Thanks Matt and Donna, That's a great clarification, and even contains the exact answers to questions not posed originally in this thread.
This will become ever-more important as HTM matures and new people (customers and practitioners) come in. Regards Fergal Byrne On Fri, May 22, 2015 at 4:34 PM, Matthew Taylor <[email protected]> wrote: > Fergal, > > Here is Donna's response to your message below. She doesn't no > subscribe to these mailing lists, so I'm sending her message in proxy. > > ======================================= > > Fergal - I appreciate you taking the time to write up your impressions > of our commercialization strategy. I am particularly grateful that > you called me a sensible businesswoman, although referring to my many > decades makes me feel old!! > > Generally, I feel you have represented the situation correctly. There > are a few minor tweaks that I would make, so let me do that here. > > First, a few corrections. We do not have 58 patents. Perhaps this > number includes abandoned applications - I'm not sure. On HTM related > technology, we have just over 30 issued patents, with more pending. > We're very pleased and proud of this portfolio as it demonstrates that > our work is novel. > > Second, the IP isn't actually "donated" into the open source project. > Basically, the open source license that we selected, the GPLv3, > includes a "patent peace provision". This provision means that as > long as you are using this technology under the open source license, > you are permitted to use the patents as well. You might think of it > as the license including both the code and the patents. If you are > not using the open source license, then you do not have a license to > those patents. Now, that being said, we have made a clear statement > of non-assertion of patent rights for non-commercial purposes. So, if > you are an academic, or you are just experimenting with the technology > for a research organization, even if you do NOT use our open source > code/license, we have declared that we will not assert those patents. > So, the only case where a patent license would be necessary is if you > are not using our open source license and want to create a product or > to use it internally for commercial purposes. I should note that it > is totally fine to start by using our open source software and > license, and then to request a switch to a commercial license. > > Okay - now to the enumerated points. > 1. "If you don't develop proprietary software, you can act as a > consultant on projects without any need for a license." Yes, this is > correct. > 2. "If you develop a fork of NuPIC, you can do so as another open > source project." This is close, but to be more precise, the terms of > the GPLv3 require that if you distribute derivative software, you must > also distribute it under the GPLv3, often called "copy left". You > cannot create a new open source project with our software under a > different open source license, such as a BSD style license. > 3. "If you have developed something commercial, contact Numenta for a > license." Yes, this is correct. Once you want to distribute > something, which in our interpretation includes SaaS implementations, > you will need a commercial license unless you are prepared to > distribute under the GPLv3. > > Next, the question of Numenta's financial status. > We are principally self-funded, with a little bit of investment from > friends and professional investors. As Fergal points out, we have a > long term view, so do not want to be inappropriately pressured into > monetization strategies before it makes sense. We do not disclose how > much money we have invested or at what valuation. > > Finally, as to the question of our "start-up package". > We are in the early stages of thinking through a strategy that would > enable start-ups to get a commercial license to our technology > (including intellectual property). The idea is to exchange a license > for a small stake in the company and a small royalty on revenues. We > recently completed our first deal along these lines with Cortical.io. > We believe this structure aligns us well in that we will be very happy > to help Cortical.io succeed, and it requires no cash from the > start-up. We are thinking through whether to formalize this structure > as a program, or whether to just handle partners on a case-by-case > basis. If this structure interests you, feel free to be in touch. > > =========================================================== > > If anyone wants to start the process of getting a commercial license > to use NuPIC or HTM technology patented by Numenta, please use > https://numenta.com/contact > > Regards, > Matt > > On Fri, May 15, 2015 at 1:03 PM, Fergal Byrne > <[email protected]> wrote: > > Hi guys, > > > > (Anyone commercial/lawyerly from Numenta please correct anything here, > this > > is just my impressions. IANAL). > > > > Jeff and others in Numenta hold 58 patents (last time I checked) on > > HTM-related technology. Much of this IP has been irrevocably donated into > > the open source project NuPIC. My understanding is that any OSS project > can > > freely use this IP, but a non-OSS project may have some issues. > > > > As one of the first people to talk to Numenta about commercial > exploitation, > > I had a great meeting with Donna Dubinsky in November 2013 (part of this > was > > on behalf of Francisco of cortical.io who had to leave SF just after the > > hackathon). Donna laid down some basic principles: > > > > 1. If you don't develop proprietary software, you can act as a > > consultant/configuration/applications expert on a project for a client > and > > you don't need to discuss with Numenta (this is what I do). > > 2. If you want to develop a fork of NuPIC, or anything derived from > > Numenta's patents but encompassed in NuPIC, you can do so as another open > > source project and inherit some version of NuPIC's license (GPL-like). > You > > are completely free as long as everything you do is open source. > htm.java, > > clortex, comportex, and a few more are doing this (I do this too with > > clortex). > > 3. If you have developed something which is proprietary and commercial, > you > > should contact Numenta and discuss how to proceed. Numenta's stated > mission > > is to be a "catalyst for machine intelligence" so you will find them very > > amenable to what is going to make your project work (yet to do this). > > > > As Matt said, this may (as in the case of cortical.io) involve Numenta > > taking a stake in your company. This is effectively a form of investment > by > > Numenta and is only one way to do it. If you think you have a billion > dollar > > idea which needs HTM technology, but don't yet have the $20m to give to > > Numenta, a small stake in your business might make sense to both parties. > > > > If you're IBM, on the other hand, you might wait until you've identified > a > > clear business opportunity, and you would prefer to offer Numenta the > $20m > > in cold cash rather than a chunk of that business. > > > > In the middle, you'll find Donna is a sensible businesswoman, who knows > how > > to strike a deal that makes both sides happy. She's been at this for a > few > > decades with great success. > > > > If I'm wrong about any of this, I completely misinterpreted my > conversation > > with Donna, so someone please correct me. > > > > On Numenta's financial affairs, I'd say that there's a reason Jeff and > Donna > > decided to keep the company a private business. The two of them have > already > > seen what happens when founders are forced to bow to corporate forces, > and > > how their vision becomes compromised. They're lucky (and/or good) enough > to > > have emerged from that with the resources to build Numenta and share this > > with us. Numenta will prosper because there is a great need for experts > in > > this particular technology, and Numenta has 90% of them in-house. > > > > Donna, > > > > the above is my interpretation in broad brush strokes of your policy. If > > there's anything grossly amiss please let us know. If it's broadly right, > > please indicate "that's broadly right" but people should actually get in > > touch. > > > > I hope that helps. > > > > Regards > > > > Fergal Byrne > > > > > > > > On Fri, May 15, 2015 at 8:20 PM, Alexander Kettinen < > [email protected]> > > wrote: > >> > >> Thanks Matt, last question: what is your current yearly run rate? (If > you > >> can comment) > >> > >> Alex > >> > >> Skickat från min iPhone > >> > >> 15 May 2015 kl. 22:12 skrev Matthew Taylor <[email protected]>: > >> > >> Contact Numenta via http://numenta.com/contact/ > >> > >> Since Numenta is currently a private company, I don't think I should > >> comment on our valuation. > >> > >> Also, no plans for an IPO in the near future as far as I know. > >> > >> --------- > >> Matt Taylor > >> OS Community Flag-Bearer > >> Numenta > >> > >> > >> On Fri, May 15, 2015 at 11:56 AM, Alexander Kettinen > >> <[email protected]> wrote: > >> > >> Matt, it's a bit premature but what is the contact point for a legal > team. > >> > >> > >> Also do you have a internal guesstimate on the value of numenta as per > >> today > >> > >> ? I know it is privately held. > >> > >> > >> Now I must attend my dinner party > >> > >> > >> Have a great weekend > >> > >> Alex > >> > >> > >> Skickat från min iPhone > >> > >> > >> 15 May 2015 kl. 21:45 skrev David Ragazzi <[email protected]>: > >> > >> > >> Talking stock... When Numenta will go to public? :-p > >> > >> > >> ..or have Numenta interest on this? > >> > >> > >> On 15 May 2015 at 15:41, Matthew Taylor <[email protected]> wrote: > >> > >> > >> On Fri, May 15, 2015 at 11:39 AM, Alexander Kettinen > >> > >> <[email protected]> wrote: > >> > >> Matt, can you define partial any further? > >> > >> > >> Meaning stock ownership. > >> > >> > >> --------- > >> > >> Matt Taylor > >> > >> OS Community Flag-Bearer > >> > >> Numenta > >> > >> > >> > >> > >> > >> -- > >> > >> David Ragazzi > >> > >> Master in Sofware Engineering (University of Liverpool-UK) > >> > >> OS community commiter at Numenta.org > >> > >> -- > >> > >> Have you tried NuPIC Studio? Just check out > >> > >> https://github.com/nupic-community/nupic.studio and enjoy it! > >> > >> -- > >> > >> "I think James Connolly, the Irish revolutionary, is right when he says > >> that > >> > >> the only prophets are those who make their future. So we're not > >> > >> anticipating, we're working for it." > >> > >> > > > > > > > > -- > > > > Fergal Byrne, Brenter IT > > > > http://inbits.com - Better Living through Thoughtful Technology > > http://ie.linkedin.com/in/fergbyrne/ - https://github.com/fergalbyrne > > > > Founder of Clortex: HTM in Clojure - > > https://github.com/nupic-community/clortex > > > > Author, Real Machine Intelligence with Clortex and NuPIC > > Read for free or buy the book at https://leanpub.com/realsmartmachines > > > > Speaking on Clortex and HTM/CLA at euroClojure Krakow, June 2014: > > http://euroclojure.com/2014/ > > and at LambdaJam Chicago, July 2014: http://www.lambdajam.com > > > > e:[email protected] t:+353 83 4214179 > > Join the quest for Machine Intelligence at http://numenta.org > > Formerly of Adnet [email protected] http://www.adnet.ie > > -- Fergal Byrne, Brenter IT http://inbits.com - Better Living through Thoughtful Technology http://ie.linkedin.com/in/fergbyrne/ - https://github.com/fergalbyrne Founder of Clortex: HTM in Clojure - https://github.com/nupic-community/clortex Author, Real Machine Intelligence with Clortex and NuPIC Read for free or buy the book at https://leanpub.com/realsmartmachines e:[email protected] t:+353 83 4214179 Join the quest for Machine Intelligence at http://numenta.org Formerly of Adnet [email protected] http://www.adnet.ie
