diff --git a/README.md b/README.md
index abbf2a6..7780468 100644
--- a/README.md
+++ b/README.md
@@ -6,7 +6,7 @@ MOOCs have given us unprecedented access to classroom learning, and from this de
 
 ### WHY?
   - **Democratize the technology**
-    - The architecture of a self-driving RC car is not, in principle, all that different from that of a full-sized vehicle. This puts the full autonomous vehicle architecture, from the sensors, micro-controllers, motors, etc. to the localization, planning and control algorithms, within reach of anyone who wants to learn.
+    - The architecture of a self-driving RC car is not, in principle, all that different from that of a full-sized vehicle. This puts the full autonomous vehicle architecture, from the sensors, micro-controllers, motors, etc. to the localization, planning and control algorithms, within reach of anyone who wants to learn. Online courses have given us unprecedented access to classroom learning, and we’re already seeing a huge surge in creativity, collaboration, experimentation and innovation that comes from a more open-source-centric culture. Let's extend that all the way down to the hardware. There’s no reason you shouldn’t be able to implement what you learned in class yesterday on your own self-driving car today, and then share what you learned in the process in a blog post tomorrow!
   - **Real world data beats simulation**
     - If you have taken or are considering taking any of the available courses from Udacity, Coursera, or elsewhere, you will have the opportunity to learn about and write the algorithms for things like steering control, obstacle detection, and path planning. And in most cases, you'll run this code against canned data or a simulator. But having access to a physical platform allows you to move from code that works under idealized conditions to code that functions in the real world with all its uncertainty and performance constraints. Not only is this more of a challenge, having this experience makes you valuable.
   - **Cost**