Bloomberg Audio Studios, Podcasts, radio News.
You're listening to Bloomberg Business Week with Carol Masser and Tim Stenoveek on Bloomberg Radio.
Well, it seems like everybody is talking about AI, but the question is is everybody using AI? We talked a little bit about this with Vinnie Catalano. Yeah, he said he's like going all in on us.
He did, Yeah, which is kind of funny.
It turns out my takeaway from that conversation is I'm no Einstein.
But neither are you.
No Exactly, Well, it turns out there are some stark differences when it comes to who is engaging with the tech, and it actually falls along gender lines. Check out this research Carol Harvard Business School Associate Professor Rembrandt Koning found that women are using new AI tools at a lower rate than men, about twenty five percent, according to the research cited in a story last month. In a post from Harvard Business School, curious what Jenny Troutman thinks about this.
She's director of Training and Certification at Amazon Web Services. Of course we refer to it as AWS. She joins us from Houston, Texas. Jenny, I know you watch the way that this falls along gender lines really closely. In your position at aws Why do you think the adoption among women is so much less than men?
Right?
Well, yeah, so it's interesting. In technology in general, women only make up about thirty percent of the workforce, but it is more stark in AI, with I'm only comprising about twenty two percent. We did do a study though, that shows that women are interested in learning about generative AI.
In fact, in the.
Study, it showed three out of four women are interested in learning how to apply generative AI at work, and eighty percent actually want to apply for jobs involving generative AI. So while I think we're seeing a gap right now in the workforce, we're not seeing a gap in interest. And I think the issue is confidence in skills, and
that's really why I'm here. What I'm focused on is how do we at AWLS help the global workforce and women across the globe get confidence working with AI and generative AI technologies.
Well, you know, this is something I'm glad we're talking. We actually had an earlier conversation about AI and the workforce. And you know, when anybody says, you know, we're hiring a lot of AI related jobs and it's AI related skills, and like, what's an AI related skill? Because to be quite honest with you, Jenny, you know, Tim and I play around certainly with like chatchipt and things, and part of it is like our job. How do you ask
the right questions right and drill down those questions. I mean, I know that's just basic level, but I mean, what what do you see as AI skills? Be it for men or for women?
Yeah, so that is a very big question.
Sorry, when we talk about when we talk about AI skills, there are kind of a breadth of them that we're talking about.
To your point, we're.
Talking about people who are just using the technology, which the foundational skills really are what we need. So things like are to the possible? What is generative AI? And what can you do with it? Because one of the one of the biggest roadblocks to getting started is really understanding how you can use it and thinking differently about what you do. So I'm you know, understanding use cases,
places to get started, where to play with it. And then the second is exactly what you just said, which is what we call prompt engineering how to write good prompts, And that's relevant for anyone looking to use the technology, no matter their role technical non technical alike, if you don't write a good prompt, you don't get a good output from the from the technology, and so that is a skill in and of itself. In fact, we offer
free training online on a to a skill builder. We can go in and take a course on how to do that.
Isn't that funny that?
Like, wait, can we take that course?
Yes? You can. It's for free. You can go online and find it. Today.
Carol's like so excited about learning how to do prompts so she can get away from her cohed.
We want to say, no, I don't want to get away from you. But I mean everybody has said that learning how to use it is like learning how to prompt and ask correct.
Well, that's why the turn, that's why engineers, that's like why it's such an important skill. I think one cool thing about this, though, Jenny, is you don't even need to be like anyone. Anyone can I've never taken a course on prompt engineering. I'm not going to pretend that I know what I'm doing. But it's such a forgiving product where if you don't get the answer you're looking for, then you change the way you ask the question or you refine it, and you ultimately get there.
Now it's not like a co host who says that was a stupid question.
Never yet, well, I bet I would do that by the way, it probably would.
I would say yes, and yes, that is very true. And in fact, a big part of using generative AI is validating the What you got is what you were looking for. The problem comes in if you weren't exactly sure what the output should be. So, for example, if you're using the tool beyond writing a document for you, but maybe to do some data analysis and you ask comparison, if you don't actually do the right prompt, the results of the data analysis you get might not be correct, and unless you a deep.
Validation, you might not know that. So that's where it becomes important.
But part of learning the technology and getting good at prompt engineering is to your point, practicing it, doing it over and over seeing what kind of results you get, and constantly refining.
I'm looking at your background. I didn't know you were a pilot. That's really pretty cool stuff Like no, I just you know, I think about how the world has evolved and continues to evolve. How are you thinking about and kind of all that you've done in your world, and like, as we it just feels like we're just scratching the service in terms of what generative AI and LMS can do for our world. And I know that there are a lot of concerns about it going off in the wrong direction, and that's all.
You know.
We could spend hours on that, but I'm just curious about from your vantage point, what you've done professionally, like how you are looking at this moment in time.
So we are actually in the same place everyone else is, which is trying to figure out how we do our own jobs differently and then using that to then help educate people on how to think differently for other roles. I would agree we're at the very early stages. And that's why the lack of skills in the industry around AI are so prevalent, because, like I mentioned, it's not
only the foundational skills the other areas. We need deep technical experts who can build good models, who can build good tooling on top of the models, and all of that is a work in progress right now. The tooling is getting better day by day, but as the tooling gets better, we need to get better and better at how we use it. So I would say we're still at the very beginning of figuring out how much this is going to change the way we work. And I
take it back to thinking about the Internet. I am old enough to remember when the Internet became big and we had no idea how much it was going to change the way our day to day lives and worked, and how we looked for things, how we found things, how we thought about getting from place to place. We're in that same same place. It's AI and generative AI is the one thing I can guarantee you is the way we do things ten years from now will be very different from the way we.
Do them now.
Yeah, but it's hard to predict exactly how quickly things will change or how much they will change.
I'm with you. I remember doing a premer on what is the Internet like early on in my career.
What did you find out?
The answer?
But it wasn't around.
It wasn't around.
No, I want to go back to you being a commercial or having a commercial pilot license, Like when you say, in ten years from now that you know generative AI and what we were talking about today is going to dramatically change our world, Like do you see AI ultimately handling you know, systems like air traffic control? Like how far does it go? Or I know it's hard, it's like who knows where it goes? But I mean, how do you think about it? Like on what scale?
Yeah?
The way we think about it is anything that can be automated will be. So anything that you can write down a word for, if it's not already automated, it will be. And the beauty of that is the things that we can't automate, like our creative thinking, like taking that and making connections with other things innovating. That's the sort of thing that doesn't automate. And so where I think it will go is we will all have AI assistance that will be doing kind of that automated repetitive
work for us. We will be guiding those agents or those assistants, and we will we will be innovating so much faster than we can even imagine because we've never had a machine in the back end doing all of the grunt work for us.
That's a down stall.
I think Jennie Diamond, I'm quoting him again, but I believe he said our kids are going to live to one hundred not have cancer as a result of these tools and be working three and a half days a week.
Is that true?
Your guess is as good as mine.
It sounds like a pretty great world.
Well, I look back before cloud and before the Internet, and we didn't work less than Yeah, so it's hard for me to imagine.
I think some people would say we work more because we always have these computer things attached to our hands always on.
Might say that, and it's hard for me to imagine, just because of human nature, that we will do less than we did.
In the past.
But but I do think that we will be able to do a whole lot more, a whole lot faster. And that's that's the part that's hard to imagine. I have a I have a daughter who's in high school and has wanted to be a doctor her whole life. And that's an area where you know, it's already using robotics, and you can imagine a lot will change it. So I question, you know, what is the what is the place to be there? But there are always going to be people needing to guide the machines.
I think, and yeah, exactly, And I think, I mean, I think radiology is one profession. That's certainly right for disruption when it comes to AI, but maybe dermatology not so much.
Fair Yeah, I think where you have to have kind of the human intervention, there will always be some level of human intervention needed. And the question is how much better can we be? How much less time do we need to spend in a doctor's office because they're able to solve problems automatically for us? And then how much will they be able to accelerate the time, you know, spend more time on innovating the solutions for us?
Yeah, I know.
And people talk about drug discovery right, being able to kind of go through a lot of information and figure out things.
What hey, Jenny, before we let you go, since you are in aviation, Also, do you think we'll in our lifetimes we'll see commercial aircraft with just one pilot rather than two.
You know, I can't predict the future, but I don't know.
Why not, whereas like maybe on the ground they're doing some of the controlling or even.
Already have so much unmanned in the military, unmanned spacecraft or sorry, aircraft spacecraft too, But I don't know why we would continue to need as much on the plane with the exception of the backup factor. Something happens to one pilot, you definitely don't want to have nobody.
Yeah, yeah, that's true. Or they can control it on the ground, who knows.
I guess I don't know, like a remote yeah, like a like a drone or something, right, exactly, exactly.
Listen, this was what I saw my car drive for me, as though, to be fair, I'm.
Not ready for that.
I'm ready.
Oh my gosh, you got to get in one of these waymouth things.
They're amazing. It was awesome.
Jenny.
Thanks.
Jenny Troutman, director of Training and Certification at AWS, Amazon Web Services, joining us on this Monday from Houston, Texas.
Well.
The Trump family is set to launch a bitcoin mining focused venture with HUT eight. It's the latest expansion of the US President's ties to the cryptocurrency sector. HUT eight is launching the American Bitcoin Corp. As a major already owned a subsidiary focus exclusively on bitcoin mining and quote strategic bitcoin reserve development this company. The company said this in a statement today. For more, let's go to Fred Tealey's chairman and CEO of the four billion dollar market
cap bitcoin minor Mara holdings. He joins us from Miami today. Fred, good to have you back with us. I want to talk the market, bitcoin mining, the health of the industry as you see at big changes. But first I want to get your reaction to this news that the Trump family is teaming up with hot eight. What do you make of it?
Oh, I think obviously the president and his family are very positively inclined to bitcoin. So bitcoin mining is a key part of anybody's crypto strategy if you're looking at the industry overall. So I think that the fact that they're interested is a good sign. Obviously, I think there could have been a better choice for partner for them, but that's just me being biased.
I understand. But do you think that there's a concern that he's a little too close in the sense of, well, if his firm is partnering up and picking partners, then he can pick winners and losers.
I'm listening. I'm a big believer in free competition and open competition in the marketplaces, and you know, my expectation is that you know, there won't be any preferences played in this marketplace. You know, the great thing about bitcoin mining is it's really a business where we all compete with our compute and our hash rate, and I think that, you know, we'll just have to see if this company
has given preferential treatment. Then I think I'm sure the media industry will cover that and in its a typical way and bring that to light, and that point it'll be up to the public to kind of deal with it.
And do you think the president should be in this business?
Yeah, I'm not going to pine on that. I think the what you know, members of government do and don't do. I think it is up for the government to police and make determinations on that. You know, there's so many people who have been in very high offices that have had financial dealings big businesses, and even though the rules call for you to kind of step away from those businesses,
are they really stepping away? I mean, I think you can go back over history and see that this wouldn't be the first time that, you know, members of high office have fingers in businesses and influenced businesses.
So let's go bigger picture here and just talk a little bit about how you view the administration with regard to crypto.
If we were to.
Speak a few months ago, maybe we'd see the eyes in the price of bitcoin and talk about the optimism that the industry had for a Trump administration. Are you seeing that promise fulfilled at this point?
Absolutely? You know, I spend a lot of time in Washington these days. I was at the White House Crypto Summit as the only bitcoin minor representing the industry. I think as we look at the things President Trump laid out in his kind of campaign promises at the Bitcoin Nashville conference, you know, obviously changes at the sec Bitcoin Strategic Reserve. You know, more bitcoin mining should be done in the US. You know, a lot of those things have come to pass. I think we're going to see
market Structure Bill be introduced here shortly. We're going to start seeing further process progress on the Strategic Bitcoin Reserve. We'll see a variety of other things around stable coin legislation as well. So I think you're going to see a lot of progress to make up for the years during the Biden administration when nothing happened in the industry was attacked. So you've got to think about it. We're not even one hundred days into his administration at this point, right,
So no, that's fair. A lot has already happened.
Hey, Fred broad Now just remind our world. I mean, you're publicly held, you're almost a four billion dollar market cap company. Remind our world exactly what your business is.
So we're in the business of transforming energy into digital gold. We consider ourselves really a digital energy company. So we take energy which otherwise isn't necessarily being used. So if you think about a lot of renewable energy assets like wind farms, a lot of the time that energy can't be sold into the marketplace because either there is too much supply and so the market is flooded, or there's
not enough transmission capacity and it's called congestion. And you know, the renewable energy boom that happened in this country over the past decade has created an oversupply in certain geographies. And what happens is that supply can't get to market and so it sits idle. So we can go in and take that excess energy and move demand to where supply is. Energy consumption is not a constant flat amount all day long, so the energy companies can just plan
to have a flat amount of energy. It varies all day long. From the morning it goes up and then it goes bottoms out midday, and then it peaks around nine o'clock at night and then drops again. And while you can plan for those flows, what you can't plan for is the weather because that affects solar farms, it affects wind farms, and that intermittent type of energy generation is very challenging to plan for. And so the only energy that really is consistent and constant is nuclear, coal,
natural gas. But all of the renewable energy forms really are what's called intermittent, and so our industry supplies what's called flexible load. And Duke University just put out a study a few weeks ago where basically, the whole AI industry in this country needs about forty eight gigawatts of power over the next few years, and yet our grid has seventy three gigawatts, nearly double what the AI industry needs.
But it's only available if the load is flexible. And we're flexible because we can shut our operations down at a moment's notice when the grid needs the energy. The AI industry has a harder time doing that. So we're a kind of a unique animal in that, but we play a very vital role in balancing the grid.
So do you have a good do you feel like tell us about your vantage point when it comes to data centers and AI demand, because I do think there has been a bit of a reset this year. So what are you seeing? What can you tell us with clarity about that demand? And certainly it's you know, draw down on energy.
So we've spoken number of times in our earnings calls and interviews about the fact that we believe that this first phase of the AI data center build out is very similar to the pattern that occurred at the buildout of the original Internet. Go back to ninety the period nineteen ninety five to two thousand and two. What happened
we were building data centers. It was like every company has to be an Internet company and you have to have a dot com after your name, and a lot of capital flowed into these businesses, and what ended up happening was that the cost initially to build websites was
millions of dollars. You had to have unique machines by Sun Microsystems, You had to write all the code yourself, and then very quickly tools came out, Things came out that lowered the cost dramatically and lowered the amount of infrastructure you had to have to do this, and so all this infrastructure that got built was superfluous for a while. It took years for that supply to get consumed. And maybe I'm just an old guy in that I lived
through that industry. I was in the data communications industry at the time, and I lived through that boom bus cycle. And so as we looked at it, we said, we're only in the first generation of this technology, really, and people are going to be very focused on how do I do this cheaper? How do I do this so it uses less energy? Because in the bitcoin mining business, that's what you focus on all day long is how do I do this to consume less energy? How do
I do this more efficiently and low? And behold, you had deep Seak came out, you had all of a sudden, all these things came out that essentially showed you could do more with less, which while that does drive faster adoption,
it drives lower cost drives faster adoption. With what happened with cell phones, when you stop being charged for minutes, you just use it, right, So what happens now is instead people like I think Microsoft has been mentioned a few times by analysts and an analyst reports that they're slowing down on their execution of some of their private data center contracts because what they're seeing is, you know, okay, there's a certain amount of demand based on current technology.
People are starting to use this next generation reasoning technology, which is much more energy efficient, and as that transition happens, the design of what you need in the data center is slightly different reasoning models. I think the easy way to think of it is open AI. Everything is in memory at all times. When you look at deep Seek, only what you're working on is in memory, so it lets you have much smaller footprint, which means less compute, less memory, etc.
Yeah, this is something thing we've talked about too with our tech guys, but we're going to need to continue this conversation because it's certainly one that dominates our world and your world as well, and look forward to next time. Fred til He's chairman and chief executive officer of Mayor, joining us from Miami, Florida.