Nov 21, 2011

Tech4Africa 2011: Bright... and Ballsy!

I recently took a trip to South Africa (where I grew up) to present on the Opportunities around Big Data at a conference in Johannesburg called Tech4Africa. The conference is now in its 2nd year and is the brainchild of Gareth Knight whom I met at SXSW Interactive 2010 and runs Technovated.

I am a bit of a social fellow and I'm all about meeting and learning from new people whether it be at a meetup, a customer meeting or a conference. The conference was great for this and I enjoyed chatting to all the fellow data and emerging technology folks, especially the folks from Quirk. The conference seemed quite like SXSW Interactive in that it was a good mix of creative, business, social and technology types but still small enough that you could really get to meet a lot of the attendees over the course of the 2 days. All the international speakers were housed at the same hotel which led to a lot of great late night conversations on both technology and Africa. Robert Nyman is quite a character.

If you're interested, my presentation is available on slideshare (to the right). I recommend you follow the click through to the actual slideshare page and download it using the Download link, as it appears that slideshare still has a few issues with converting slides in landscape format. I hope to be doing a detailed blog post soon on the data science demo quantifying US Venture Capital investment at the end of the presentation.

Positive Indicators

I got to chat to Guy and Jason from 10Layer in the speaker room while they were getting ready for their  IGNITE presentation and I liked what they were doing, namely, creating technology. They built a new Content Management System (CMS) from the ground up and then open sourced it on GitHub the day of their presentation. Solution plays are great but technologically they typically involve glue code around existing technologies. Something new built from the ground up speaks a lot more to the technical depth of the team as not every team has the deep technical skills required to achieve this. From a software innovation standpoint this is a good example of African innovation following similar patterns to the valley. A valley example of this would be the explosion of businesses built around Apache Hadoop which was created and open sourced in a similar way. Solution plays can be just as disruptive as new technology plays but the latter is able to take advantage of opportunities in the market that have a higher on-ramp due to the deep technical skills required to address them.

However, I would like to be honest in my commentary and I found the current state of non-mobile broadband in South Africa a little worrying. One framework to break down the impact is to consider technology comprised of compute, storage and network. Due to the Open Source revolution, software is largely free and effective. Hardware and Storage is inexpensive due to the advent of scale out architectures of commodity hardware powering a large majority of modern implementations. In addition, if you can't afford the capital expenditure required to purchase the commodity hardware and storage that you need you can always go to a metered operating expenditure model in the cloud. So I think we can say that South Africans have no inhibitors to compete on the world stage when it comes to compute and storage, all that is left is the network. I found myself pretty frustrated whenever I tried to get online. Generally, the service was too slow for me to be fully productive and I often just gave up trying to actually do any work (downloading & hacking on new software/technologies for my job) and just kept it to light weight largely text based social media (Facebook and Twitter) and browsing. Now, I'm sure there are places (like your workplace) where you can have faster access, but, to truly lower the network barriers to create fertile soil that could enable South African equivalents of Hulu and NetFlix, there needs to be consistent and decent (good enough for it to just work) ubiquitous network access at affordable prices. I think the lack of that type of access passively & actively hampers innovation in South Africa. I will add though that I thought data access on mobile devices worked well for what I use mobile devices for.

Interesting Conversations

I enjoyed having the opportunity to spend time with other African speakers and attendees. I don't often get to meet other first generation African speakers at conferences and one of my favorites was Ndibuisi Ekekwe from Boston (by way of Nigeria). Ndi (pronounced "Indy") is a true hardware head (semiconductors/microchips) and provided some pretty poignant insights in a conversation we were having about scale out technology. When we got around to discussing reasoning engines (such as Apple Siri or IBM Watson) he moved the discussion around to the difference in computing power required by modern reasoning engines versus the human brain. For instance, the human brain's energy consumption to compute an answer to a question is massively smaller than the energy consumed by a technology like Watson. Ndi posited that in order to reduce computational energy consumption we need to redesign computational processing to operate similarly to the mammalian brain. This is also known as neuromorphic processing or Cognitive Computing and is currently being explored  by the DARPA SyNAPSE project.

Another interesting conversation I had was with a young man by the name of Victor Rikhotso. It was at the end of the conference and all the international speakers were in the front row getting miked to go on stage for a panel and Victor walked right in front of the crowd and just came and sat right down between us, 5 mins before we were about to go on.  He then proceeded to pepper me with questions about machine learning. The first thing you should know, Victor is 17. The second thing you should know, Machine Learning is an elective class for a Masters degree in Computer Science.This guy is both ballsy and bright ! Victor is from the African Leadership Academy (ALA) and is currently applying to a number of universities to study Computer Science or Engineering next year. I'm planning to stay in touch with Victor. If Victor and his classmates from the ALA are an indicator of the next generation's technological aptitude and propensity, then the future for South Africa is bright.... and ballsy.

No comments: