AU Guys! A Recap of Autodesk University 2016

By: Darrell Westcott, AIA, LEED AP

As my first time attending a major industry conference, Autodesk University was quite the experience. The bombardment of new technologies (and experts utilizing them) was initially overpowering, but as AU2016 ended, I left feeling more excited than overwhelmed. One of the most concise speeches that helped put the conference in perspective was the opening keynote by Autodesk Chief Technology Officer, Jeff Kowalski. 

"Technology has always helped us express our ideas to the world around us, but the specific tools we use have always been the regulator by which we could adequately express those ideas. They have been the filter through which all of our best ideas had to pass."

The general belief at AU is this is no longer the reality, in fact, I would witness an emerging toolset that helps amplify our ideas.  

Jeff Kowalski gave examples of four radical technologies that are converging right now, changing the future of work, and giving us infinite expressibility: Machine Learning, Generative Design, Virtual Reality and Robotic Systems. Out of these four technologies, I was most fascinated with the portion about machine learning.    

Machine learning, otherwise known as artificial intelligence, has the clearest examples in the gaming industry.  

- More than 60 years ago, a programmer taught a machine to beat humans at tic-tac-toe.  

- 45 years later, IBM's Deep Blue beat reigning world champion Kasparov at chess

- In 2011, Watson beat two humans at Jeopardy. This last accomplishment is an incredible leap when you realize that rather than working from predefined recipes and algorithms, the computer had to use reasoning to overcome its human opponents.  

- This year, a program called AlphaGo, beat the world's best human at the board game, Go. Go is a game so complex, it continues to be a topic of mathematical research and is theorized to have more possible moves than atoms in the universe. To win, the program had to develop a sort of intuition about the game; during the match the programmers were unsure why the program was doing some of the things it was doing.  


“So, in less than a single human lifetime, computers have gone from learning a simple child's game to mastering the game recognized as the pinnacle of strategic thought. With technological advances and increased computing power, we have taught computers to teach themselves.”  

Kawalski used one more gaming example to make this concept hit home. He used the Atari game, Breakout, to draw a comparison between human learning and machine learning.

"We learned breakout by spending hours in front of the computer honing our strategies. This is how a recent program (Deep Q-network) developed by Deep Mind learned breakout- It was told that it could only maximize the score by twisting one knob. Then, it learned how to play the game better than any human had ever played it- overnight. How did it do that? It played in computer time. Which means playing millions of games in parallel in the course of a single night."   

"Compare that to how humans share knowledge, just because your friend learned how to get good at breakout, didn’t mean you did. Once this machine learned how to master breakout, all machines learned how to master it, forever."    

In architecture, there are always things to learn. Our projects are incredibly complex, involving a vast number of industries and people. To do our job well, we have to know a little about a lot of different fields, not to mention codes, systems, building technologies and design processes. It is a profession that (thankfully and appropriately) gives value to experience, and because there are so many complexities, usually experience is the only way people can learn.    

However, technology is now available to allow us to better utilize the vast amount of knowledge existing within an organization and disperse it to more people. This helps relieve the bottleneck created when one or two people become the only people who can do a certain task. The experience of a few individuals can be used to help build tools, allowing more people to do specialized tasks.

Image courtesy of Nathan Miller (Founder of Proving Ground) and Matt Goldsberry (the Digital Design Principal of HDR)  |  AU2016 Class: Computational Design-From Specialized Skillset to Core Capability

Image courtesy of Nathan Miller (Founder of Proving Ground) and Matt Goldsberry (the Digital Design Principal of HDR)  |  AU2016 Class: Computational Design-From Specialized Skillset to Core Capability

But, let’s be honest. Sometimes, trying to learn a new set of tools can feel like another addition to the list of things to do “when I have some free time.” And because learning a new technology is a complex task, seeing the immediate value is extremely tough. So rather than carving out the time to learn the new tool, you commit your focus and energy on getting your project done. Unfortunately, it becomes human tendency to find the quickest solution and move on. When left unchecked, this can stifle one's growth and exploration.

We frequently view new technologies as a scary and even threatening. As Kowalski put it:

"These technologies should not be looked at as a threat, they should be looked at as super-powers. The only threat is that a competitor that adopts these super powers more quickly than you do." 

"The reason the prospects of these machines and technologies is scary is because they are so powerful. These tools of imagination and creation are challenging our thinking. Something we haven't really experienced before. But that is the consequence of exponentiating technology- it stretches our thinking, and, our capabilities. It should be exciting, not daunting."   

As the keynote speech began to wrap-up, Kowalski ended with the last and most important piece- the people that use the tools.    

"Just as we should be embracing new technology, we should be welcoming new kinds of talent. Talent used to be about stability, now it’s about mobility. 40% of the US workforce is composed of freelancers, consultants, and other contingent workers. What does this mean to my industry? All of this mobility means we now have access to a vastly larger pool of talented people than we have had in the past. So now imagine the flexible resources you can now bring to bear on any new challenge we face." 
Image courtesy of Nathan Miller (Founder of Proving Ground)   and Matt Goldsberry (the Digital Design Principal of HDR)  |  AU2016 Class: Computational Design-From Specialized Skillset to Core Capability

Image courtesy of Nathan Miller (Founder of Proving Ground) and Matt Goldsberry (the Digital Design Principal of HDR)  |  AU2016 Class: Computational Design-From Specialized Skillset to Core Capability

Autodesk intern investigating new Generative Design technology

Autodesk intern investigating new Generative Design technology

"Today the increased speed of change has created pressure on all of us to learn more quickly. If you're going to keep up with tech and talent you're going to need to up-skill them both at the same pace. If your education stops when you get that one monolithic degree, you're doomed."   
"In this dynamic environment, you can never stop learning. Continual learning is the antidote to fearing technology and new talent. It also happens to be the key to embracing and using it. New technology, new talent, new ideas- for millions of years we have been using this powerful combination to shape our world, but never have we had such an abundance of opportunities, so much to learn, debate, incorporate and create."

AU2016 was proudly described as a time in the earliest moments in an amazing new chapter in the history of making things.  “So, what role will you play in the future of making things?”






Darrell Westcott, AIA, LEED AP is an Associate and Architect with STG Design. He has nearly a decade of architecture experience and is Design Lead for STG's Commercial Practice Group. Darrell enjoys investigating new tools and technologies to help push the limits of design.