Jobs of the Future, Harnessing Earth Observation, & Gaming Tech Advances AI transcript and summary - episode of podcast a16z Podcast

Go to PodExtra AI's episode page (Jobs of the Future, Harnessing Earth Observation, & Gaming Tech Advances) to play and view complete AI-processed content: summary, mindmap, topics, takeaways, transcript, keywords and highlights.

Go to PodExtra AI's podcast page (a16z Podcast) to view the AI-processed content of all episodes of this podcast.

a16z Podcast episodes list: view full AI transcripts and summaries of this podcast on the blog

Episode: Jobs of the Future, Harnessing Earth Observation, & Gaming Tech Advances

Author: Andreessen Horowitz
Duration: 00:41:59

Episode Shownotes

As 2025 begins, industries are evolving at unprecedented speed: robots are revolutionizing manufacturing, terabytes of earth observation data are driving new possibilities, and gaming technology is transforming how we design, train, and innovate across sectors.In this episode, a16z General Partner Erin Price-Wright, Engineering Fellow Millen Anand, and Partner Troy Kirwin

discuss the trends reshaping the future of hardware, software, and beyond.We explore:How robots and full-stack engineers are driving the next industrial renaissance.The explosion of Earth observation data and its potential to revolutionize industries.How gaming technology is moving beyond entertainment to reshape training, design, and more.This is just the beginning of our four-part series on 50 Big Ideas for 2025—don’t miss the full list at a16z.com/bigideas.Resources: Find Erin on X: https://x.com/espricewrightFind Millen on LinkedIn: https://www.linkedin.com/in/millen-anand/Find Troy on X: https://x.com/tkexpress11Stay Updated: Let us know what you think: https://ratethispodcast.com/a16zFind a16z on Twitter: https://twitter.com/a16zFind a16z on LinkedIn: https://www.linkedin.com/company/a16zSubscribe on your favorite podcast app: https://a16z.simplecast.com/Follow our host: https://twitter.com/stephsmithioPlease note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.

Full Transcript

00:00:00 Speaker_03
How do companies take the magic that happened when we all got up at 5 a.m. and watched SpaceX catch a massive multi-story rocket with chopsticks? Yeah.

00:00:14 Speaker_00
Terabytes and terabytes come down from orbit every single day, soon to be probably petabytes.

00:00:19 Speaker_03
We're talking about job creation for a whole class of folks that don't necessarily need a degree to participate.

00:00:26 Speaker_01
All the technology exists today to have that experience in an amazing, intuitive way, and yet it doesn't exist.

00:00:33 Speaker_03
What is the next generation water treatment engineer look like? It's someone that knows their way around a robot.

00:00:40 Speaker_04
In a matter of days, we'll say goodbye to 2024 and start 2025. It's hard to believe. It's been 25 years since Y2K, 21 years since Facebook was founded, and even nearly 10 years since OpenAI was brought to life.

00:00:55 Speaker_04
So if it feels like things are moving quickly, you're not alone. That's why every year, we ask our partners, who are meeting every day with the people building our future, what they think is in store for the following year. Last year, we predicted

00:01:08 Speaker_01
A new age of maritime exploration. Programming medicine's final frontier. AI-proof schemes that never end.

00:01:15 Speaker_04
democratizing miracle drugs. And on deck this year, we'll be exploring... Infrastructure independence and hypercenters. Super staffing for healthcare. Regulation will become code.

00:01:27 Speaker_04
And throughout this four-part series, you'll hear from all over A16Z, including American Dynamism, healthcare, fintech, games, and more. However, if you'd like to see the full list of 50 big ideas, head on over to a16z.com slash big ideas.

00:01:44 Speaker_04
As a reminder, the content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund.

00:01:57 Speaker_04
Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see a16z.com slash disclosures.

00:02:15 Speaker_04
Today, we focus on the increasingly interesting intersection of hardware and software.

00:02:20 Speaker_03
Up first? We're seeing a renaissance in technical disciplines that cross the hardware-software chasm. The robots are coming. Someone will have to build, train, and service them.

00:02:31 Speaker_03
That was Aaron Price-Wright, a general partner on the American Dynamism team. And here's Aaron's big idea. In the 2000s and 2010s, if you weren't coding, it seemed like you'd get left behind.

00:02:43 Speaker_03
The number of computer science majors exploded while degree programs like mechanical engineering and electrical engineering shrunk on a relative basis.

00:02:52 Speaker_03
Now we're beginning to see a crucial shift amid the push to reshore manufacturing, the mass retirement of skilled workers across unsexy industries like water treatment, commercial HVAC, and oil and gas, and the rise of autonomy across defense, enterprise, and consumer applications.

00:03:11 Speaker_03
People are building physical things again. It's really exciting to see. And that's going to require this full-stack skill set of people who can cross the hardware-software chasm.

00:03:21 Speaker_03
So I think we've really seen over the last two decades a mass migration of engineers to the software and computer industry.

00:03:30 Speaker_03
And what I think we're going to see is huge demand for engineers in areas like electrical engineering, mechanical engineering, controls engineering.

00:03:40 Speaker_03
are going to be in really high demand over the coming year as industries from defense to industrials, manufacturing, even things like HVAC and water treatment are looking for people that are really ready to wrestle with and bring AI software into these really complex hardware contexts.

00:03:59 Speaker_04
Over the last few decades, we have seen a lot of people migrate to software, and now maybe we're seeing a shift. Can you speak to the macro trends that are really underpinning that?

00:04:08 Speaker_03
I think the last 20 to 25 years has really been the software is eating the world, not to steal our own tagline, but it's really true. Most of the growth in the economy came from the software industry.

00:04:21 Speaker_03
So with first the rise of software as a service and then more recently the kind of explosion of AI, the skill sets that have really been in demand have been software focused.

00:04:31 Speaker_03
But with the kind of very recent crossing of the chasm from software to hardware, there's huge demand for engineers that can both speak software and speak hardware.

00:04:42 Speaker_04
Yeah, and as we do feel this pivot back, are you seeing that in the data when it comes to the degrees that people are getting?

00:04:48 Speaker_04
Are people no longer getting software engineering degrees at the same clip and maybe choosing to take a mechanical engineering degree, for example?

00:04:56 Speaker_03
We're not seeing that necessarily yet in degree programs, but we're seeing that in the demand in companies. And so I think that's going to filter out into what people end up studying in school.

00:05:06 Speaker_03
So the highest demand jobs in our portfolio today, other than perhaps AI engineers, are people who have full stack hardware and software experience, whether it's mechanical, electrical, etc. And what we're finding is that actually those

00:05:23 Speaker_03
companies are having to go outside of traditional tech feeder schools in order to find that talent. So we're actually seeing an interesting rise of schools like Georgia Tech, Colorado School of Mines, University of Michigan.

00:05:36 Speaker_03
Some of these really hardcore engineering programs are really coming into talent pipelines for some of our top tier tech companies in the portfolio, which is really interesting.

00:05:44 Speaker_04
So interesting. And let's actually touch on that specifically, this idea of us needing all of these new kinds of engineers. Some of them previously existed, but there seems to be a gap. And there's this question of who's going to train these people?

00:05:56 Speaker_04
Is it the four-year degree program?

00:05:57 Speaker_03
Is it something different? There's definitely a spectrum and I would frankly really credit Elon Musk both from the SpaceX and Tesla perspective with a lot for really training up engineers that have fluency in hardware and software.

00:06:09 Speaker_03
As more companies that are building physical things grow, companies like

00:06:14 Speaker_03
Skydio and Andrew and others are really taking on that mantle and training a new generation of full stack engineers that are really comfortable and fluent across hardware and software.

00:06:24 Speaker_03
I do think the university systems or some sort of education system has to catch up. I expect that schools like Stanford and MIT have great engineering departments outside of computer science programs.

00:06:36 Speaker_03
I'm hopeful that those programs will continue to grow as the demand in the market continues to grow. And then I also think it's not just engineers with four-year degrees that are going to be valuable.

00:06:47 Speaker_03
Like, we're talking about job creation for a whole class of folks that don't necessarily need a degree to participate in this sort of reshoring AI or autonomy-driven hardware economy.

00:07:00 Speaker_03
technicians who will help service and test robots on the manufacturing floor. That is a well-paid, secure job of the future, I think. Another example that we're really excited about is robotic teleoperation.

00:07:13 Speaker_03
So you have someone remotely operating a robot in an environment where maybe it's dangerous or hard to get to or just really difficult to staff humans. It's really difficult doing robot teleoperation. It's a valuable skill set.

00:07:26 Speaker_03
I expect that there'll be an entire kind of class of jobs related to that arising over the next few years.

00:07:33 Speaker_03
And again, that's an example of something you don't need a four-year degree to get really good at robot teleop, but it's a very valuable skill if you do it.

00:07:39 Speaker_04
Yeah, and there's no current university degree that would offer that, right?

00:07:43 Speaker_03
Precisely.

00:07:43 Speaker_04
Are there any other jobs, new jobs, that stand out that maybe we don't even see quite yet?

00:07:48 Speaker_03
I think that when you look at, for example, the recent results that came out about the TSMC factory in Arizona, Governor Katie Hobbs has introduced this big apprenticeship program where they're helping to train new

00:08:02 Speaker_03
semiconductor manufacturing employees alongside folks who have come over from TMC. So with kind of the rollout of the CHIPS Act, I expect as chip manufacturing comes back to the U.S.

00:08:11 Speaker_03
that it's an entire class of jobs that we just have literally never had here. Maybe we did for five years in the 60s right at the beginning of the industry. And I expect reshoring, we'll see a lot of those coming back.

00:08:22 Speaker_03
I think another example, more in the industrial sector, is you have lots of industries that hired people when they all modernized roughly in the 80s.

00:08:34 Speaker_03
So there's a large class of, it's oil and gas, it's water treatment, it's chemical engineering and HVAC.

00:08:42 Speaker_03
There's a class of these heavy industries that introduced a lot of new equipment and new ways of operating in the late 80s, hired a whole bunch of people to work on it, and then essentially never really had to hire. anyone else. They're great jobs.

00:08:57 Speaker_03
There's low turnover. They're well paid.

00:08:59 Speaker_03
And now we're seeing kind of mass retirement across some of these heavy industries and companies really looking to incorporate more autonomous kind of control systems and the way they manage some of these processes. But there's no one to operate them.

00:09:14 Speaker_03
So what is the next generation water treatment engineer look like? It's someone that knows their way around a robot.

00:09:20 Speaker_03
a control system, a PCB board, they're going to have to know their way around sensor data, and they're going to have to be able to understand how all of that integrates together to be able to troubleshoot when things go wrong.

00:09:31 Speaker_03
So huge opportunities in some of these industries where we're just seeing massive labor gaps over the coming, I don't know, three to five years.

00:09:38 Speaker_04
Wow, and that's pretty soon, actually. Yes, it's happening now. Yeah, and so that sounds like an incoming shortage. Are there any other shortages as we think about the supply and demand? Maybe not so much of completely new jobs, but existing jobs.

00:09:51 Speaker_03
The thing that really stands out in our portfolio, especially when you look across industries like aerospace, defense, robotics, it's this notion of a full-stack engineer. And I don't mean a full-stack software engineer. I'm talking about a full-stack

00:10:09 Speaker_03
hardware engineer, so someone that can write some code, they can write some firmware, they can fiddle around with electronics, maybe design a really simple sort of PCB board with components, figure out how it fits in a mechanical system, sort of troubleshoot when things go wrong.

00:10:29 Speaker_03
There are just very few people who have that kind of end-to-end skill set. And many of these industries are still relatively early on their robotics or autonomy journey. So there's still a lot of iteration to go around. Things aren't fixed in time.

00:10:44 Speaker_03
So the biggest demand that we're seeing in our portfolio is that fluency across multiple different domains that includes hardware or software.

00:10:53 Speaker_03
So it's not necessarily that you need to be the literal world's best electrical engineer and have a PhD in electrical engineering. It's more that you're flexible and comfortable and familiar moving across different parts of the stack.

00:11:09 Speaker_03
You're able to unblock yourself and you're really able to try things and implement things yourself as all of these companies are figuring out what the next generation of hardware systems looks like.

00:11:21 Speaker_04
Fascinating. And as we think about closing that gap, right, producing more of these people who are full stack in the hardware sense, what do you think is needed?

00:11:29 Speaker_04
And let me preface by giving the example of we saw this wave of software engineers over the last few decades, and that was pulled by the market in many senses.

00:11:37 Speaker_04
You had these large companies like the FANG companies providing a lot of really impressive benefits for these people to come in. And it became the kind of job that everyone wanted for a period. Do we need the same kind of, you could say, marketing?

00:11:50 Speaker_04
Are there other effects that you think need to influence the larger system?

00:11:55 Speaker_03
I think a lot of it will be market-driven. It's also about capitalizing on this sort of romantic desire to build stuff that I think is in the air right now. So how do companies take the magic that happened when we all got up at 5 a.m. and watched

00:12:18 Speaker_03
SpaceX catch a massive multi-story rocket with chopsticks in the air. That's so cool. And I think the companies that can figure out how to translate that feeling into

00:12:35 Speaker_03
Pulling people who might otherwise go down a safer route of, oh, I'm an ML researcher, so I'm going to go work on LLMs because I know I can go get a really big paycheck at one of these research labs.

00:12:49 Speaker_03
I'm a smart person, and I'm going to go code a chatbot. I think it's up to these companies that are building in these hard and gritty spaces where it is harder because having to worry about hardware is harder than dealing with software alone.

00:13:04 Speaker_03
But if they can pull that talent and inspire people to really get into the guts of what does it mean to deploy a model on a physical system and what is the complexity and challenge in that.

00:13:14 Speaker_03
I think it's like capturing the lightning in a bottle that exists in the moment right now and turning that into early movement from the pure software or pure AI industries.

00:13:25 Speaker_03
We're already seeing this with larger robotics labs that have gotten well funded over the last 12 months or so. They're managing to pull really incredible talent.

00:13:33 Speaker_03
I think that companies like Anduril and Waymo and Skydio and others are pulling incredible talent still.

00:13:42 Speaker_03
And that's going to hopefully translate into more and more 18-year-olds being inspired to take a mechanical engineering class or build a robot rather than just study pure software like they maybe have been for the last decade.

00:13:56 Speaker_04
Only since you just mentioned the kind of 18-year-olds who would be starting their skill journey from scratch, do you also think that the folks who are older, who have maybe gone through their career already partially, is there a role for them to play in the reskilling and entering the workforce maybe for a second time?

00:14:13 Speaker_03
Yeah, for sure. I think I see the rise of autonomy.

00:14:19 Speaker_03
the benefit that's going to bring to the labor market in terms of manufacturing jobs in the United States, in terms of high-skilled technical work replacing low-skilled labor, huge job market opportunities.

00:14:34 Speaker_03
Like, how to manage that training process and training pipeline, I think it's going to take a lot of different forms, whether it's apprenticeship programs, teleoperation, outposts, I think we're going to see a whole bunch of different things emerge.

00:14:46 Speaker_03
But for people who are curious and like to build and are highly technical, even if that doesn't mean that they've gotten a college degree, electricians, for example, like there's a huge shortage of electricians in places like Texas and Georgia, etc.

00:15:02 Speaker_03
I think I was talking to someone from Microsoft recently who said that as they were standing up some of their new data center work, In Georgia, they employed, at one time, one-third of the registered electrician's state, which is crazy. A third?

00:15:17 Speaker_03
A third. So there's just a huge dearth in some of these skilled trades that are going to become really important as the AI and autonomy expands across the United States.

00:15:28 Speaker_04
Absolutely. And even in your big idea, you cover so much ground, all kinds of jobs, all kinds of industries, whether it's mining or energy or things like autonomous vehicles.

00:15:38 Speaker_03
Everything physical that we interact with or even everything physical that affects our life that we don't even know about is going to be impacted by autonomy over the next decade. And that's a lot of jobs to fill.

00:15:49 Speaker_04
Is there any particular area within that that you think is especially important to get right?

00:15:55 Speaker_03
I think that it's really important that we figure out how to build things again in the United States at scale in production.

00:16:08 Speaker_03
And I don't think that's going to happen with the way sort of labor economics work globally if a huge part of the way we manufacture in the future is not driven by autonomy. And it's kind of a chicken and egg problem, like we need the robots

00:16:26 Speaker_03
to build the factories, and we need the factories to build the robots, because right now all the components for robotic arms, everything that might go into an autonomous factory is all coming from Shenzhen.

00:16:39 Speaker_03
So how do we start bootstrapping this supply chain native to the United States or the U.S. and our allies so that we're less fundamentally existentially reliant on another power? But we have to start somewhere, and I think it's going to take people who

00:16:55 Speaker_03
really want to build and get stuck in and don't mind the hairy complexity of having to deal with AI that doesn't quite work and hardware that doesn't quite work and supply chains that aren't quite ready but are willing to just tackle that problem vertically head on to do it.

00:17:12 Speaker_04
Absolutely. And this was a big idea for 2025. So as we prepare to start that journey, what are you looking out for? What are you thinking about? What do you hope to see?

00:17:22 Speaker_03
I hope to see more founders and teams that are willing and ready to tackle this problem head-on vertically.

00:17:33 Speaker_03
So they're not outsourcing their hardware, they're not outsourcing their supply chain, but they're really recognizing that in order to build something new here in the U.S., they have to figure out all of those things and have to develop skill sets across all of those categories and become kind of generalists.

00:17:52 Speaker_03
I think something Elon Musk has done really well, not to just be, you know, an Elon shill on here, but it's something he's done really well. to really own the kind of end-to-end supply chain for birthing something into the world.

00:18:04 Speaker_03
And I think we need more of that in founding teams.

00:18:07 Speaker_04
We hope to see many more founding teams in 2025 crossing that hardware-software chasm.

00:18:13 Speaker_00
One frontier with a flurry of activity is space, with... The number of Earth observation satellites doubled in the last five years from 500 to over 1,000.

00:18:21 Speaker_00
There's more data than ever downlinked to the Earth, and it is easier than ever to access imagery, although still far too difficult. that was. Mellon, I'm an engineering fellow on the American dynamism team.

00:18:32 Speaker_04
Mellon thinks there's an opportunity in 2025 to harness all this Earth observation data.

00:18:38 Speaker_00
My big idea is really around building verticalized Earth observation tools.

00:18:42 Speaker_00
There's been a lot of work over the last decade or so to actually get a lot of Earth observation sensors and set up the infrastructure to send pixels down to Earth from space, which is really a remarkable task.

00:18:54 Speaker_00
For 2025, I'm really excited to see entrepreneurs build verticalized solutions that go into different industries and actually solve customer problems.

00:19:01 Speaker_04
There has been a boom in Earth observation satellites going up. What's really driving that growth?

00:19:06 Speaker_00
I think first off is launch costs and access to space. In decades previously, it was very, very difficult and expensive to send satellites up to orbit.

00:19:15 Speaker_00
These days, it's honestly getting as easy as sort of like a bus service with things like SpaceX's transporter launch service. So the cost per kilogram of satellites getting to orbit has really, really come down.

00:19:25 Speaker_00
Also, architecturally, satellites have changed a lot in the last few decades. They're no longer sort of school bus-sized, $100 million per unit satellites. They've come down to the sizes of a loaf of bread.

00:19:37 Speaker_00
And so with that, you get a lot of proliferated sensors, you get a lot of coverage over the earth, and costs have really remarkably come down. And then I think the last thing I would point to is communications infrastructure.

00:19:47 Speaker_00
There's been a lot of work to set up ground stations and to actually have satellites communicate with each other in space to have a more efficient way to send pixels down to the Earth.

00:19:57 Speaker_04
And so as we think about those economics and if the economics have come down so much on getting the satellite up there, how does that ultimately ladder to the applications that can be built and the economics around that?

00:20:07 Speaker_00
To me, it's a really exciting positive flywheel here.

00:20:10 Speaker_00
The more satellites you have up there, the more pixels that are collected of the Earth's surface every single day, there become more products that you can actually build and more use cases that you can unlock.

00:20:19 Speaker_00
Prices sort of come down, and that allows and unlocks entrepreneurs to really cheaply solve problems for the entire Earth at once.

00:20:26 Speaker_04
Since we're talking about economics, maybe you can give listeners a sense of what those are. Like, how much does it cost for us to get this data, process it, or for real end applications to be made?

00:20:37 Speaker_00
So the unfortunate answer is that it depends a little bit. Broadly, there's actually a huge amount of freely open data. So NASA's Landsat program provides free open data for the entire earth.

00:20:47 Speaker_00
The European Space Agency has Sentinel, which also provides free data. And then there's a lot of commercial companies as well who provide medium resolution. The prices range in the sort of dollar to $5 per kilometer squared.

00:21:00 Speaker_00
And so that's really not too expensive. And then it tends to be more expensive as you go up into the high, resolution 30 and 10 centimeters per pixel, but those are typically for more specialized use cases.

00:21:11 Speaker_00
And I think it's not very well known that it even could be this cheap. And there's a whole suite of archived data that goes back years and years from these companies who have collected data from orbit every single day for years and years.

00:21:22 Speaker_00
That's a goldmine waiting to be really leveraged.

00:21:25 Speaker_04
And we've seen different private entities, public entities, governments, universities really pour in resources to get a lot of this data. But speaking of the applications that in some cases do already exist, what are some of those?

00:21:38 Speaker_00
There have been some really, really exciting applications so far. I think agriculture is one of the ones that is sort of most talked about.

00:21:44 Speaker_00
Farmers across the world are really using earth observation data to monitor their crops, predict crop yields, and understand ways that they need to irrigate or fertilize their crops to really produce better outputs here.

00:21:56 Speaker_00
So I think that's really, really exciting, especially as we face food shortages across the world. Other applications include defense. Governments across the world use it to monitor things like troop movements, ships and ports,

00:22:07 Speaker_00
and fleets of their equipment across the world. Energy use as well.

00:22:11 Speaker_00
There's a lot of interesting work that can be done in forecasting clouds and large scale grids and looking at how much solar production will happen on any given day, planning of utility solar farms, like looking at land and where you can actually place utility solar farms.

00:22:25 Speaker_04
Obviously, this is your big idea for 2025. Why do you think this maybe has been underexplored to date, earth observation in particular?

00:22:33 Speaker_00
I think it's really, really difficult. And I think that it's not so easy to work with the data. Companies are making it a lot easier these days. It's sort of a new thing that you can even go on to a website and buy an image through a portal.

00:22:44 Speaker_00
You used to only be able to talk to a salesperson to be able to do that. There's a lot of open source work. There's a lot of companies chewing off different parts of the puzzle.

00:22:53 Speaker_00
The main thing in my eyes that I'm looking forward in the next year to two years is entrepreneurs actually going into specific industries and really taking solutions vertically.

00:23:02 Speaker_04
So in one sense, if you take something like SpaceX, and they have Starlink, and they're the ones who create the satellite, put it up there, and then also sell the data that people can use.

00:23:12 Speaker_04
Are you basically saying that you expect to see more of that specific solutions for, let's say, like you said, agriculture or energy, instead of one company putting the satellite up there and then other companies in the middle distributing that data, processing it, et cetera?

00:23:26 Speaker_00
It has traditionally been hard for incumbents and natural satellite manufacturers to go into industries like agriculture and really get granular in terms of solving problems.

00:23:36 Speaker_00
Just an example of that would be like automating any sort of farming equipment. It's hard for an existing Earth observation player to go really, really vertical all the way straight from orbit to the farm.

00:23:47 Speaker_00
And right now that's traditionally stopped somewhere around providing analytics or insights, but I'm looking for in the future to actually provide automation.

00:23:55 Speaker_00
And maybe we can drive a lot of that farm equipment, for instance, or actually irrigate fields directly through a closed loop Earth observation.

00:24:04 Speaker_04
Maybe a follow-up there is we've obviously seen a bunch of machine learning and AI tools come up in this last few years.

00:24:11 Speaker_04
Does that change the game in terms of being able to parse data, process it, make sense of it for these verticalized players, for example? Or how does that really reshape the ecosystem?

00:24:20 Speaker_00
Certainly, yeah. I mean, terabytes and terabytes come down from orbit every single day, soon to be probably petabytes.

00:24:26 Speaker_02
Wow.

00:24:27 Speaker_00
And we don't have enough humans on the earth to actually look at all those images. So we really need advanced techniques here. And I think there's been some exciting progress here.

00:24:34 Speaker_00
A really famous example, a couple of years ago, where a company actually found the Chinese spy balloon through training their model with AI prediction of exactly what the balloon might look like.

00:24:45 Speaker_00
And they were able to sort through pictures of the entire US, taken like over days and days, and actually track down the source of the spy balloon for the first time. So we really can't do that with humans.

00:24:56 Speaker_00
And I think new use cases like this will be unlocked every year or so.

00:25:00 Speaker_04
And as we think about the roadblocks or the challenges that may be on the road to us, really this Earth observation economy, you could say, proliferating, what are those?

00:25:08 Speaker_00
One of the most pressing is how difficult it is to work with the data. I think right now, typically it still requires specialized knowledge of orbits and different types of sensors and how they are calibrated and correlated.

00:25:21 Speaker_00
And I think it's not so easy to apply techniques from one Earth observation constellation to another right now.

00:25:27 Speaker_00
We almost might need some sort of middleware here where we can abstract away the nuances of each Earth observation constellation and make data sets that are really sensor agnostic for non-space engineers to use.

00:25:38 Speaker_00
Changing this from a very specialized climate scientist, GIS scientist, to any sort of ML engineer can start to use these techniques.

00:25:46 Speaker_04
What about regulation?

00:25:47 Speaker_04
I mean, I think about maybe regulation mostly impacting the satellites that go up there, but maybe you could give us a wider perspective of is regulation hindering us at all, whether it is to get the satellites up there or utilize the data, proliferate it, etc.

00:26:00 Speaker_00
Yeah, there's been a lot of exciting regulation changes recently. I think most famously, NOAA agency lowered the restriction for the maximum commercial resolution from 30 centimeters per pixel to 10 centimeters per pixel very recently.

00:26:13 Speaker_00
So this kind of unlocks a lot more higher resolution products. And there's a whole suite of applications that you can start to target.

00:26:19 Speaker_00
On the imagery sort of side, I think that there's been a lot of conversation about licensing and different data sort of rights.

00:26:26 Speaker_00
Right now there's a lot of complex contracts where customers will buy exclusivity and there'll be 24 hours where nobody else can access an image.

00:26:34 Speaker_00
Yeah, and it really hinders folks that are trying to build products with like daily applications or daily monitoring.

00:26:40 Speaker_00
So I think licensing and opening up archive data and making it easier and less restrictive to use could be a good regulation change as well.

00:26:48 Speaker_04
Where are you looking next year and what applications in particular do you find really exciting as we look toward the future?

00:26:55 Speaker_00
I'm really excited, honestly, about the energy sector. I think that's one of the main areas that we can really make a difference with earth observation data. We already touched on predictive analytics for solar farms.

00:27:05 Speaker_00
There's a lot of work with renewable wind sources as well. So I'm just looking for entrepreneurs to take earth observation and really solve our most pressing challenges with these new data sets.

00:27:15 Speaker_04
Speaking of new datasets, we're at a unique juncture where we finally have the data and technology to build entirely digital worlds. But also, these virtual simulations are increasingly having a direct impact on what we can do in the physical world.

00:27:30 Speaker_01
Anduril leverages game engines for defense simulations. Tesla creates virtual worlds for autonomous systems. BMW is incorporating AR in future heads-up display systems. Matterport revolutionizes real estate with virtual walkthroughs.

00:27:47 Speaker_01
That was Troy Kerwin, partner on the games team at A16Z.

00:27:51 Speaker_04
And here's his big idea.

00:27:53 Speaker_01
The big idea is games are essentially virtual simulations, and those virtual simulations have been designed for fun over the last couple decades, but increasingly we're gonna be seeing them used in the real world for all kinds of use cases, whether it's training and learning and development, or training grounds for robotics and other autonomous systems, or visualization to allow folks to see things come to life in real-time 3D.

00:28:21 Speaker_04
Amazing. And I love this prediction because we've actually seen versions of this from the past, and this idea that we all have games and the gaming industry to thank for a bunch of technologies that exist outside of gaming.

00:28:33 Speaker_04
So can we talk about those of the last couple decades maybe? And also, are there any that you think maybe are overlooked? Everyone references GPUs today as maybe one example, but are there others?

00:28:44 Speaker_01
Totally. People forget that Nvidia was a gaming company.

00:28:47 Speaker_01
Almost all of the revenue in the early years was for gaming graphics cards, these new processing units for computationally intense and matrix multiplications, which were great for rendering images and animations and videos.

00:29:01 Speaker_01
But then we soon found that it was useful for things like cryptocurrency mining.

00:29:06 Speaker_04
And of course, now it feels like everything, digital biology. The idea of this accelerated computing is now being used just about everywhere.

00:29:12 Speaker_01
Totally. And I was looking back at some of NVIDIA's earliest websites and the headline was the future is 3D. And it's so funny that 25 years later, while it's been slower than one would have hoped for this real time 3D to intersect.

00:29:28 Speaker_01
with all these other industries. And we'll talk a little bit about why we think now is the right time. But, you know, if we go back, the 90s was about text on the internet. The 2000s was about images. The 2010s was about video.

00:29:41 Speaker_01
And we feel pretty strongly that the 2020s is going to be about interactive 3D and gaming technology being used in the enterprise.

00:29:49 Speaker_04
Super interesting, and maybe just to take a step back, why is it that games or the gaming industry and the technologies that are derived from that, why is that a crucible for innovation?

00:29:59 Speaker_01
I mean, Jensen said himself that he allowed consumer spend to fund the R&D to bring it to what it is today. And I think that that's an interesting lens to think about gaming technology. In the gaming industry, technology innovations are celebrated.

00:30:15 Speaker_01
It's new technology, whether it's new platforms or new features or evolutions that allow new game designs to emerge and flourish. And at the end of the day, like the gaming community, both players and developers, it's a hacker mentality.

00:30:29 Speaker_01
And so it's no surprise that's where big breakthroughs have emerged in the past, and we're going to see them continue to emerge.

00:30:35 Speaker_04
Yeah, and some of those breakthroughs aren't always obvious as breakthroughs. A good example is multiplayer, right? Multiplayer has existed forever in gaming.

00:30:42 Speaker_04
And then it took a while for that to really penetrate or really companies were built off the idea of multiplayer. You take something like Figma, right? Exactly.

00:30:51 Speaker_04
So on that note, today in the gaming industry, there's still lots of innovation happening that maybe again in a decade or so we'll see elsewhere. And so let's talk about those tailwinds. You talk about three in your big idea.

00:31:02 Speaker_04
Maybe you can just talk about each one and how you're seeing that reshape.

00:31:05 Speaker_01
Before A16Z, I was at Unity for close to five years and got a front row seat to seeing how all these various industries were beginning to experiment with real-time 3D for some of the things that I talked about before, whether it's like visualization for architects to be able to walk through their design before it's constructed and they can see if there's errors or other sort of imperfections that they wish they had known when they were designing or for automotive manufacturers.

00:31:34 Speaker_01
They use real-time 3D for also the design, but also virtual test driving. And now the sort of heads-up displays that you see in Rivian is powered by Unreal. BMWs is powered by Unity.

00:31:46 Speaker_01
And then there's the virtual training, whether it's for heavy machinery operations or other operations tasks.

00:31:55 Speaker_01
But some of the bottlenecks for a lot of these use cases that seem so obvious are really some of the same constraints that game developers have faced. And so it's bottlenecks on the content creation side.

00:32:07 Speaker_01
Within a game studio, more than half of the spend goes towards creation of the assets and the art and the content that goes into these virtual simulations.

00:32:17 Speaker_01
And the same is true for these non-gaming use cases, except they don't have 3D artists on staff to build those. And so now when we have AI for asset generation, whether it's images or audio or now 3D assets, it makes that so much easier.

00:32:33 Speaker_01
So that's one. The second is for 3D capture techniques. So of course, for a lot of these non-gaming use cases, they want to capture the physical world as it's built and as it's seen.

00:32:46 Speaker_04
There is a correct version.

00:32:47 Speaker_01
Yes.

00:32:48 Speaker_04
In a way.

00:32:48 Speaker_01
Yes.

00:32:49 Speaker_01
And there's been technologies in the past that have allowed this, things like photogrammetry, or in the case of Matterport, for instance, where it's basically just a 360-degree image, but you can't actually interact with the environment the same way you can with a video game.

00:33:04 Speaker_01
Well, now with newer technologies, NERF or neural radiance fields of a couple years ago, and more recently, other radiance fields technologies like Gaussian splatting, which allow consumers to capture in a much more efficient manner, and it's photorealistic, lifelike.

00:33:22 Speaker_04
And it's immediate, right, in terms of the capture?

00:33:25 Speaker_01
Exactly. And so it allows these use cases to be unlocked. So it's the second.

00:33:29 Speaker_01
And then the third is for some of these non-gaming use cases, this is where we're going to see the prevalence of XR and being able to, going back to the construction or architecture, but put on a headset and see how the BIM model overlays on the construction site.

00:33:46 Speaker_01
or for medical surgery simulation or other use cases like this.

00:33:52 Speaker_01
And as we have better headsets, lighter with eye tracking and other amazing technologies, there's still lots to come in terms of development there, but I think that's going to unlock some of these. Absolutely.

00:34:03 Speaker_04
And as we talk about all three of those tailwinds, so again, the content creation, the capture that you mentioned, and then the devices, it feels like each one of those has their own cost curve.

00:34:13 Speaker_04
And we're traversing down that cost curve pretty quickly across all three. Can you speak to the economics there? I mean, you touched on it a little bit in terms of even games, you said 50% goes toward content creation.

00:34:24 Speaker_01
Yeah.

00:34:24 Speaker_04
So how quickly is that dropping? And then same thing for the devices.

00:34:28 Speaker_01
Yeah, it's interesting, particularly for these non-gaming use cases, some of which photorealism is everything. And that's why as Unreal and other 3D engines have progressed towards photorealistics, these use cases have been unlocked.

00:34:41 Speaker_01
But for other use cases, actually you don't really care what the BIM model looks like so long as it has utility for you. And so as some of these asset classes are up to par with what they would expect to use for these, The cost dropped dramatically.

00:34:58 Speaker_01
But more importantly, I think, particularly if you think about virtual simulation, virtual training use cases, where let's say we wanted to train our workforce on how maintenance and repairs for a robot or some other piece of equipment.

00:35:15 Speaker_01
Well, you would build this experience and you would fund the development of this virtual simulation.

00:35:20 Speaker_01
But then after the fact, if the team wanted to update it or add content to it, they'd have to go back to the outsourced agency who built them the original digital twin. And now they'll be empowered to do that themselves internally.

00:35:34 Speaker_01
And so this content and curriculum doesn't go stale, but they can constantly improve it and update it over time.

00:35:41 Speaker_04
So what you're pointing out is it's not just a one-to-one, how is the economics of creating one thing, how is that changing, but also how it's integrated into the entire system. That's really interesting. Let's talk about applications.

00:35:52 Speaker_04
You've already touched on a bunch, but you've mentioned several companies which are very different, right? Andoril, Tesla, BMW, and then you've also talked about workforce, training.

00:36:02 Speaker_04
Tell me a little bit more about those applications and where does it end? Or is it really, we're seeing it everywhere?

00:36:07 Speaker_01
Autonomy is deeply rooted with these virtual simulations. Anduril is a great example. Funny enough, Anduril's first acquisition was a game studio.

00:36:17 Speaker_04
Really?

00:36:17 Speaker_01
Which would be surprising for a defense tech company.

00:36:19 Speaker_04
I guess if you take Palmer's past.

00:36:21 Speaker_01
True. But they were interested in acquiring it for the game engine that this studio had developed. And they used that technology for strategy simulation and other autonomy workflows.

00:36:34 Speaker_01
And then with other companies, so applied intuition as an example, it's just impractical with the scale of training data that you need to capture this in the real world.

00:36:43 Speaker_01
And so when you have these virtual simulations, you can not only scale the amount of data, but also the fringe and edge cases that you would never

00:36:53 Speaker_01
be able to experience or capture in the real world, whether it's extreme weather or human intervention that is one in a thousand situations. But of course, for these things to be deployed, they need to take into account all these edge cases.

00:37:06 Speaker_04
I remember when we talked to Waymo a year or so ago, they were talking about that, how they ingested all of the crash documentation.

00:37:12 Speaker_04
which exists somewhere on pieces of paper or not in the real world that a Waymo can necessarily interface with at that moment, right? But again, like you're saying, these virtual environments allow you to simulate it.

00:37:23 Speaker_04
And so speaking to that, one way to put what you just said around applied intuition is that you can actually do something new that you couldn't do before with the ability to simulate at scale.

00:37:34 Speaker_04
Are there other downstream opportunities or like second, third order effects that you can think of that we get from these virtual environments?

00:37:40 Speaker_01
Totally. While in the past, we had the ability to use these virtual simulations for physics, training environments, or the learning and workforce development that we talked about. But these were mostly either physics simulations or hard skills.

00:37:56 Speaker_01
But now with, we call them AI NPCs in the gaming context, Whereas before, NPCs were scripted, but now with autonomous agents and LLMs, these agents can take on a life of their own.

00:38:10 Speaker_01
They can observe the environment, they can reason and plan, and then they can act. Well, when you have a multi-agent simulation,

00:38:19 Speaker_01
Now, when we think about the next pandemic response or immigration policies and how those impact a civilization, we're going to be testing these in a virtual environment with these agents who can interact with each other and decision tree out all these different developments.

00:38:34 Speaker_04
Yeah, and instead of just decision tree on paper, what you're getting at is that we actually get to simulate these ideas. Exactly. That existed in the ether. So far, a lot of the applications you've mentioned have been more enterprise-focused, right?

00:38:46 Speaker_04
A company like Android or Tesla. Gaming obviously existed to begin with in the consumer sphere. And so do you see more consumer applications already also coming up?

00:38:57 Speaker_01
So one of the ones that I've just been so excited for is I just moved into a new apartment. And as I wanted to plan out the space, I was still using grid paper and pen.

00:39:09 Speaker_04
Same.

00:39:09 Speaker_01
Despite the fact that we've had the Sims for 25 years, where we can, in a 3D environment, drag and drop furniture and see how it fits. And all the technology exists today to have that experience in an amazing intuitive way, and yet it doesn't exist.

00:39:26 Speaker_01
But we should be able to, and we can, scan our space and develop a digital twin of the 3D environment. We should be able to show it design inspiration that I find from Pinterest.

00:39:37 Speaker_01
have it find the pieces of furniture, the artwork that closest matches my inspiration, fill the scene, and then either be able to walk through it in a virtual world or use augmented reality and see how it fits into your space with your dimensions.

00:39:52 Speaker_01
Like a 3D Wayfair, if you will, where the end consumer has a lifelike digital twin visualization of their space.

00:40:01 Speaker_04
So looking to 2025, so far we've talked mostly about technologies that have been invented over the last few decades.

00:40:06 Speaker_04
But there's obviously this wave of new technologies that are really exciting, haven't really found their footing necessarily in terms of applications. Is there anything you're paying attention to there and maybe how that intersects with gaming?

00:40:18 Speaker_01
Yeah, there's some really interesting research and work being done in the HMI, human machine interaction space, where you can imagine all kinds of different use cases.

00:40:29 Speaker_01
But as with most emerging tech, there's probably going to be initial use cases in gaming that are the wedge for these companies to use consumer spend to fund their R&D, similar to Nvidia.

00:40:40 Speaker_01
So obviously, Apple Vision Pro made huge progress this year with eye tracking. But we're going to see soon BCI type technology that reads energy signals from your brain to actually control and interact with the computer in the virtual environment.

00:40:55 Speaker_01
So we can think about VR use cases where I can use strictly my brainwaves to interact with the scene, which is amazing. And then the inverse is true too.

00:41:05 Speaker_01
And we've seen technologies that allow sort of sensory or digital touch based on solely wearing a ring on your finger for

00:41:13 Speaker_01
increased immersion in the virtual world, which is sort of like the dream of every gamer, to be able to be fully immersed with that haptic feedback, not just in the game controller, but actually throughout your body.

00:41:27 Speaker_04
All right, I hope these big ideas got you geared up and ready for 2025. Stay tuned for parts two, three, and four, where we discuss... The search monopoly ends in 2025. On-device and smaller generative AI models.

00:41:41 Speaker_00
Romanticizing inorganic growth.

00:41:44 Speaker_04
Again, if you'd like to see the full list of 50 big ideas, head on over to a16z.com slash big ideas. It's time to build.