SXSW Day 4. Randomness Meets Substance
- March 16, 2011
- 0 Comments
Day 4 started off right, and got better from there. Parking was quick, and coffee in hand I strolled into Ballroom D just as Craig Venter was starting to talk about synthetic life. I whipped out my iPad2 and started taking notes. Yes, I was feeling pretty good about my new charger-free existence (I really do love the long battery life).
Lest you think SXSW is all about frivolous social media and twittering away late into the night (literally and figuratively), Sunday’s sessions brought a random substantiveness to my SXSW experience, and had me thinking about things outside my usual domain. And lest you buy into the hype by some (Vivek Wadhwa you know who you are), that Silicon Valley (and startups in general) are focused on ephemeral, superficial things, Craig Venter just completely shattered my notions of what can be done today in genomics.
As I entered, Craig was discussing their goal: to make a living, self-replicating bacteria cell, driven by artificially generated DNA.
Craig’s construct: to discuss this in terms of hardware (virus, bacteria, yeast) and software (DNA). Several minutes were spent explaining how DNA strands were spliced together to form strands more than 100,000 “letters” long. Fascinating process of experimentation, followed by investing in discovering an automatable process that a robot can perform. Literally. (Not just automated in the BPM sense of the word, by software).
Eventually they achieved a one-step, in vitro assembly that worked at 50’C, which could be automated. More examples could be pursued, larger pieces, more automation, more robotics.
A study in 2007 showed that just by changing the DNA in a cell, they could convert one species into another. And still have a fully functional organism.
Paraphrasing a quote: “We think that the software recognized the DNA of the original species as foreign and ‘ate it up.’ All the characteristics of the original species are gone.” He reminded us how often DNA is replaced – 20% an hour, or something to that effect (I might have missed the exact timing/numbers…) and asked, imagine if you had to change 20% of your car parts every hour to keep your car running. It is amazing stuff that goes on inside our bodies.
To be sure that they had the real, unique organism, they encoded messages within the genetic code – an “easter egg” in the genome – quotes from James Joyce and others.
So now they’re focused on the software for designing new cells and organisms – there aren’t enough scientists to understand all the possibilities, so software will do much of the combinatorics and analysis. But why bother? what’s the point?
- World population of nearly 7 billion has demands our earth and tech cannot supply without continued technological change. 3 people alive today for every 1 alive when Venter was born. Genomics can help address the food requirements of such a population.
- Additionally, flu vaccines, and eventually an HIV vaccine, may be formulated. HIV presents a challenge due to its rapid genetic code changes, so we need new approaches. A flu vaccine using their rig takes 24 hours to produce, rather than the 3-6 months required by today’s techniques.
- Energy demand is increasing even faster than population growth. The Keeling curve shows increases in carbon dioxide in the atmosphere. So bio-engineered micro-algae can do a great job synthesizing sunlight and CO2 into combustible fuels. The questions is can we scale this to get billions of gallons of fuel in 10 years?
If we based our energy needs on corn sugar replacement, we’d need a cornfield the size of 3 United States’. But if we can do it with micro-algae, we’d need a field the size of Maryland. Not easy, but at least it is achievable.
Clearly, Craig Venter is focused on how to solve the big problems. One might say that he is on the path of a “Smarter Planet” but with a completely different approach than our friends at IBM. It is fascinating.
My second session seemed to just gel perfectly with the Venter talk. From Biological artificially created life to synthetic artificial life. The second panel was a discussion of the “Singularity” – the point beyond which we can’t imagine what our life(lives) will be like, because the rate of change is so fast we can’t understand the implications. Usually this is considered to be because an Artificial Intelligence achieves sentience (the Vernor Vinge hypthoesis), but this is not the only thought model. Our three panelists, experts in the field of AI, discussed their views.
First, the difficulty of context for computers. The following two statements:
- Mary and Sue are sisters
- Mary and Sue are mothers
To us, it is easy to see that in statement #1, the obvious implication is that they are related and sisters, not that they both happen to have sisters and are not related. In statement #2, it is obvious they are both mothers, not the mother of each other, and not necessarily related in any way. But this sort of context is hard for computers. Hard for Watson. Doug Lenat theorizes that we are now close to modeling computer systems that will be able to leverage fact systems for context and inference engines for generating outcomes that make sense.
Vassar took the point of view that if the singularity is in 2030’s, the point is to live well now – to enjoy your life, not to worry about what happens after 2030. He had an oddly humanistic angle to his arguments:
- We are not very good at deliberate thought, but we seem to get a lot of mileage from the little deliberate thought we engage in.
- Machines are good at deliberate thought, but not very good at the non-deliberate thoughts that humans do quite naturally.
- If you have a good understanding of intelligence, by definition you have super human intelligence, whether you have an AI or not.
- If we understood ourselves as well as we understood societies, we’d already be creating people.
- Society does things – it does not decide to do things. Society isn’t intentional.
- We tell good stories about caring about things – but looking at our actions you would not conclude that we DO care about these things.
Vita-Moore was focused on the artistic, and the idea of extending life through cybernetics, rather than Ray Kurzweils view of uploading intelligence into a computer. She talked about aesthetics and how we would feel about some of these changes as they happen.
The third session was focused on Business Models -but it was such an introductory discussion that I couldn’t last through it. I left and joined a friend (Amar Rama) for lunch at the Driskill Hotel’s 1886 Bakery and Cafe. We thought it would be far enough from the convention center- we were wrong. It was packed. We got a table, and looked over only to find ourselves sitting right next to Jane Kim of Hashable again.
We missed the keynote of the day, Felicia Day’s talk. From what our friends wrote about it, it was a great session. But we did stop by Vast, saw some of the Startup Bus competitors, and a couple of small Austin startups in that same building. Good conversation followed, before we headed back to the convention center for another session.
My next session was “Abolish the Hourly: How Value Pricing Wins Clients” – a great content-rich session by a couple of successful professional services (agency) firms. Jon Lax and Lee Dale led an entertaining discussion and case walk-through – in a room packed literally beyond capacity – with people sitting in chairs, on the floor, and in the hallway outside. But there two key questions left unanswered, and I’ll have to track down the panelists:
- When we say “value based pricing” – is “value” defined as the value the customer accrues from a successful project? Or is it the value the customer sets as the amount they’re willing to spend to get the project implemented.
- If we take value as customer-accrued value , then we have two values – customer-accrued and customer-willing-to-pay. The question is, what is a spread between those values that works? ie, if the project has $10MM in value to a customer, what should that customer be willing to pay to get the the help they want?
After this session, I walked down to the Four Seasons bar for a Trilogy Alumni reunion. It was great to see all the alumni that were back in town because of SXSW-interactive. It is astonishing to see the level of success and influence our alumni have within their companies – startups in NYC, SF, Bay Area, and Austin. Not to mention the influence they have in the investor community. I think this influence and network is only going to improve over the next 10 years. Thanks to Joshua Baer and OtherInbox for sponsoring the happy hour!
After several hours of catching up with some of the smartest and most engaging people I’ve ever met, we decamped to Blue Ribbon BBQ (reviews on Urban Spoon here, another Trilogy Alumni founded service, seems appropriate). The BBQ was amazing, but they need some work on their BBQ sauce (I know I know, good BBQ doesn’t need sauce, but I still believe if you’re going to put it on the table it better be good!).
Having taken my two visitors (and friends) to fine Mexican dining and fine BBQ dining, I felt my mission of representing Austin’s eating establishments well had been met!
From there, we proceeded to the Hashable party. And then, to the Mashable party. At that point, we declared we were only going to parties that end in -ashable. And yes, a good bit of the value of SXSW is reconnecting with your friends (old and new) at the end of the day to share stories about the day and your lives. Something you can’t do with livestreamed, virtual conferences.
I have one more SXSW post in me (besides my writeups of the Lean sessions), which will focus on my conclusions from SXSW… look for that tomorrow.