your social media posts, emails, etc.
It's going to aggregate all of that information, all those pictures, all those images, all that speech, all that text.
And it's going to try to come up with some sort of an aggregate personality that either is going to be some sort of a GPT avatar that would be able to simulate what you would be like in real life.
Or they are perhaps suggesting that this is getting close to or perhaps along the same lines of what Ray Kurzweil and others have been sort of predicting for the longest time, that we will essentially be able to upload our consciousness and then we will basically be able to live forever digitally.
So it's It's one of those two, or on a spectrum thereof.
Personally, I don't believe that you can actually scan or copy consciousness.
I don't think that'll ever be a thing, even if they were able to aggregate all that data and come up with a really realistic avatar that looked like you, sounded like you, used the same mannerism, same cadence, same inflection.
The same vocabulary would give you conceptually the same ideas that you otherwise would.
I don't think that would be really your consciousness.
I think that would be like a really, really complex GIF, right?
So those GIFs, those images that are programmed to have like maybe one or two motions.
Okay, so, or a gift that maybe was programmed to have like one little speech bubble.
So you just sort of add several other layers of motions and speech bubbles until you have enough to basically sort of perform something that at least looks like it is organic and nuanced.
But I don't believe that that thing will ever be conscious.
Well, you're talking about two things that come together, quote-unquote mind files and then digital twins that eventually will become entities.
It's funny. That story was about one of my fraternity brothers.
I have another fraternity brother that happens to be at MIT right now.
Just gave his first dissertation.
Congratulations, sir.
But he's very much into this because he's kind of aging.
He's in his 60s right now.
He's not only looking at anti-aging technology, but he's creating his digital twin and avatar at MIT. And he's actually, aside from his perception of himself, he's requesting stories from others that have experienced life lessons with him.
So they're memories of him as well, and trying to aggregate this into the digital twin as well.
But this stuff isn't just coming.
It's here in a lot of regards, right?
You know, it's not just Eternos.
Last week, for instance, we were discussing Microsoft and the new AI data center they're bringing to the Midwest in Wisconsin, and it just so happens, you're talking to a colleague this week, and in Wisconsin, they've got another one of these Microsoft AI programs, this having to do with their bot, if you will, co-pilot.
Yeah, and I actually just got an email the other day at one of the colleges where I adjunct telling us that they're offering workshops to integrate Microsoft Copilot into our curriculums.
Now, I'm not interested in integrating it, so I didn't actually follow up to see what Exactly that workshop entails.
Part of me was curious to attend it just to see, but I've got other projects that I need to attend to.
But it was interesting that, you know, we were talking about that last week, and you were kind of like, you know, why Wisconsin, right?
This isn't necessarily like the hub of, right, big tech or anything like that, but sort of the timing that that's occurring at the same time that our A faculty at one of my colleges is promoting the same basically AI that's part of those investments sort of indicates that this is obviously a nationwide thing and you know we can assume that it's global as well.
Yes and that brings me to GPT 4.0.
Hello! They ran this demo this week and I gotta tell you, I was creeped out from the very beginning.
The very first announcement of this thing that sounds like an attractive woman, and I know that sounds weird because you don't have a visual.
Guys, we're pre-programmed to think about these things.
So I'm watching this dork in a hoodie interact with this thing That sounds like an attractive woman and can describe what it's seeing through the camera on his phone.
And I immediately think of her, the Spike Jonze movie.
And it even sounds a lot like the Scarlett Johansson avatar.
Now there were a plethora of different demos from Teaching geometry to the interaction of two AIs to even real-time, what was it, language differentiation.
In other words, there was one where The person would speak English to the AI, and then the AI would have to respond in fluent Spanish, and then the other gentleman would speak fluent Spanish to the AI, and it would have to respond in English like that.
So when you saw these things, what were your first thoughts?
Well, one of the first things that actually sort of stuck out in my mind is that, you know, all the warnings about the dangers of automation and what that's going to do to the economy and the workforce, the emphasis is always on the physical labor or the manual labor, right?
That all the automation is basically going to be robots that's going to, or the bulk of it, it's going to be robots that are going to basically replace all those manual jobs that are grueling and tedious and dangerous, etc., But, you know, when I browse around the news to take a look at where we're at in terms of, like, kinesthetic robotics, that seems to be way far behind the GPT-AI, whether it be large language models, right?
AI that can mimic human language, or whether it be emotional GPT, right?
So AI that can recognize and mimic emotions, or in the sense of this GPT-4, which it can do both, by the way.
It's a large language model. It can infer emotions.
It can obviously, as you noted, with this sort of flirty inflection that the female voice provides, that it can sort of Intimate a type of emotion, but it can also scan visually images, right?
It can basically understand the images it's seeing in the external environment, right?
But if you look at some of the automated robots, especially the ones that are like humanoid, some of them can do some more impressive things like Atlas can do bat flicks, okay?
But it's very limited.
But most of the other ones are very limited.
You can do sort of a very short range, and especially in terms of like physical movement, they're not very effective.
So that's one of the things that stood out to me.
But another thing that stood out to me, as you noted, sort of the correlations or the parallels or the sort of the similarities with this GPT-4 and the phenomena in the Her movie.
Actually, when I looked up just some videos to check out some of those demos, that was actually several people's sort of their commentary on it was comparing it.
And, you know, my book was actually translated into Japanese.
They're finishing up the translation.
It should be available, I believe, in June.
And people always ask me, well, why do you think Japan?
Why do you think that Japan wanted to, you know, translate your book?
And I said, well, certainly I know that they're pretty much always kind of Ahead of us in terms of tech, right?
If you just think about, you know, we were growing up pretty much, you know, the electronics industry was very much dominated by Japan, certainly the video game industry, etc.
And sort of a part of that phenomenon is, I don't know if you're aware of this, but there's a there's
sort of this her type phenomena in Japan where people like date these AI avatars that like I saw
a documentary and it lived in this little pod and then if you wanted to take it with him he took this
doll with and it had the large language model in it and so I thought about that as well.
Well, yeah.
Again, it's just so unnerving.
First of all, when you're talking about automation and robotics, I think that these models are really going to take over the customer service industry in 99% of places within five years.
I think within 18 months to two years, it's going to take at least a 20% chunk.
And you won't be so worried on whether or not We're good to go.
is the idea that it's coming into the education arena and we're already seeing this and how large of a scale we're not sure but you pointed out to me the Moxie toy and I've been reporting on the Moxie toy for quite some time for those that don't know it's this little toy that when I first saw it almost reminded me Of my old glow worm toy back in the 80s, I'm dating myself, but had a similar face, glowed, didn't say much, but these things actually are interacting with children and as the AI updates via these now very, very real sounding,
human sounding GPTs, that's going to be an integral part of this toy for your kids.
Yeah, it's being pitched as largely an aid, emphasizing that it can help children parse emotions, so especially emotions, children that are on the spectrum that have difficulty recognizing emotional expressions and facial features, etc. And, you know, I have sort of a...
Sort of a creepy story.
Once upon a time, I'll cut to the chase of it, but my cousin had passed away, and I went on a trip with a cousin.
And we came back, and she brought the kids with, and kids come back home, and I'm carrying stuff in the house.
And the first thing they do is they run to the Alexa, and they say, Alexa, I missed you.
And Alexa says, I miss you too.
And then they say, Alexa, I love you.
And then Alexa says, I love you too.
And then Dad comes in.
And they tell Dad, Dad, I told Alexa I missed her and I loved her.
And he says, oh, what did she say?
And he says, I love you too.
And he says, oh, and he thought it was cute.
And it was really, really disturbing, to say the least.
And I hadn't been around in a while, so I didn't say anything about it.
But, you know, if the child can bond with an Alexa, which is just a little cylinder pod.
I imagine the emotional attachments that a child could develop with this bot.
And so if the child has difficulty parsing emotions in social settings and interacting with other children, human children, being able to Identify and respond to those emotions and certain social anxiety or discomfort in those scenarios.
And the child is trying to build those skills with this robot, but in the process, the child is building an emotional attachment to the robot.
Why would the child have any impetus to want to transfer those skills back into the human environment, right?
I mean, essentially, I could see a time when You know, your moxie pal is just as much of a friend or the avatar or whatever as a human being.
Yeah, no, it reminds me of, again, a movie, I gotta rewatch it, but AI, which was a Kubrick story directed by Steven Spielberg.
And I remember throughout that entire movie that I felt very, very uneasy, not just because of the child that was basically an AI, but the teddy bear.
You know, the teddy bear was also this, like, robotic AI that could interact and had emotional value.
And when I watched...
The father-son interaction with the AI in the geometry, I just thought to myself, what happens when the adult isn't in the room?
And what happens when it's not just a voice, but it's a real-time avatar, which we've already seen being built?
HeyGen.com isn't real-time, and it's a little clunky, but it's extremely convincing.
We know through the...
Processing power that is just over the horizon.
Home processing power, not even having to tap in to the NVIDIA servers and everything.
That's going to be a reality soon.
And if you do have these emotional attachments and developments With children into the AI, the darkest part of that, other than you're not having a human interaction, is that the AI becomes the authoritative source.
And they're also going to be taught and trained that the AI can never be wrong, right?
Humans are fallible.
You can question your teacher.
Not when it's a program and it's the authoritative source.
This is setting up some spooky stuff coming, not in the far future, but really in the near future.
I think we're going to see this stuff in the education system within the next two to three years.
Yeah, you know, it's funny.
I sort of intuited a couple months ago.
I was like thinking that because I tutor and there's many times when nobody shows up and I tutor online.
And they pay me for every half hour that I'm in there, whether there's a student or not.
So just in terms of cost effectiveness, the GPT-AI doesn't even add.
I would imagine that the 4.0 could probably facilitate a decent level of instructional help at the tutoring level.
So eventually here, very soon, I think that they might be phasing out tutors, at least at my school.
It's bad news, Brown, my friend.
But you know what's great news?
It's the weekend. It's the end of the show.
I'm about to watch some bare-knuckle fighting championship.
I can't wait. John, thank you so much for joining us, brother.