Loading summary
Transcript39 lines
- [00:03]
Samantha Murphy Kelly
This is the Next Innovation and I'm your host, Samantha Murphy Kelly, a senior tech correspondent. In the last few episodes of this series, we've unpacked how different innovations are influencing trends across the business and tech industry, from automation in warehouses to how AI can impact office culture for the better. But today I'm catching up with some fellow reporters who closely monitor the tech world. I'll ask them about the most notable emerging trends and what's on their radar for the year. Joining me today is my longtime friend Rachel Metz, an AI reporter for Bloomberg, along with Adam Rogers, a senior correspondent at Business Insider, and Matt McFarlane, a former CNN tech reporter and substack writer of Greenrocketship.com let's take a.
- [00:53]
Rachel Metz
Foreign.
- [01:00]
Samantha Murphy Kelly
Thank you all for joining today. All right, so based on your reporting, you guys are really deep into these worlds. What would you say were some of the most interesting stories that you covered or some of the most interesting stories that you've been covering and that you think will trickle on into 2025? Rachel, let's start with you. Since AI is certainly most top of mind these days this year, I mean.
- [01:22]
Adam Rogers
For me, I was focused a lot on how the technology was changing, how the companies are struggling to keep up really rapid pace of innovation. We saw some of the biggest players stumbling a bit. And that was something that I reported on deeply with some of my colleagues at Bloomberg and sort of like what are the power dynamics and the shifting power dynamics at play there?
- [01:43]
Samantha Murphy Kelly
Do you think we're going to see a big difference of what was big last year into 2025? Like, do you think by the end of this year thing? I mean, AI is just moving at such a rapid pace. Do you think that the landscape is going to be quite different?
- [01:57]
Adam Rogers
I'm not so sure. I mean, like development is, is going fast. There's tons and tons of money pouring into this space. However, it's not all going everywhere. I mean, a lot of what we're hearing about is money going to just a handful of companies. We've seen a lot of talk about agentic systems, AI being increasingly able to handle like more complicated tasks. That's something that people have been working on for a number of years now in various ways. But like it remains to be seen for the most part a how useful these things will be and also how much people will actually want to use them. Like, I'm not 100% convinced that people want to automate away all kinds of parts of their lives.
- [02:39]
Samantha Murphy Kelly
Matt, why don't you walk us through a little bit about what you've been tracking.
- [02:42]
Matt McFarlane
I'm really curious about, you know, if we can keep this pace up, if things are really going to keep going. Do things, know, have to slow down it. I, I'm kind of of two minds if I think there are a lot of uses of AI that people just haven't uncovered yet of how like, can large language models be used by say your average worker to improve their job? And like when I talk to people just doing all sorts of work in different sectors, there's that slice of people who are experimenting and finding new things and, and a lot of people who aren't. So I, I expect over time we'll start to, people are going to understand better how they can use new AI tools in great ways. At the same time, I definitely think we're in this hype cycle and I keep waiting for things there just to be a bit of a pullback of like, oh wow, Microsoft's spending like 80 billion on AI infrastructure this year and AI is impressive and growing, but there's also a lot of excitement.
- [03:39]
Rachel Metz
Yeah, I was just thinking about that. I do think that I'm really struck by how the difference between what it gets branded as AI broadly and what may be the useful parts of this technology versus the stuff that, like Rachel said, is TBD at best about whether people will want it and whether it will actually even deploy. Because I don't know Matt and Rachel, if you're seeing this, but like the press releases I see are all like, we've signed a deal to plan to possibly produce something in the near future. They're not like, here's the new product. You know, the new products were large language models. They were things like chatgpt. And those are obviously incredibly capable and wild to play with and sometimes really weird and mess up in kind of shocking ways also. And that is not, I think, what a big enterprise thinks about as like what might be a useful kind of AI. And I'm not even sure they're really talking about what those are. I'm not hearing it appropriately, I guess. I'll tell you like the, the technology that I, that really fascinated me on a bunch of different axes was robot cars, was robot taxis deploying for wide use in a bunch of cities, San Francisco especially. But I really think that the, the deployment of those is sort of experimental technology in a city with a couple of companies trying to get it done and not really succeeding. And what that kind of says about how cities are going to be structured.
- [05:00]
Samantha Murphy Kelly
Yeah, there's also the element of Safety. Using AI as a safety tool or for conscious living. You know, one of the companies we've been tracking, cameramatics, they have an. A powered camera that helps basically drivers detect blind spots or avoid collisions and, you know, even monitor when the driver is tired or distracted. As we kind of evolve over time and add more of these tools into cars or parts of our life, I mean, is this. Is this the solution? Like, we're moving right toward, you know, problem solving here? Are there. What are the concerns that are associated with these types of things as well?
- [05:34]
Rachel Metz
I thought that spending a lot of time riding around robot taxis in San Francisco, I thought, wow, this is an incredible boom for safety. Actually, you know, one of the huge problems. Problems in the United States right now, I thought, like, you know, there's something to be said for, like, after a couple of drinks, not driving, right? Like, not having a. Having a robot, you know, not make as many mistakes as a person will. Because I'm pretty convinced that. That if you had. If you. If you have equivalent numbers and you have a technology that's working, that it'll be safer than human drivers. And then kind of a week after I published that story, there were. There were stories coming out about people on the street, like walking, hassling and harassing people in the robot cars because all you have to do is stand in front of them to make them stop because they're hyper cautious. Obviously, the company's building them want to be hyper cautious about that, especially because the person inside the car has already essentially has signed a waiver when you sign up for the app. So if something bad happens to you in the car, you've waived your rights to sue. But if it hits somebody on the outside of the car, it doesn't. So they're very cautious about that stuff. And, you know, somebody. I said something about that I think maybe on Blue sky, and somebody came back like, yeah, you got to do the. The ex. Like, the ex girlfriend test is like, is this a technology that could be misused by someone's harassing and stalkery ex boyfriend? It's like, wow, yeah, actually, I should have seen that, because it's true. Safer for the people in the car, maybe around it, but also less safe for the people in the car because of how you can. Because how you. Because of how it's possible to affect that car in a way that you couldn't do if you were in an, you know, an Uber or a Lyft or a taxi cab. I do think, though, you. You asked a bigger question, which is about adjunct technologies of AI being an adjunct to help things be safer broadly. And I do think that's sort of, that's coming. I don't know what that looks like, but having kind of artificially intelligent assistance in almost any role saying like, hey, did you remember to turn off the dangerous thing? Seems like a pretty obvious use case for those.
- [07:20]
Samantha Murphy Kelly
Matt, when we were at cnn, you were covering self driving cars and technology related to this pretty closely. Have you been following this over the years too? Do you have any opinions on how this might, you know, where we are now and where it's going?
- [07:31]
Matt McFarlane
I think it's a huge growth opportunity, obviously and it's, you know, it's taken a lot longer than expected. There was so much hype and talk over the years about, you know, by now all of the cars should have been self driving. But it is just really tough technology. It's expensive. I think there are big questions too about can it be just profitable. The one thing always in the back of my head, you know, if I have any concern about autonomous vehicles, obviously everybody talks about things like privacy, what's the impact on traffic and congestion. But also, you know, robotic cars remind me a bit of like nuclear power. And it's a type of technology that's really awesome and incredible unless things go wrong. But I just always feel like this is a realm of technology that has great potential but, but also risks. So, you know, I hope it works out because of obviously the, the safety benefits. But obviously there are big questions still and challenges.
- [08:23]
Adam Rogers
What's interesting, I will say just as like a, an observation, it's so I work on the Embarcadero in San Francisco, which is a really interesting spot because it's like the nex. Every type of transportation you can imagine, which is really funny. Like you've got BART here, you have the ferry which is like behind our office. You have buses I think go by on the street here. You definitely have an old fashioned streetcar line which goes by, you know, like in front of the office throughout the day. And then you also get like a ton of Waymo taxis. And it's just so striking to me that I see like, you know, the old fashioned like streetcar like toodling by and then like you know, these robot cars plus you know, all these other things are all happening at the same time. My expectation is that we'll have some form of that going forward. Like I don't expect any of those things to 100% go away.
- [09:13]
Samantha Murphy Kelly
We recently did an episode that looked at a sort of an unprecedented four Directional automated pallet shuttle which was basically revolutionizing the way warehouses work. Blue collar industries, how it's changing and just the idea of automation. We're also seeing like sweetgreen, the restaurant chain here in the US has rolled out automated salad makers. I'm curious if you guys have been seeing, you know, automation in general, how it's been able to creep up in surprising ways across many different industries.
- [09:44]
Adam Rogers
I've seen it ebb and flow over the past, I'd say 10 years or so. It seems like there will be a lot of interest in automation. Then maybe people will be like, oh actually it's really hard to get like 100% of the way there with certain things. Like you mentioned food, food automation stuff I think is super attractive to especially like quick serve restaurants. But it's extremely hard to get that 100% right. And then you still have to have workers to, you know, make sure everything is going okay. Like if you like a burger flipping or even burger assembly robot for instance, something like that, you're still going to need people to oversee these things. But I think what you see with automation, like this pallet robot and a lot of these other kinds of systems that like I don't, I haven't really seen any exceptions to is that they're all going to need humans. Like they will always need humans in some way, shape or form. I mean we see that with even just software based systems. It doesn't have to be like any kind of physical system with like robotics. That's always for me, something that I try to like, keep in mind is that like even if people aren't needed for certain things, they'll always be needed for other things. These aren't just going to be systems that are 100% doing, you know what they will. Even the autonomous vehicles we have that are roaming the streets of San Francisco, they still use plenty of remote operators and could not as far as I can tell, operate at this point without them.
- [11:03]
Samantha Murphy Kelly
Another thing we've been monitoring as well is just the idea of drones. Obviously it's been top of mind in the tech industry.
- [11:11]
Rachel Metz
What's really interesting about that to me just in the context of this conversation is, is again like how people relate to these technologies that become ubiquitous. It's, it is possible to buy for not really relatively not that much money, an incredibly sophisticated almost autonomous flying vehicle. They're right. This technology isn't. If you told somebody about it 10 years ago, they would be like, that's not, you're not, that's going to be a $30,000, that's something the Department of Defense is going to own. You're never going to own something like that. And, and so what happens when you, when that's available is people don't really know how it's deployed. And I think with the, you know, with drones, that's another one that I know. A lot of companies, big ones are looking for other uses for them.
- [11:50]
Samantha Murphy Kelly
And it's not just the big companies too. There are smaller companies like Mana is a drone company and they use it for food and groceries and pharmaceuticals. And this is bec bigger. But I'm curious, do you think that like, will that ever become a mainstream thing? Like will the pizza ever dangle in front of your entryway to your home? Or like at the same time like will regulations and scalability challenges keep it relatively small?
- [12:14]
Adam Rogers
But my guess would be that it won't be something that you will see most of the time, that it would be used very sparingly for a whole bunch of reasons. Privacy issues, the fact that they're super, they can be super annoying. You don't want people damaging them. Generally, I feel like you're gonna have a lot of problem. Also there's like, you know, all kinds of local rules, I would imagine.
- [12:34]
Matt McFarlane
I think we're all like naturally a little uneasy with robots especially you know, in a way like your dishwasher is a robot, but it just doesn't move at all. So we're like okay with it. It's contained. But right when you go out on the street and you see a self driving car or a drone, it can be, you know, unsettling.
- [12:50]
Rachel Metz
I do think with any of these, any, any of these kind of automation technologies, they solve a problem for a company which is labor. And so the idea is like, well, this is going to be cheaper or faster, which is also cheaper or less unionized than a human being who wants, who would do the same job. And that means that it, but that means it has to do that job equally or better than the human who it's replacing for, for the less amount of money or for less management. So you know in a, you mentioned like a pick and place, robots are pallet moving robots. And those they, they can do more and they can do it faster. You still have to have a whole robot floor that humans aren't allowed on because they'll crush all humans on the front lines of a fast food place even if you can get that technology to work. And restaurant after restaurant, business startups and large ones alike have tried to have human Beings flip burgers or make salads and stuff. And it tends to fail every time because it's really expensive to maintain and it's not precise, not as precise as a human. Plus, just getting back to how like a city is going to relate to this. If you're a city that wants to fill your empty storefronts, the only ones that are going to be using automation are highly capitalized national chains. They're the only places that have the money and like the startup funding to put, put robots in there instead of just hiring, you know, instead of somebody hiring their children to come in and, you know, make burritos, which is like some of the best kind of food you can get in San Francisco. But also, all these drone companies now don't want to be regulated. And the only way to have them be like, do things like disaster, you know, rescue work or helping with fires in California or something is if they're tightly regulated and being used by places like fire department.
- [14:18]
Matt McFarlane
So one thing that's been really interesting to me, I followed this company, Skydio, for a while. They had just some great talent in terms of building drones. There were some people who worked at Google and gone to mit and they started out as this company making this nifty computerized drone that's going to follow you and film you. It was powered by great AI. And now when I look at what they're doing, they do a lot more of engaging government and supporting first responders and getting those government contracts. Things like that. I'm not expecting anytime soon where I'll just be having Amazon packages or our burritos dropping in our backyards. Or it's, can you safely get a drone delivery where you live? I mean, I thought about this at my house. I'm like, maybe there are trees, there are power lines. Sometimes my dog's in the backyard, sometimes one of my kids may run out there. So there are just a lot of challenges. I think it's what's so tough about robotics and growing it. It works a lot better if you're in a controlled environment like an Amazon fulfillment center versus out in the real world.
- [15:23]
Samantha Murphy Kelly
Let's shift a little bit to obviously technology has applications across many different areas, but climate, you know, mitigation efforts is another really big one. We've been following many different companies for years and as you know, it's sort of evolved. Different solutions have kind of popped up, you know, in terms of sustainability as well. There's a company called Wetushure that is basically offering like insurance solutions for Carbo Credit markets. So they protect investors against things like fraud and loss and threat. And you know, these are sort of examples of, you know, potentially other use cases for, you know, in the climate change crisis. So I'm curious if there are other interesting trends or innovations that you've noticed in this space as well.
- [16:04]
Rachel Metz
One of the issues is that in California, because of the threat of fires, a lot of insurers have been refusing to insure certain areas or certain neighborhoods for, for fire. Part of the reason that they've said that's so expensive and part of the negotiations between those insurance companies and the state of California, because it states that do that kind of licensing have been whether they're allowed to use AI models to figure out where the higher risks are and how much to charge for that. The state's been reluctant to do that because while there are a lot of these AI models that can predict climate risk, you know, sometimes down to the smallest, to an area as small as a single house or a place where you might cite your next warehouse or company, they're black box models. Nobody really knows how they work and they all come up with different results. There have been some tests that different journalistic places and others have done testing these models and they, they'll show different amount of risk for the same place and nobody really knows how they work. Part of the problem with large language models and other large AI models now is that it's very difficult to tell what the math actually is inside those because it comes from training and they sort of come up with their own approaches to solving these problems. A sustainability approach here would be can you make it cheaper and smarter to figure out the models for where climate threats are and how to insure these places? Because insurance business is a very old one, but it runs on the very expensive risks are also rare. That becomes less true in a more climate impacted world. Can they be using AI type systems to figure out better ways to amortize and apportion that risk? It's possible. Nobody knows if it's going to work, but that was one of the issues even before the buyers with insurance in California. It's going to become even more of an issue now that all those insurers really see that they were kind of right to not be in some of those neighborhoods in terms of their business model and how they're going to pay these folks, these thousands of people who've lost their homes.
- [17:45]
Samantha Murphy Kelly
That makes a lot of sense. A lot of climate innovations aren't necessarily climate friendly. What do you think about, I guess the Intersection of technology and climate efforts and the idea of the ethics as well, it's tough.
- [17:59]
Matt McFarlane
I think so many, the tools, these new products and services we love, often they do have negative climate impacts. And if you think about how much energy goes into say a ChatGPT query versus a Google search, and even I think about the growth of Bitcoin and how popular that's been, but also just uses a ton of energy. And often we don't always know how much energy is going into these products and tools that we're using. Right When a lot of things were just, you know, interfacing with the large language model, like through my iPhone. And I really don't know, you know, how many Nvidia H1 hundreds are just like sucking up lots of energy to answer some question I have. So I do think it's, it's really challenging in that, you know, so much of our great way of life has depended on negative impacts on the environment. How do we find this balance of trying to like, live in harmony with nature, of enjoying new tools, great innovation, while also, you know, not being, you know, too destructive to the world we live in? I think it's, it's hard for us to, to find that harmony.
- [19:11]
Rachel Metz
What are people going to use ChatGPT for? What are large language models going to be? You know, that's not. If that is tbd, they're already spending a lot of money to build new power infrastructure or not take old power infrastructure that was heavily carbon emitting offline, that was going to come offline to supply the power to the data centers where all the chips are that get really hot to do the AI computation. So, you know, you can ask, well, okay, is that going to be worth it to us? Are we going to get so much back from what AI can do that it'll be worth it to keep burning coal or natural gas? And companies are saying like, no, but we're going to use it. They're going to get. The AI is going to get so good that it'll figure out how to solve climate change. It's like, well, that, that would be a trade. That would be, that would be great.
- [19:53]
Samantha Murphy Kelly
Let's talk a little bit about President Trump's current and growing allegiance to tech companies. You know, how big of a role do you think the administration is going to play in the development and implementation of these new technologies? And perhaps like, what are some of the biggest challenges, you know, in a second Trump administration could pose for the tech sector?
- [20:13]
Adam Rogers
I at the moment would expect that a lot of the Regulatory stuff. I mean, we're already seeing a lot of the, like, potential regulatory stuff is going to kind of go away, at least for the time being. And we saw with this Stargate announcement that it's really important for the Trump administration to look like it is shepherding in like a new age of AI. Some of it is still TBD though, because with Trump you don't quite know what's going to happen. You know, like you can expect one general thing due to party affiliation, but that doesn't mean that that's what's going to happen. I think it's going to be some, you know, ups and downs and a bit of a roller coaster, but it should be interesting.
- [20:52]
Rachel Metz
One thing that I think you can expect is the abandonment of the aggressive antitrust and anti monopoly efforts against technology companies that the FTC under Lina Khan and the Department of Justice were undertaking. They were becoming quite aggressive and that's part of the reason that a lot of the sort of richest tech folks stated that they thought that that kind of regulation and that kind of aggressive antitrust intervention that a lot of economists were asking for, there was some pretty good work showing that, you know, Silicon Valley tech related companies had these problems that they were going after. And you know, that's part of the reason that they supported a Republican presidential campaign. That and the, the further threat apparently of other regulation, regulation of artificial intelligence and cryptocurrencies. And you know, those are things that, like if you're Marc Andreessen and you're heavily invested in those kind of companies, you were looking for less regulation of those.
- [21:46]
Samantha Murphy Kelly
Absolutely. I mean, that's a good question. So, you know, obviously maintaining a close circle of tech leaders, but what does this mean for smaller startups? Do you think that this is going to incentivize innovation or do you think it's going to be counterproductive?
- [22:01]
Adam Rogers
I think it is hard to say because this could be good in some ways for smaller companies, but obviously it will be very good for the current leaders in the space. I mean, if nobody is cracking down on them with legislation and then they can keep, keep staying and they will at least stay where they are, I would expect. Unless there's some sort of breakout hit, some sort of network effect within the startup community that pushes a smaller company to the top. I mean, yes, it's possible that someone could come up with like some incredible AI innovation. I'm always thinking about what is the next architecture that people are going to be looking at or how are people going to be combining these things. Differently, there could be an agent company or there could be a company working on reasoning type stuff that is able to pull ahead. But, and it's a big but. If they don't have huge financial backing, they just can't. They won't be able to scale, you know, they won't be able to get that product out to lots of people. So I think this is a great thing for the biggest companies and I'm not sure how much it will really benefit the smaller ones.
- [23:03]
Rachel Metz
You know, it's an open secret. It's not even the secret, I suppose is well understood now in the kind of, in the Silicon Valley economic infrastructure that like companies that get venture capitalized do not go to ipo. They do not end up like some vanishingly small percentage of them, 5% or less end up, you know, listed on an exchange someplace. They get acquired or the people get acquired. And the reason that incumbents do that is first of all, the way that the system exists now, they just have a tremendous amount of cash and can do that and they have their own venture fund. So it's like, yeah, you're going to do it. Plus just the pressure on you if you're an incumbent is you have to crush every potential competitor in its crib because you never know which thing is going to be the thing that, that overturns you. And now they have the resources to do that. So one of the reasons that the ofTC and the DOJ were being so aggressive about antitrust was just this exact argument to make to open space for, you know, smaller startups. But those threatened the incumbents, those, you know, candidly threaten those people who were sitting on the inauguration stage. You get a company that can do that stuff better and needs those kind of resources so they can just invest in them and they become part of it. Which is what Microsoft did with OpenAI really. But like they do it because they can and because the pressure is on them to do it. So it does, you know, it puts a, it puts people wanting to do a startup in a, in a position where at minimum they have to only do things that are not being done by those incumbent companies. You know, it narrows the, the range of what they can try to do.
- [24:28]
Samantha Murphy Kelly
So with all that said, are there any other areas that you think that we should be monitoring in 2025, whether it's crypto or, you know, obviously we've talked a good bit about AI as well, but just curious where you, where your will be in the next 12 months or so.
- [24:44]
Matt McFarlane
We were just talking about big Companies and the advantages they have. Like, I wouldn't be surprised to see some smaller AI companies or other companies get bought and brought into these big companies. Wonder maybe, you know, Apple could probably use a stronger AI company. Maybe that's an option. I've been waiting for a while for someone to buy Lyft. It's no longer founder led. It feels like it could plug in well somewhere. The real obvious one is like TikTok. And is that going to end up in Elon Musk's hands? Even a company like intel, what's going on there? Could they be bought by somebody? Is this the best moment they'll have? Any time? Yeah, those are things I'm curious about.
- [25:24]
Adam Rogers
This year I'm going to be paying a lot of attention to what some of the AI industry leaders, in particular companies like OpenAI and Anthropic are doing as far as the kinds of technology that they're focusing on. That will show us where they see the industry going at least in like the coming months, in the coming year. What I've noticed over the past number of years is, and a lot of technology companies do this, right? They, they lay the groundwork for larger things in little bits and pieces. And so if you kind of just watch and see sort of what features they're rolling out, what technologies, what types of AI models they're rolling out, you can start to get a sense for like where they see things going. I mean they make all these grand predictions. But I think I try to keep in a closer eye on what are they actually doing. Like what can I actually do with this technology? And not, I'm less interested in what, you know, I may eventually be able to do with it. I mean, who knows if I'll even want to use it. So just keeping an eye on what reality is, I try to pay close attention to that and increasingly the environmental impact. I mean, I think that as these systems become more and more powerful, it's ever more important to think about, well, what is the trade off here and is it worth it? Because in some cases it may not be. And there are some people that are doing some really interesting work sort of on the opposite side. They're making AI models smaller, for instance, smaller, more capable networks of models. There's a nice amount of work in that area to try to figure out how can we make these things really capable but not actually consume a ton of energy. There's a, there's a lot of interesting stuff going on.
- [26:57]
Rachel Metz
I'm actually interested what happens with other sustainability related technologies. What happens with grid scale, solar, wind as the, you know, the federal government seems to be pulling back on those kind of contracts. But things like solar and grid scale batteries were huge lines of business for companies, some of them American grid scale batteries are a huge line of business for Elon Musk. So even though the current administration has, you know, very clearly from out of the gate been pretty down on green technologies, on any kind of energy technology that's not, doesn't involve, you know, extracting oil from the ground and burning it. So I'll be interested to see what happens, you know, over the course of the year to see if those trends maintain even in the face of what become headwinds from federal government instead of a tailwind.
- [27:42]
Samantha Murphy Kelly
That's true. We'll have to keep an eye on it. Thank you so much to everyone for sharing what's capturing your attention most these days. It'll be fascinating to see how a lot of this plays out. Thank you for listening to the Next Innovation. This series was produced by Situation Room Studios and powered by Enterprise Ireland. Investing in the next wave of innovation. Our executive producer is Christine Barata and our senior producer is Sharon Barrero. Emily Beeman is our associate producer. Additional production assistance by Global Situation Room. A special thanks to Rachel Mets, Matt McFarlane and Adam Rogers. I'm Samantha Murphy. Kelly until next.