Emergent Properties
We have news! Our Medium posts are now syndicated on Towards Data Science!
Collaborating and grouping into collectives like this give my company more exposure, and also helps in the formation of a community. My idea to start writing these posts was spawned watching UCIC CEO Leor Grebler spend time every day writing about AI voice technology. Writing your ideas regularly can work as a long term goal.
When we watch the progress of the machine learning world, specifically in developing a community, there is already excellent progress. For example, I submitted a journal article where the programming used the theano framework with an OCR that does not use state of the art deep learning. The first comment of the first reviewer was basically “Why didn’t you use tensorflow? Everyone uses tensorflow. Use tensorflow for OCR.” Most people would view this comment as a small-minded bad review. I don’t. We are moving to higher and higher levels of abstraction, and to do that the community needs to pick winners and losers among the tools that we use. This is like the slow slide of assembly language into obscurity, but in fast forward.
When we group together we will find that the sum of our parts is worth more than the parts themselves. Classic examples of building up levels of functionality on top of each other are subatomic particles to planets (e.g. atom, molecule, protein/DNA, cell, tissue, organ, organ system, organism, population, community, ecosystem, biosphere) and hardware systems (e.g. transistor, logic gate, datapath, core, multiprocessor, chip, package, board, system). To make this work, we need to all agree on how to get from one abstraction level to the next. This is the invisible hand of the marketplace of ideas, pushing us to standardize, organize, and grow. As we reach high levels of abstraction, cool things happen that would be otherwise impossible. This is happening now in machine learning and IoT.
Exciting times!
-Daniel
daniel@lemay.ai ← Say hi.
Lemay.ai
1(855)LEMAY-AI
Other articles you may enjoy:
