The Cloud: Inherently, Unpredictably Transformative
At Chesapeake Systems, we’ve been moving “the edges” of our clients’ infrastructure to the cloud for some time now. Some use cloud storage to collect assets from the field, which are then fed into the central management platform (often still found on-premises). Others use the cloud to host a review and approve web application, which might tie into a post-production workflow. The cloud is obviously used for delivery to both partners and the public at large, and all we have to do is look to YouTube to see how much of a shakeup to traditional M&E that has caused.
This makes the cloud transformative. It is more than “someone else’s server,” even though on one level it is that. But I believe that technologies as fundamental as “the cloud” are often inherently unpredictably transformative. It is difficult to imagine the kinds of shakeups they foretell. And this notion is exemplified by the famously controversial invasion of West Coast cities (and beyond, including Baltimore) by electric scooters.
For those catching up, Uber-like services have been renting electric scooters for short, “last-mile” type trips in major American cities. In classic “disrupt-at-all-costs” style, companies, like Bird Rides and Lime, dropped their rentable scooters off in metro area test markets by the hundreds (and maybe even thousands), without any permitting whatsoever. And Santa Monica, California, embraced them full on. These scooters are EVERYWHERE! Ditched on sidewalks. Being ridden down streets and bike lanes. I can only compare it to such “future shock” moments as watching people play Pokémon GO in NYC, the week it was first released. Essentially, one day there really wasn’t anyone tooling around on electric scooters other than maybe as an occasional novelty, and then BAM! Scooters scooters everywhere, scooting to and fro.
What was the confluence of factors that aligned the stars for rentable electric scooters to accelerate from minor influence to “THEY’RE EVERYWHERE!” practically overnight?
It’s simple. The scooters work easily. The consumer downloads the service’s app, which shows you the location of all rentable scooters around you and their electric battery charge level. Thanks to built-in GPS and cellular networking technologies, the scooter ‘s unique QR-code with the iOS or Android app “unlocks” the scooter with payment kicking in via credit card, Apple Pay, etc. The scooter is not literally locked to anything, but it will not function until you pay for it with the app, which is connected, of course, to your identity (you even have to scan your driver’s license when setting up an account). These services are affordable. And when you’re done, you finish your ride, which locks the scooter, and you put it… wherever. The scooters are gathered up every night, charged, and redistributed around the city by non-employee contractors, akin in a way to how Uber or Lyft contracts automobile drivers.
With lithium-ion battery technology reaching certain performance and pricing levels, GPS and cellular data tech expanding, high smartphone ownership (over 75% in the U.S.), easy mobile payment processing, and QR code technology, scooters went from zero to near ubiquity overnight.
But not without the cloud. What does the cloud have to do with electric scooters? The cloud brings to a startup operation the ability to weave the aforementioned technologies and services into a cohesive system without having to make a major technology infrastructure investment. And for a widely-distributed system, it makes the most sense to put the IT backbone in the cloud. It’s scalable and can easily talk to a fleet of thousands and thousands of mobile devices that happen to also be modes of transportation.
I would submit that without the cloud there would be less of a – or even nonexistent – rentable electric scooter craze. It’s a major supporting piece of the puzzle.
Similarly, that is what the cloud is doing to the media technology space.
Now, we can begin to plan out how to put more of the “guts” of a client’s infrastructure up there, and on-premises systems will soon enough be there only to touch systems which make sense to house at the edges of a deployment. Maybe your MAM’s database, viewing proxies, and application stack will be next to go up to the cloud. Maybe the cloud will house your disaster-recovery data set.
It’s even fairly easy to imagine more of high-performance post-production taking place without any significant on-premises infrastructure beyond a workstation. Or will that, too, become a virtualized cloud system? We can already do this, in fact, in a way that works for some workflows.
What’s further out? Here’s just one scenario:
In five years, significantly more of the software application and platform stack that we all rely on today will be “containerized,” and thus ripe for cloud and hybrid-cloud deployments in a much more sophisticated way than is currently done (in our M&E world, at least — other industries already do this). Software containers tend to use a technology called Docker. You can think of a Docker container almost like a VM, but it has no operating system, just a piece of the overall software stack (a “microservice”) and any software dependencies that piece of the overall stack has.
Management platforms, such as the popular Kubernetes (from Google), allow one to manage the containers that make up a software platform, even auto-scaling these microservices as needed on a microservice-by-microservice basis. Say the transcoder element of your solution needs to scale up to meet demand, but short term? Kubernetes can help spin up more container instances that house the transcoder microservices your solution relies on. Same could go for a database that needs to scale, or workflow processing nodes, and on and on.
All of this is basically a complicated way of engineering an infrastructure that on a fine-grained basis can automatically spin up (and likewise, spin down) just the necessary portions of a unified system, incurring additional costs just at the moments those services are needed. This is, as we know, essentially the opposite of a major on-premises capital buildout project as we currently envision it.
What I described above, by itself, is going to be extremely disruptive to our industry. That’s not to say it’s a bad thing, but it will significantly impact which vendors we work with, which ones are with us a decade from now, how we fund projects, what type of projects can even be “built,” who’s building them, etc.
The notion of a “pop-up broadcaster” with significantly greater capabilities than today’s OTT-only players becomes possible. Want to see if a major broadcasting operation can be sustainable? Rent some studio space and production gear, and essentially the rest of the operation can be leased, short term, and scaled in any direction you’d like.
Many, many organizations do the above every single day. Facebook is doing this, and Google/YouTube, Amazon, etc. just to deal with traffic loads for their websites. In fact, you don’t build a mass-scale, contemporary website without using the above-described approaches.
What’s more interesting than the above, or pretty much anything else we can think of today? What will be more challenging? It’ll be our “scooter” moments. It’ll be the confluence of cloud technologies and many others that will lead to innovators coming up with ideas and permutations that we can’t yet anticipate. One day we’ll be doing things in a way we couldn’t even sort of predict. One day, seemingly out of nowhere … the “scooters” will be everywhere.
To learn more about what AI can do you for you, contact Chesapeake at firstname.lastname@example.org