SXSWi 2013 recap

Sessions recap, by Nick Martin

Elon Musk Keynote

Really interesting dude in an interview format. He mostly talked about where SpaceX was going, where Tesla was going, and Solar City, which I hadn't heard of before. Basically he's setting out to tackle some of the planet's really hard problems. We got the first glimpse of a self-landing rocket they've been working on. He's (optimistically) thinking we can put people on Mars in 10-15 years.

Niki Weber: Why Your Brain Needs Negative Feedback

She's a business strategist/consultant with a knack for design. Her presentation was actually on using negative feedback to incur positive growth, and many, many examples of that. Some of the more pertinent were giving the feedback in one on one's with her employees. Apparently the millennials cry at first.

Swiss Miss Keynote

She presented on how she went about creating her tattoo business. She really had a great bent on not letting anything get you down - and turn haters into revenue while at it, if possible.

Xobni & Gravity: A Robot in Your Pocket: AI-Powered Applications

Ok, this was just insane. The stuff Xobni is doing is way beyond anything else of seen in the realm of graph theory mixed with machine learning and AI. Basically, they have a large corpus of factors that they make relationships between, pretty much like neurons in the brain. The most they see the same connections, the stronger that connection becomes. This is standard stuff. Where it gets better is the speed that they can do it at (something like < 1ms per email). They claim to gather 500 pieces of information about someone per email. They are actually focusing much more on the enterprise side of things as of late.

Peter Thiel Keynote

He had an extremely interesting take on charting Positivism/Negativity vs Determinism/Indeterminism. Basically, what it would mean to exist at any place on these axes. A positivist outlook believes the economy (the world, really) is getting better, while a negativist believes it will not. Determinates believe they can influence outcomes, while indeterminists believe it's all just chaos. A positive determinist would believe that the world is getting better, and their actions can make things better for themselves too (the US in the 50's & 60's). A negative indeterminist would believe that things are getting worse, and they have no control over outcomes.

Google Glass

I originally forgot to add this one was because I was so disappointed. They sloppily revealed the upcoming Glass API to us. Unfortunately, it's extremely restrictive. You can insert new timeline cards, which are sort of like events that glass can display. However, it's all REST API-based and goes through Google's servers, so no running applications right on the hardware. Maybe they'll let javascript fly...

The Others

Some of the presentations were basically ego trips and infomercials. The only thing of note is that you should make a 30-1 minute video when doing kickstarter-esque things, and make sure to continually post updates, every three or so days.

Streetcap, by Nick Martin

Alto brunch

This was more a chance to inhale margaritas and breakfast tacos than a product pitch. I did learn that Also has ~100,000 users, and 10K concurrent connections, though I couldn’t get a number of active users. They are in the process of building native and mobile versions of Alto, and just completed some sort of rollout that necessitated stopping inactive accounts.


We stopped by the LeapMotion tent at some point and got to play with the actual controller. The current applications were mostly gaming-related, and I thought a bit weaksauce, but they do offer a dev SDK and the device itself is $80. I could see this being a thing in a few different applications - as a musical instrument, creating/shaping things, anything composition related, or even as something overlayed over keyboard/mouse for additional gestures.


I was present for one of their events that basically was all about drinking and OpenCloud. It's basically a mix of Heroku, backups, email, etc. The actual framework is OpenStack I believe, but they are providing a cloud-based variation.

Lessons learned, by Claire

1) There is plenty of room for new communication apps, provided they fulfill a need that isn't adequately met yet. Eg, I could already group text using iMessage, but it only worked with iOS4.1+, and didn't have search. Groupme ended up being a much more useful way to reach different groups of people. It would be awesome if we could add group chat to Gander, though that's likely outside our near-future scope. 

2) Most people are at Sx for parties and networking, though they'll also stand in line for 2 hours to hear Elon Musk. Seriously, though–there are so many boldfacers at this conference, and the few I managed to nab were there for the express agenda of learning about new companies and meeting with people. 

3) Going off 2)–SXSWi is not the place to be discovered. It is an excellent place to grow rapidly, provided you make the conference easier for attendees and go in billing yourself as such. That being said, there weren't any new apps I came across that were filling this need this year. The one new app everyone was using was a one-off Austin Party List app that was pretty terrible, but comprehensive. 

4) Make sure you have an iPhone or an Android. Said party app didn't work on Nick's Lumia; and lots of the apps people are using to connect with eachother aren't Ballmer-friendly either.

5) A lot of people are using or waiting for Mailbox. This being said, nobody I talked to gave it a full endorsement–consensus seemed to be that the swipe categorization was really nice, defer was good, especially in at a conference, but also sometimes detrimental, and the lack of ability to leave older unread emails in the inbox was a real problem. I guess the takeaway here is that people like to have agency (or the illusion of it) over their email, and Mailbox's rigidity prevents that. 

6) On the agency note, there is a real call for DIY, open-source, adjustable, hackable applications and hardware. I guess learn to make is a natural offshoot of the learn to code movement. Some products (eg OUYA) are capitalizing on this quite nicely, positioning themselves as by-the-people, for-the-people, against-the-man. 

Final Observations, by Nick Martin

I hadn't ever heard of the majority of the companies present, but there they were. Apparently there is a large ecosystem, startup bubble or no, and it's getting stronger. There's basically no point to going stealth, because there are probably ten other companies building your idea - much better off getting publicity. Everyone has a terrible code base, and everyone is fighting fires all the time, in smaller orgs. It's better to completely nail a problem for some customers than half-ass it for a bunch. Twilio has an extremely capable API. Apple had zero presence, except for the fact that almost everyone was using their devices. People are willing to pull awful/awesome marketing stunts. Lots of folks seemed to pay attention to what someone had built, or where they worked. You should always know who is throwing the party, and preferably know someone there. Austin is a cool city, and the people there are mad chill. I'd live there.