Some of the things I learnt at SXSW

April 5th, 2017 by Bryan Hoedemaeckers | Around the OfficeEvents

AR and VR, AI, Smart Cities & IOT, and Driverless Cars. Literally all of the buzzwords!

A couple of weeks ago I went to SXSW, which I can only describe as a brain-overloading, sleep-depriving, inspiration-fest that happens over 10 days in Austin Texas. Every big name in tech, music, and film descends on the very hipster town of Austin to show their wares and help people push to the bleeding edge of each industry.

Let’s just say I heard every single 2017 buzzword at least 10 times, and that the start-up culture in the tradeshow needs to be seen to be believed. Amongst all the partying and networking I did learn a lot, some of the topics I was interested in include: AR & VR, Artificial Intelligence, Smart Cities & IOT, Driverless Cars. There were others, but this post would turn into a book if I talked about all of them!

Below is my high-level brain dump, some of it is my thinking, some of it is direct from speakers, as I said, brain-overload.


AR and VR
I’m going to get this started by saying…

Don’t build interfaces in VR/AR based on things we’ve built for screens!, I know it’s intuitive, but you wouldn’t limit yourself to 1D objects in a 2D space, so don’t do it in a 3D space.

In AR, an apple or book can be part of your UI, you need to get creative with the things that people are surrounded with, and use objects as the interaction points. In VR, buttons are hard to push unless you’re interacting with something physical, so use motion as your interface with haptics as the feedback mechanism, think Minority Report.

In VR you also have to consider physics, well, actually, the fact that you can break traditional laws of physics. Don’t design environments in VR that mimic the everyday world, the power of VR is it’s ability to take you outside of the realms of everyday life, if you’re in VR you should be amazed, you should see magic and other things that don’t exist in everyday life.

Designers working in VR and AR shouldn’t constrain themselves with the same stuff that inhibits us on screens.


Artificial Intelligence
A lot of people are worried about things like Skynet and the Terminator being built by AI once we hit the singularity. Let’s just say that AI isn’t going to create a war between humans and machines. It’s more likely going to be humans — with enhanced intellectual capabilities thanks to AI and synthetic synapses — still fighting the same wars we’re currently fighting, but with one side at a huge advantage over the other. There’ll be way more strategy and potentially less actual fighting. I reckon the military will be the ones that push AI the farthest, but they won’t share it with the rest of us.

For everything ‘non-machines killing us’, Google will take AI the rest of the way, and they will share it with the rest of us. There’s going to be a point in time where we move beyond computer vision and chatbots and the AI agents during that period will be able to do insane things with data, so the companies that are, right now, gathering as much data as possible, are going to hit the jackpot. Those that aren’t, will have an ‘UH OH’ moment.

Don’t disregard data that you can’t do anything with right now, when quantum computing and artificial intelligence mature, that data will be invaluable.


Smart cities & the Internet of Things
Technology is permeating every aspect of our lives, we all have screens we stare at all day long, it’s also taking over our wallets, our cars, our homes, our music collection, the way we travel, the way we sleep, pretty much everything. If you’re not into technology, you should be, it’s everything now. But that doesn’t mean we need to live in a digital world, it’s actually the opposite, technology is getting to the point where it’s driving physical experiences. The world of the future is going to be full of digital things.

Smart cities are no good without smart humans. Just like with home automation, local authorities that want to get into smart cities need to start simple. Most people that have some sort of home automation start with lights, you buy a few light bulbs, hook em up to the internet, and they add incremental value to your life by turning on an off when you get home / leave the house. No-one jumps straight into having a conversational chat bot that wakes you up, addresses you by your name, reads you the weather and plays snippets of the news.

Smart cities will start simple, and as the residents warm up to the idea, they’ll get more advanced. For example, bus stops that converse with you about more than just the bus route, or streetlights that dim when no-one’s around, or even traffic lights that talk to each other and adjust traffic flows accordingly.

When it comes to IOT, one of the big leaps forward will come from low-bandwidth networks. We don’t really want to connect all of our ‘things’ up to our WIFI routers, and we certainly don’t want to buy SIM cards and data plans for our toaster. Low-bandwidth networks will allow us to connect thousands of devices together very cheaply and very effectively across a large area.

If you’re looking to design for Smart Cities or IOT, think about the things that surround us and how they can connect to each other to add incremental value to our lives. My Google Home connects my TV and Stereo to my Netflix and Spotify accounts so I can consume content by asking for it, not by pushing hundreds of buttons.


Driverless Cars
I went to a great talk put on by NVIDIA, the people that make GPU’s so powerful they enable nerds around the world to game at the speed of light in environments that put most CGI movie scenes to shame. The talk, however, wasn’t about gaming, it was about driverless cars, and how NVIDIA are making hardware (computers for cars) and software for the devices that help companies build driverless cars. The main advantage for car manufacturers in the future is going to be the capability of their AI agent in driving a car safely. Safety is the key. Manufacturers are, as I write, training their AI agents to drive cars in a host of different conditions. If you think about it, an infrared sensor and LIDAR can see better than humans in a snowstorm.
One of the issues with this competitive advantage is that car companies won’t share their agents with each other. Which means we’ll have cars out there, driving by themselves, that are more dangerous than others.

A great insight I got from the talk was that NVIDIA is building a system that can take multiple camera inputs, and that could include a camera that tracks the eye movements of the driver. So in instances where the AI isn’t fully in control of the car, it could become a smart-copilot, warning the driver of dangers that are out of their sight, or even recommending autopilot to drivers who are distracted by their phones or other passengers.

If you’re going to push into driverless car territory, don’t ignore the fact that a driver will still need to be present for regulatory purposes for the next decade or so. Think about how your products can enhance the experience of that person, or extend the abilities of that person in either driving the car or existing in the car.


All in all, SXSW was an insane experience, and I would recommend it to anyone that is interested in tech, the future, music, or films, pretty much everyone alive should go to SXSW. Don’t expect to get much sleep.

blog comments powered by Disqus