- 10 hours ago
Category
🥇
SportsTranscript
00:00Joining me now, Maya Vujinovic, CEO of Digital Assets at FGNexus.
00:04Maya, thanks so much for joining me at the desk.
00:06It's great to see you again, Caroline. Thank you for having me.
00:08So we're talking future of work, AI disruption.
00:12We hear a lot of talk about AI being a productivity booster and a job killer.
00:17What do you think? What narrative will win out?
00:20I think in the past year, a narrative of a job killer has definitely won.
00:26I would say we're slowly starting to see that it's a job booster.
00:31It will augment certain types of jobs, right?
00:34Certain types of jobs may go away, but in about 10 to 20 years, right?
00:39At the next kind of short-term span, year, two, five years,
00:43I see it more as augmenting our lives and helping and being a productive asset than anything else.
00:48Let's dig into that a bit more.
00:50How do you see AI changing the way we work over the next five to ten years?
00:55Yeah, so right now you have AI in healthcare that is really impacting healthcare
01:01in diagnostics and MRIs and, you know, research.
01:05You have AI right now really impacting our transportation.
01:09That's kind of obvious.
01:10You have in finance, in algorithmic trading and et cetera.
01:14You have in HR, in legal.
01:16That is quite here.
01:18Now, is it here at scale in those last two that I mentioned?
01:21No, but I think in the next two years, you're going to start to see assistants of assistants.
01:26You're going to start to see legal research on AI.
01:30You're going to start to see definitely HR and companies trying to push more of AI tools
01:36onto their employees to learn and be more productive and kind of make them superhuman in a sense.
01:43Can you give an example of AI at work right now that maybe we wouldn't even notice?
01:47I mean, co-pilot, Microsoft co-pilot is something that is ChatGPT.
01:51I think most people take that for granted, but think about it.
01:54Everybody I speak to will say, I don't really use Google anymore.
01:58I just go to ChatGPT and I, you know, do everything I need to do in legal.
02:03I obviously all those things have to be double checked.
02:06But there are people that are using ChatGPT today for $200 a month to do medical diagnosis,
02:12to double check medical diagnosis that they've been given from their doctor.
02:15I mean, that is an incredible advancement of what we have.
02:20And it is pretty cool.
02:23And it brings us kind of forward and makes us more efficient in many ways.
02:27When it comes to automation, though, how do you balance AI with also having a human touch?
02:33Because as someone who leverages ChatGPT often, it's wrong a lot.
02:39Yeah, absolutely.
02:40You almost have to tell it that it's stupid and don't be stupid in order to give you good answers.
02:44Look, jobs like yours, for example, or jobs like in therapists, in school teachers, those will not be replaced anytime soon.
02:54I don't envision a robot in your place, you know, interview me anytime soon.
02:58So I think things will evolve, but they are not going to just suddenly change in the next year or two.
03:05I think a lot of people out there are making a lot of money by saying we're all done, right?
03:10But reality is there are certain emotional things that we still don't have with AI and we won't have.
03:15Now, I think Dennis from DeepMind, Google DeepMind has predicted that, and they've been quite consistent actually with this.
03:22And in the next 10 years, they'll probably have AGI.
03:25Now, there is a thing where you can say, how do you define AGI, right?
03:29But I do kind of turn to them quite often and turn to him to kind of tell me when does it become really serious when AI starts to develop these emotions that humans have.
03:41Still, my prediction is we won't be seeing any of that at scale for another 20 years.
03:46You make up a good point, though, because when we first chatted months ago, you told me I was being too nice to AI, and you said you need to almost tell it it's stupid.
03:57And I told you I was saying please, and you said don't say please, it will work harder for you if you're mean to it.
04:03Explain that to me.
04:04Yeah, good memory.
04:06So I noticed when JustChatGPT came out that I was saying, okay, now let's do this, and please, could you, right?
04:13And I realized that every time I said that, it was kind of repeat itself.
04:18And so I think that's a basic fact now.
04:20Everybody knows this.
04:22I mean, you have to almost tell it, please don't be wrong.
04:25You know, you're acting pretty stupid.
04:27This is, you're not AG, you're not AI quality.
04:32You're, I've even used prompts like saying you are a low-class human.
04:37You do not have a good education.
04:39You need to act like AI.
04:41And it would actually dig deeper into answers.
04:46I don't have an explanation of why that happens, but I can only tell you a sci-fi answer, which I do think that, you know, at some point you will start to see a bad AI and good AI.
05:00And that will depend on how we train AIs.
05:02There will be some bad actors in this world, and I do think that AI is almost capable of sensing when something is good and something is bad.
05:11And so I do see a future where we will have AI that will act in a way that is not going to be good for humans.
05:19So I know most people don't want to believe that, but maybe I watch too many sci-fi films, but I do think it's not anytime soon.
05:27However, though, it would all depend on our training of AI.
05:30Well, you say that we're in act one of AI right now.
05:33What can we expect in act two, hopefully not these bad actors?
05:37And what do we need to do to prepare for that?
05:40Yeah, so act one, when I refer to that, I think it's infrastructure, right?
05:45It's solid infrastructure, electricity, chips, right?
05:50Everything that we're seeing now and what we've seen within NVIDIA.
05:53I think act two is applications.
05:56I don't know if Dennis is correct.
05:59If in the next five to 10 years we start to see AGI, then that act two may start to see some bad actors through those applications.
06:08Because remember, a lot of the AI is trained on real data, but it's also trained on synthetic data.
06:15And synthetic data can be manipulated.
06:17I'm not an expert in synthetic data, but from the basic understanding, you know, that act two, you could have a lot of bad actors.
06:24And I think there are other issues regarding AI really picking up steam.
06:33And one of those is just, do we have the power to support all of these, you know, all of this technology?
06:41And do we?
06:43Yeah, it's one of the most important questions you ask.
06:45And I think every single TV channel should be asking this question about electricity.
06:50I think what we've seen in the last couple of days with the fight with China and minerals and all of that, you know, we should be doing the same around electricity.
06:59There should be a really a large awakening around how much electricity we need for AI.
07:04Just in the next year or two, we are going to need as much as a whole New York City be put on a grid.
07:11That's how much electricity we would need just for the U.S. consumption in AI.
07:15And so globally, that's a lot more as well.
07:19And so I think we need to be, and I think this administration, by the way, has done an incredible job of focusing on that, right or wrong, in many people's eyes, depending on, you know, which take you align with.
07:30But I believe that we should be pushing on electricity development all the way.
07:35And I think they've been correct when it comes to that.
07:37Is that even possible, though, putting a whole New York City on a grid in the next one to two years?
07:43Yes, but it's extremely fragmented.
07:46I think it is possible, but our U.S. grid is fragmented.
07:51Our infrastructure is fragmented.
07:52And I think if I was anybody in AI space and looking to invest, and I think you've also seen in the last, you know, four or five days, there's been a lot of discussion of are we in AI bubble?
08:04Are we over-investing in AI?
08:05And I said it the other day on CNBC, depends what you're investing in.
08:10If you're investing in infrastructure, then you need to be doubling down on that.
08:14If you're investing in applications that actually are backed by infrastructure, real utility, real impact, then you need to double down on that.
08:21We are in the real race with China.
08:23There is, there's, it's not anymore a question.
08:26And I think when, when people come and say, well, you know, U.S. will win in culture and, you know, U.S. dollar is dominant, I think we live in a bit of an illusion.
08:35You know, there is U.S. and there's China, and then there is China with the rest of the world.
08:42And there is U.S. kind of on its own.
08:44And so, so I think we need to start thinking about it that way.
08:47And really, I think there is something that Dennis mentions, I think he mentioned it a couple of years ago, even, where he said that even that 10 minutes of 15 minutes, who comes out with AGI, that first 10, 15 minutes is really going to kind of lead for years to come.
09:05Who's leading right now?
09:06I, look, 90% of AGI companies in the world that we're using are U.S. based.
09:14Okay.
09:14Now, a lot of those scientists that work in those companies are Chinese.
09:19And so, or are Chinese nationals, right, or are U.S. nationals that are from China.
09:24And we don't know if they're going to go back.
09:25And I think this is really, to me, this should be a core question of where we're at.
09:30I know it's gotten to that level, but that is the level we've got to discuss these things.
09:35And so, I think China leads in industrial and China leads in the rest of the world in covering AI.
09:42But I believe still U.S. is in a little bit of a race because 90% of companies that people use around the world are U.S. companies in AI.
09:52And that's why I wrote a piece about a month ago, two months ago, where I called it an AI dollar, basically saying that because these companies originate in the U.S., right, all of those subscription services have to be paid in the U.S., right, they all have to settle in USD.
10:06And therefore, you know, that's another way of doing diplomacy and pushing U.S. dollar globally.
10:11Okay, so just finally, from an investor perspective, not that you're a chief investment officer by any means, but how should investors be thinking about this as an opportunity?
10:22As you mentioned, there's the questions, is AI a bubble?
10:25But should they really be homing in on the electrical play?
10:30Should they be focusing on investing in China or here in the U.S.?
10:33What should investors be doing right now?
10:36Yeah, I mean, it depends what your mandate is, right?
10:40Some people don't have a mandate to go beyond infrastructure.
10:43Some people have only to focus on applications.
10:47I think if I was an investor, I would focus on infrastructure.
10:50It's a lot harder and a lot longer play.
10:52I would look at applications that have to do anything with legal, with trading, with health care, certainly, with HR.
11:01Everything that helps the bottom line of a company, that is going to increase the bottom line of the company, look for those applications, because that's exactly where they will have an impact and they will do really well.
11:16All right, we have to leave it there.
11:17Maya, we hope you'll come back and talk to us more soon, because you have so many interesting insights.
11:21Thanks so much.
11:22Thanks so much.
11:23That's Maya Vujinovic, CEO, Digital Assets at FGNexus.
Recommended
2:30
|
Up next
0:42
1:05
0:50
0:32
0:40
1:00
0:57
8:41
14:33
3:49
2:07
0:53
0:43
Be the first to comment