SXSW 2014 Trip Report

SXSW Interactive 2015 is two months away… you can still get a discount on registration.  I will dust off some useful content to get my blog going — this was a trip report for my  team at work to highlight some pertinent observations (for the process control domain).

Takeaways from SXSW Interactive 2014

This is the second SXSW Interactive I attended – last was 2002, attendance was 3,015 and everything fit in one floor of Austin Convention Center. Buzz topics were blogging, Flash, and wearable computing.

This year, attendance is 10x (30,650 last year), sessions were split between 15 venues, some miles apart (I didn’t attend workshops at UT, but did walk across the river for some panels). Buzz topics were security/privacy, responsive web design, 3D printing, healthcare, and wearable computing. Granted 2002-era wearable was a laptop vest, and 2014 was Google Glass & Fitbit, but I still had a few déjà vu moments.

In terms of other HCD-related conferences, SXSW is where you go to get inspiration and make out-of-the box connections; keep up on the future trends, and hear from the big names in tech. Like CES, but you get to hear about the stuff rather than look at it.
Comparatively:

• UXPA is where you go to get practical knowledge in methods and case studies;
• CHI is where you go for the latest academic research and trends in HCI (though they had initiated a practitioner track when I went in 2012); and
• HFES is where you go for the broader perspective

I’d like to plan on attending every year. Since it is so huge now, it takes a veteran to figure out how to navigate and get the most out of it. I started planning my schedule a month in advance and I would have been lost otherwise, e.g. attending product launches/marketing masquerading as information sessions.

Top Line

• Adobe’s ‘Pipeline Team’ 90 day sprints to innovate and test out new product ideas works
• Most (but not all) getting the difference between User Experience (UX) and UI/Visual design
• ’10 foot experience’ and ‘second screen’ have relevance for control rooms (especially iOps)
• 3D printing (and other ‘personal manufacturing’ technologies) is here, now what?
• Personal sensors and wearables are getting mainstream, with implications for bandwidth, big data, and privacy
• Robotics and Bioengineering are about to break out
• Responsive websites vs. platform apps – blurring the lines

Details

Friday

‘What Do We Build Next’ – Adobe

http://www.afhill.com/blog/sxsw-2014-how-adobe-decides-what-to-do-next/

Innovation Pipeline teams consist of: Product Manager + Developer + Designer + Researcher (UX). Team works with entire organization to get ideas, then they do 90 day sprints to research solution, build prototype, and test it out. Many times they fail, but parts of failures end up in other projects. Presenter also mentioned that in the beginning they got hung up on defining ‘The Process’ and it quickly became the tail wagging the dog – unwieldy, so they caution against it. To smoothly transition idea to implementation, they pull in the product manager and engineering manager to participate with the Pipeline team.

They use an ideation process called ‘Brainwriting’ which is like the Round Robin, but less direction: 8 participants – 122 solutions – 30 min.

Their prototypes were based on a common browser-based interface, which probably made it faster to develop. They also mentioned that the prototypes were instrumented , so they could get quantitative feedback.

The projects they showed were interesting, because they address (for the web page designer) common issues for control graphics builders:

How to aggregate all the different ‘stuff’ in a page (font family, size, colors, CSS tags, etc.) and select them in the style browser to see them highlighted on the page.
How to share a design and get review feedback.
Hard to see here, but they did a project ‘Floto’ which was a command-line driven prototyping tool. You type and the auto-suggest will pop up elements to click on. Then you can go back and style them.

UX & Video Games

Since many of the new generation operators are gamers, or grew up with gaming, I was interested to see what relevance this panel had for traditional UX – I was impressed to hear they didn’t see a difference: UX in games wasn’t just UI, look and feel, visuals, or subset of rules/patterns specific to the domain, but application of the traditional role. Glad to hear somebody mention the 150 year history of psychology (Taylor etc.) in a UX talk, not just start with the browser-era.

The role of UX on the team is a “dedicated person to fight for the user and help developers realize this”.

Some specific takeaways:

• Design for the intended experience: Porting directly from console to PC or vice versa is always a bad idea. Console (Xbox etc.) is a ’10 foot interface’ where users can’t scroll through pages of settings, where PC gamers have mouse + keyboard dexterity and demand ultimate configurability.
• Don’t make it ‘simpler’. Sinclair: “old UX joke—the only intuitive interface is the nipple. Everything else is learned.”
• ‘fixing it in the sequel’ is an expensive way to learn
• Doing good UX doesn’t mean spending more money – UX is the ‘measure twice, cut once’ approach to development.
• Second screen experience of WiiU gamepad is not there yet, but shows the potential of a connected device synchronized to the real-time gameplay to do ancillary tasks (help, change settings) or even have a second player cooperate with you

Future of Making

Tim Brown (CEO of IDEO, wrote the book on Design Thinking) and Joie Ito (Director of MIT Media Lab). Demonstrated a few of the projects from the MIT Media Lab (see them at http://madeinthefuture.co/ ) and pontificated on art & design and the new manufacturing. Innovation comes from divergent thinking where our systems generally promote convergent thinking (Brown). Art is the creative brain, Design is the symbolic brain (Ito). For art, the tool has to disappear (Ito); today we don’t build interfaces that let people get into flow (Brown). That sort of thing.

The new manufacturing technology was interesting – they talked about going beyond what you normally think of the ‘Makerbot’ 3D printer, where you can create a 3D printer ‘larger than the gantry’ to create large scale objects, and nano-scale assembly – using small modules repeated enough to make strong aircraft wings – or anything (think honeycombs). Bullets:

• Maker movement doesn’t replace large scale manufacturing – brings it together (Brown)
• We all need to learn bioengineering like we needed to learn internet 20 years ago (Ito)
  • Bioengineering is moving 6x faster than Moore’s law
  • It is now easy to print genes, stick in a cell and reboot
• ‘economy of frugal reality’ – how do you design in the field with the people who are making
  • HCD is long cycle – prototyping directly in the field is faster
  • Example: Asian cellphone markets, design, build and sell outside the building in 30 minutes. Go back and tweak and get another version out an hour later.

Future of Networked Humans

My expectations were to see how this could apply to mobile workers, outside operators etc. However, this was a session on health sensors (see http://www.uscbodycomputing.org/ ) and how we have the potential soon for continuous, integrated experience where a person could ingest or implant a durable (2-3 year) sensor that is networked. It would continuously upload data to the ‘cloud’ where it is aggregated, resulting in continuous diagnostics and tailored, targeted therapy (gold standard) vs. one-size fits all today.

Brushing aside the health domain (and how Snowden et. al. would respond to all our personal bio data streaming around the net), I thought about my original take: how does the plant operator experience integrate human sensors (ID cards with location, wearable HAZMAT sensors, etc.)?

From Every Screen to No Screen – Next Generation Responsive Design

This was well attended by all the web/app guys, hoping to learn the holy grail of design once, run on many different platforms. The UI manager from Netflix, who has had to do this on 1000+ devices (web browsers, tablet & phone apps, hardware, etc) spoke from experience. ‘Wearables’ is the next headache – how do you fit your web site on a watch face?

Some bullets:

• Mobile first in developing countries – they start with smartphones, not desktop browsers
• In the global market you can’t rely on app stores – in some countries the government determines your OS not the user
• More people are using the Xbox for apps/web browsing than games – with gestures, too
• Question: should experiences differ across devices?
  • A #1: wrong question; think functionality + context, not hardware
  • A #2: we are in dark ages of understanding the customer; companies are making assumptions based on the device—maybe I don’t want the simple mobile website just because I am on a smartphone, I just don’t want to walk into the living room to get my laptop
• Big Question: Should I make a Responsive Web or an App?
  • Web is open, App store is walled (you have to go through Apple if you want your app on iOS)
  • App gives more access to hardware capabilities (gyro, GPS, camera, etc) – but platforms are beginning to expose more of them through the browser.
    • In the future, the app vs. web experience is going to blur even more – “browser will win back what it lost to apps within 3 years (Tim Hayden – strategist)

DARPA: navigating the rapidly expanding infospace

Daniel Kaufman is the director of DARPA Information Innovation Office, but came from entertainment (Dreamworks Interactive) industry, so one of his goals was to make it easier for the government to do business with entrepreneurs and innovators (removing the red tape). He gave an interesting presentation, with the theme ‘the future is here, we just don’t recognize it’; DARPA’s mission ‘prevent and create technological surprise’; and ‘how to think impossible things’. He presented 6 impossible things:

• Safe & Reliable
  • Cyber Grand Challenge – SW that sees an attack before it gets to your PC
  • High assurance cyber military system – produce a reliable real time OS (why do you by a PC that is broken—always patching itself every week?) including SCADA
  • Social Media Strategic Communication – how can you tell if a tweet is from a bot or a human? Benford’s law has been used to identify fraudulent data
•  Empower the End User
  • Memex – allow deep search of the internet (Google only shows you 5% — what makes them money)
  • MUSE (Mining and Understanding Software Enclaves) – computer doesn’t understand Pascal (e.g.) so why do we talk to it that way? Just tell it what you want – system takes high level instructions and pulls from open source code to assemble solution
  • Big Mechanisms – e.g. the cure for cancer exists, but it is spread out in bits and pieces in thousands of sources. The system will read – assemble – interpret, including presenting information to human experts: define research topics for oncology PhD students to fill the gaps.
• Kaufman also talked about how to think impossible – identify the disconnects, look at random things, use active listening (try to have no opinion), look for trends. Change in the world view is uncomfortable, so when you hear a lot of naysayers you are on the right track (this came up in a few sessions).

UX Anti-patterns: Hidden Traps in Sites & Apps

https://speakerdeck.com/mvboeke/ux-antipatterns-sxsw-2014

Anti-pattern: 1) appears beneficial at first, but later has unintended problems, 2) there is a better solution that is demonstrable and repeatable

See the slides for examples – these are all well and good, but it became clear that the speaker was came from a user interface design and not the larger UX/HCD. When it was asked ‘how do you identify an anti-pattern?’ there was no mention of usability testing, even though he did an impromptu test during the talk! UI designers get into this problem because they borrow widgets from other sites/apps, not testing them out in context. It may be that the source used the pattern for a specific reason that doesn’t apply to the borrower’s case. So my rules for avoiding ‘anti-patterns’:

1. User test – verify the pattern works in your context
2. Know the guidelines, standards, and history
   • Boeke’s anti-pattern ‘double jump’ is solved by field masking like that adopted by iPhone:
     ->
     great, but Apple didn’t invent this—it was in the Smith & Mosier guidelines from 1986, and input masking was common in green screen clients.
3. Do human–machine function allocation
   • The anti-pattern of ‘robotic repeat’, where a site asks you to fill in the same data multiple times, or similar validation bug where a form errors out because it expects dollar amount including a decimal (55.00 works but 55 errors), is generally because it is easier for the programmer to make the human type it in again, or type it the right way (3/14/14 or 03-14-2014??) than have the computer either store the input and transfer it (sometimes to another system) or make some simple assumptions and allow entry in multiple formats.
     
     (on a related note several sessions on privacy/data sharing pointed out how dangerous it is to share identifiable data (like birthdate) vs. non-identifiable data (age))

Computing the Future: MIT Scientists Tell All

MIT Computer Science and Artificial Intelligence Laboratory (CSAIL). Tim Berners-Lee got a nerd ovation, but was the weakest speaker of the bunch. His message (besides the 25^th anniversary of the web) was to keep the ‘web we want’ open, neutral, innovative, private, etc.)

Andrew Lo, a financial economist, demonstrated how simple algorithms can be used to get ‘secret’ data from hedge funds and shadow banking without exposing their proprietary info.

Russ Tedrake, director of the center for robotics, showed how far robotics have come, and what they can’t do yet (move like ballerinas, manipulate unknown objects, recognize objects, high level decision making). Google’s robotics buying spree is an indicator of the coming ‘dot com era for robots’; we have COTS robotics and plenty of open source software.

Top Tech Innovation Trends for 2014

Robert Scoble’s job is to scout trends and future applications – he arrived onstage wearing a fitbit, google glass (quite a few at SXSW wearing these), Muse brain-computer interface headband, and a $4K Tshirt OS he got the day before (and gave them a prime demo plug in return).

Some bullets:

• Wall sized displays (TVs) are getting cheap
• 3D sensors are getting cheap (50 cents)
• High resolution audio, including ‘3D audio fields’
• Car is becoming an API
• Home automation is no longer for nerds only (products like Revolv unify different devices..

Creativity in Innovation & Entrepreneurship

John Madea of MIT Media Lab, RISD, and now design partner at Kleiner Perkins VC. This was a conversation between Madea and the editor of Fast Company; no real theme but just a lot of picking his brain on thoughts about design and business. Quick bullets:

• Business (Silicon Valley) doesn’t know the flavors of design – 80% of bias is on visual design
  • “thinking a brand logo is design is like thinking a haircut is health”
• SEC rules for Annual Reports drove the paper and graphic design industry…
• Flipboard is an example of a venture started with design from the beginning
• Google Ventures Design Studio (recently redesigned their web site for the worse…) is a good place for design inspiration
• Make ‘trajectory statements’ instead of goal statements – goals may change
• When talking to customers/users – listen for signs of great pain or suffering
• On leading creative: they aren’t joiners, they are individuals and don’t want to be on the team; though businesses don’t realize the strengths of designers/artists – lateral thinkers, hard workers, lifelong learners
  • Eat with them (Madea’s example of how eastern companies negotiate – eat a meal together first to establish commonality, then negotiate; western – eat the steak dinner afterwards)
  • Personal touches matter (social media is like spray glue – sticks fast, but then peels away. One-on-one is Elmers – long to dry but holds)
  • Vision (at RISD) – build a case for designers – ‘fight for the weird ones’
• If design doubled in speed like Moore’s law, technology would be more human and it would boost the economy.

I attended a number of other sessions, many interesting but less practical takeaways. Several good ones too crowded to get in (would have to skip a session to stand in line) or in venues too far apart.

Some of the information can be found online in Slideshare, twitter, etc. Several sessions were filmed — YouTube has SXSW Interactive 2013 playlist, so 2014 might show up later. [Ed. yes, they post a month after: SXSW Interactive 2014]