Why Moving Video Between Devices Still Feels Broken in the Age of AI and IP Workflows
By Stephanie R
Despite living in a world filled with powerful smartphones, 4K cameras, cloud storage, and AI-driven video processing, everyday media workflows can still feel surprisingly outdated. Industry expert Andrew Cross highlights how even simple tasks — like transferring video from a phone to a computer — remain frustrating despite advances in IP-based audio and video systems. His talk explores why modern technology hasn’t yet delivered the seamless media workflows many expected and what innovations may define the next stage of connected video production. Watch the full breakdown:
Read the full transcript below:
So I now have the pleasure of introducing really truly a legend in our industry. There are very few people that I get to introduce who I can honestly say literally changed how we do our jobs and what tech we use to do it. Andrew Cross is one of those from the early days of new tech but probably in many ways the thing that we see the most today every day no matter what product you’re using is NDI. And so Andrew is going to talk about that journey today. and how they invested and how long it takes for this kind of technology to become mature and get out in the world. It’s a pretty interesting story. So, with that, I’ll pass it to you, Andrew. Thanks, Dave. I’m gonna talk about chapter 2. I realized when I started making slides about chapter 2 that it would be far more fun to actually start talking about chapter 1 first. And so I I wanted to think back to the things in the past that that we did that actually made a difference. And so with the benefit of hindsight, I’m want to go right the way back to when I started in this industry, which was when I graduated from my PhD in 1999 and got my first job. So this is me and I’m a software engineer at heart. So um that’s software, right? And um you know I I got out of um university um in England if you can tell by the accent and I’m like I need a job. I’m a software engineer. Um and you know it was really really awesome. Um I got hired by a company called Newekch. A few of you might have heard about them. It’s a big celebration. And I go to this company called New Tech who who were building these things called video toasters and and kind of cool stuff. And I knew nothing about the industry. Nothing. Um and I kind of knew that the goal was to make stuff that goes onto TV. And so you know back this is my 8-bit art because this is 1999. Um you know TV was done in complicated ways. Typically all the products um you know that that that made up TV in the year 2000 were made out of hardware. They were had FPGAAS and complicated stuff things called resistors and capacitors and all diodes like I don’t know what chips um and they were really really expensive. Um and you know so I I didn’t know anything about this industry. I worked for this company. I just got hired. I was happy. Um got hired by New Techch and and then I learned that all all this complicated stuff is done in hardware. And the problem is um I’m a software guy. Um so I’m the software guy working in an industry that is totally dominated by hardware and expensive stuff and and you know, so I went to my first NAP. Um and this is what everybody told me. Um, Andrew, computers are far too slow to do real-time video on. Like, how on earth can you take a computer and actually process images on it in real time? Like, you need FPGAAS to do this. Um, you never ever ever trust Windows on air. Oh my gosh, what happens when the TV station goes down? We would, you know, software is super unreliable. It crashes. Um, has things called memory leaks and Doom. really awesome. Um, kids play it. Like, what? Seriously, you got to run a TV station on on this thing that my kids play games on and then the other half of the world are doing boring office stuff, you know, the first versions of Excel, spreadsheets, and like Andrew, what are you software guy? How are you going to use this to make TV? No idea. But then we had an idea. Um, here is the bet we made and it is super easy in hindsight to understand this bet, but you have to go back to the year 2000 and look forwards. Here was the bet. There’s hardware, the traditional approach. There are thousands of TV stations in the world and they update equipment every five years, every 10 years, something like that. But here is the bet. The bet is software is going to get driven by millions of users. There’s going to be millions of users playing Doom and Computers are getting fast really quickly, aren’t they? So, people replace their computer every year. And if there’s one thing I know is that the laws of economics are always going to win. And so the bet was that a market that was going to get driven by millions of people is always going to grow faster than the one that is driven by tens of thousands. And so this is me. Here’s the thing. This is the bar needed you need to cross to make real time video stuff. Um, and we were not there. When I started, we were using 486s. You cannot do a lot in real time on a 486. But here was the bet. It takes years to build a product. And in the time it takes to build a product, things are going to happen. We’re going to have computer games are going to drive computers faster. Doom is going to even offices, they’re going to invest in more computers. there’s going to be you know the the the we the economics of the market are going to be such that in the time it takes to build a product that oh I forgot kind of important we bet that networking would go from the thing that you build these token ring networks around to connecting everything together and our bet was that even though it didn’t work today that in the time it took to build a product that we would be able to build all of this and make it run in real time even though it was impossible and we did that and you know that is how the video toaster was born that’s how tricastaster was born that’s how Indi were born was just based purely on those bets the bet was that something that didn’t work today if you just look at where the economics of the market are going to go that you win. And um you know so the bet was that computers would grow faster with software and all the things that everybody told me at my first NAB this is why it will never work but they will get solved where hardware will not. And so we take for granted today that um we take this so much for granted today that this is what TV looks like that you have these amazing virtual sets that you have 3D stuff lying around on sports fields. Um none of that was possible before and it was software that enabled this and by getting out in front and understand what the economics that are going to drive a market are. you have this ability to build products that that change the world because you’re there first in a way that if you follow they don’t. So I I’m going to show you some real numbers now that still blow my mind. This is the shipments of PCs starting in 1999 and um exponential growth, right? by anybody who talks to investors, they love stuff like this. It looks like there’s exponential growth. And for most the industry today, most of everything we’re still doing here, we’re looking at this this this rise in compute, this rise in the accessibility of computers and we’re using it to drive the decisions we make. you know, NDI was based on this um tririccaster, you know, like like virtual sets, everything is based on this curve and we’re using it today. A bell 10 years ago, it’s a little bit more now. Um Apple shipped this thing just down the road from here, the the the iPhone, and it started by selling a few. What’s this? I’m now plotting the sales of the iPhone against the sales of desktop PCs. And um you know, I’ve not finished this curve yet, but obviously they they they beat desktop PCs. And so you you now put this to scale and this is what happened. And this is what happened to desktops. And we say that green curve a few slides ago I showed you exponential growth, right? Does it look exponential anymore? What does look exponential today is what mobile technologies have done for the world. That’s where the exponential growth is. So in 1999, here was the bet. And it’s hard to fully realize this. We started building before it was possible. We knew where things was going and that allowed us to build things that changed the shape of this industry today. And so here we are now. I’m actually onto the subject of my actual presentation. We’re here in chapter 2. This is what the world looks like. It looks like this. We’re still building in large part for that green cuff. And I know one thing. It is that the laws of economics are always going to be what win out. That is because when you’re driven by markets of now billions, that’s how many cell phones are in the world. Um people work out how to make better displays, how to make them cheaper, how to make them faster. Um and the things that we need for our market, which is video, can piggy back on that. And so the question that I think every one of us, literally everyone in this room in every field needs to go off and ask is first of all, what is the bar? Just like then there was this bar. We’re going to get to software. Today it’s not clear what the bar is. And if you’re in production, you need to be thinking where are the laws of economics going to take production. If you’re in products, start thinking about what the products look like. Work out what is next. Think about the time it’s going to take to you for you to get there. And if you can predict what that will be, you have this amazing opportunity to go and build what’s next. Most of us are not doing that. But most of us know full well. It’s easy to look in hindsight and go it was so obvious that computers were going to get faster. They were going to get driven by Mo’s law. Everybody knew that 10 years before they became fast enough for real time. If you know that you can build the products of the future that will shape the future. And I think today we know so much and what we have is amazing. Here is what every talk here is about. AI is solving things we didn’t think were possible 5 years ago. Cameras. I gave my phone to somebody who’s looking after. It’s got a 4K camera running on a battery that we walk around here with. That’s amazing. The that phone is interconnected with every one of you like like our phones all talk the same language and are connected to each other and to every website and cloud service in the world. It has a display on it that has smaller pixels than my eye with my eyesight certainly can see and it’s all cloud with got data centers that can process and store huge amounts of stuff. We have every ingredient like if you think what all of that should enable it is science fiction. We have every ingredient we need for what we would consider science fiction. We don’t have the products to use it yet, though. It’s still really, really hard. If I asked one of you to record a video of me presenting and send it me, you’re going to have to go through 45 steps. It’s not easy. You’re going to have to upload it somewhere. You’re going to have to get my email. Like, somebody’s going to have to work out how to go and take all those ingredients and solve whatever is next. And there is this massive opportunity in every market to do that. And it is totally totally driven by the, you know, the the the laws of economics that are going to get us there. And we know what every pieces and we’ve just got to go out to actually build it. So we have every ingredient. It’s easy to look back and go, computers got faster. We managed to build this. We got to go out and work out what’s next. We’ve got every ingredient. We need to just plot the trends and I know one thing you the answers are going to be given to us by the things that are driving volume that are driving you know and that will take us where we go next. That’s what we all need to go do all of us in every market and if we do that we can absolutely build what comes next and what is chapter two. So that’s everything I’ve got time for. I hope that I’m 15 seconds left. Um, thank you.