New to Numenta? Bit of a beginner in brain theory? Then this episode is for you. In their 20 minute conversation, hosts Christy Maver and Matt Taylor break down the top 5 things you need to know about Numenta, including key resources for all levels of interest – from AI enthusiasts to neuroscientists – and how you can get involved.
This episode is designed for beginners. It does not require any previous knowledge about Numenta, brains, or brain theory. If you’re looking for more detailed conversations about our neuroscience research, listen to our first two episodes, where host Matt Taylor interviews Numenta Co-founder Jeff Hawkins to describe the latest research developments.
You’ll hear a couple repeated terms in this episode that are core to Numenta. For those who are unfamiliar with them, we’ve included definitions to help you understand. You’ll also hear Christy and Matt refer to multiple resources, which you can access through the links below.
Hierarchical Temporal Memory (HTM) – HTM, which stands for Hierarchical Temporal Memory, is our biologically based technology. At the core of HTM are algorithms that learn time-based patterns in unlabeled data on a continuous basis. HTM is well suited for anomaly detection and prediction in streaming data.
NuPIC – NuPIC stands for Numenta Platform for Intelligent Computing. It’s our implementation of HTM, available in an open source project where you can find all of our core algorithms and sample application code.
Top 5 things you need to know about Numenta
[1:58] #1 Our Mission
- [2:35] Our primary scientific mission and secondary machine intelligence mission
[4:31] #2 What We’ve Learned so Far
- [4:42] Our first area of discovery: Sequence Memory – how your brain memorizes spatial patterns over time
- [5:46] Matt conducts a thought experiment: Invoking sequences with Beethoven
- [6:40] The “Neuron Paper” – Why Neurons Have Thousands of Synapses, A Theory of Sequence Memory in Neocortex
- [6:49] Our second area of discovery: how your brain learns objects over time through movement, object recognition
- [9:11] Research paper documenting this discovery: A Theory of How Columns in the Neocortex Enable Learning the Structure of the World
- [9:17] For additional resources on this area of our research, visit our Cortical columns page.
[9:21] #3 Applications of our Technology
- [9:42] HTM Applications
- [11:11] HTM Studio – our free tool that allows you to test our technology (HTM) on your data with one click and no coding
[12:55] #4 Our Commitment to Open Science and Research Transparency
- [13:01] Increasing Research Transparency (blog by VP of Research Subutai Ahmad)
- [13:07] Subutai on Twitter
- [13:41] Numenta Research Papers
- [13:55] Numenta Events page
- [14:13] HTM School
[15:24] #5 How You Can Get Involved
- [15:44] For non-scientists: Learn more about us with these 30,000 foot view pieces by Jeff Hawkins: Thousand Brains Model of Intelligence, Secret to Strong AI, What Intelligent Machines Need to Learn From the Neocortex (IEEE Spectrum article)
- [16:36] For scientists: Read our peer-reviewed papers, posters, and preprint journal papers
- [17:00] Join the HTM Forum