Cool Toys Pic of the Day – Grok

Numenta: Grok:
https://www.numenta.com/grok_info.html

You might know the word “grok” from Robert Heinlein’s novels, in which “grok” is used to describe a fairly deep understanding of a topic or question. These days, with the infamous information explosion, it is harder and harder to find someone who can really grok any important question, in part just because virtually all questions have
information spaces that are too large for any one individual to deeply understand all aspects.

Grok, the tool, is less of a “Wow, cool thing you can use now” and more of a “Heads up! Look what’s coming.” Numenta is a company that has come up with a nifty idea for connecting the big data space with artificial intelligence and real time interactive use and modeling of the data. The artificial intelligence part, specifically what Numenta refers to as “Adaptive learning, automated action,” is the part that makes this all easier and accessible for us humans.

One of the responses I keep hearing in my work in healthcare technology is people saying in many different ways, “We need the big data to help us answer all these important questions, but so much happens so fast, we need new tools to help us work with the data.” Enter Numenta, stage left.

Jeff Hawkins on the future of data (short version):
https://www.youtube.com/watch?&v=mP_AbIKXlsg

That’s the short introduction to what they are trying to do. Basically, allowing people to connect their data to outcomes with real-time analytical tools available anywhere / anytime through whatever tech you have in your pocket or in front of you. A grand vision brought to us by Jeff Hawkins, the scientist behind the technology, who was quoted recently as describing this as tech that will, when integrated into our surrounding environment, make it so “everything is intelligent, everything is learning.” The science behind the technology is supposed to be based on neuroscience. You can get a brief overview of the neuroscience aspects here.

The Technology Behind Grok:
https://www.numenta.com/technology.html

If you want to dig deeper, there is a technical report available here.

HIERARCHICAL TEMPORAL MEMORY, including HTM Cortical Learning Algorithms
https://www.numenta.com/htm-overview/education/HTM_CorticalLearningAlgorithms.pdf

If you are feeling really nerdy or geeky, you might want to watch his recent keynote on the tech.

Jeff Hawkins – ISCA Keynote, June 11, 2012

If you are feeling less nerdy/geeky, then you’ll possibly prefer his earlier TED talk.

Jeff Hawkins: How brain science will change computing

Right now, Grok is in closed beta, so I haven’t had a chance to play with it, or see how it is supposed to do what it says. The idea of the product is what is interesting to me, and that someone believes they’ve gotten close enough to making it work to actually say they are in closed beta. Whoa.

Meanwhile, there are a lot of people who are pretty hesitant about whether or not this can possibly work. There is a piece from a a couple years ago by Michael Anissimov in which the tech is described as a disappointment, followed by a rich and energetic discussion about the flaws in the tech over at Hacker News.

Ben Goertzel on Numenta (April 2010)
http://www.acceleratingfuture.com/michael/blog/2010/04/ben-goertzel-on-numenta/

Hacker News: The Grok prediction engine from Numenta announced http://news.ycombinator.com/item?id=3933631

Since they’ve gone into beta, things have been quieter. I don’t know if that is good or bad, but I still find it encouraging that people are working to move along this path in a generally forward direction.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s