Thanks Matt, I think this more or less answer my question: Custom License for Software and Intellectual Property only
The HTM theory is both a general theory of cognition as well as the basis for machine intelligence. The NuPIC open source project is one implementation of HTM. We are confident that the ideas behind HTM can be implemented in many other ways. Certain implementations of HTM, such as creating custom microprocessor architectures, will require optimized versions of HTM. In these cases, developers may wish to negotiate a license for intellectual property only without using our software. Inquiries for IP licenses may be sent to [email protected]. A note to researchers: We are excited to promote the use of HTM for research and exploration. Consequently, we want to make clear that we will not assert our intellectual property against developers who are using our technology, including our intellectual property, for research purposes. If you are working on a non-NuPIC HTM implementation, you only need to engage in an IP license if you plan to use the technology for production or commercial purposes On Mon, Feb 22, 2016 at 11:33 AM, Matthew Taylor <[email protected]> wrote: > Hi Sebastian. This is a very good question for our business mailing > list: http://lists.numenta.org/mailman/listinfo/busn_lists.numenta.org > > It doesn't get a lot of traffic, but our CEO Donna answers > business-related questions like this one. Maybe you can post there? I > can tell you that we have been forming a strategy about how to treat > businesses that use HTM theory, but not our OS codebases. > > In the meantime, you can read through through this if you haven't > already: http://numenta.com/assets/pdf/apps/licensing-guide.pdf > --------- > Matt Taylor > OS Community Flag-Bearer > Numenta > > > On Sun, Feb 21, 2016 at 6:51 PM, Sebastián Narváez <[email protected]> > wrote: > > Hello Numenta. So, I'm not really thinking on doing anything at the time, > > but out of curiosity, what kind of policy do you have in respect to > > applications or algorithms that based (fully or partly) in the HTM > theory, > > but that doesn't use Nupic? > > > > I'm talking about technologies heavily inspired by the HTM theory, but > not > > completely faithful to it (hybrid, maybe?) > >
