Professor Gernot Heiser has released the second part of his response to my blog post and the originating Australian IT article, and you can find it here. On reading both parts, I have amended my original posts to include links to Gernot’s responses, because they address both of the key questions from the original posts: by identifying why a process like this would be followed and how national benefit is served by it. The most important point to realise is that NICTA still owns a great deal of associated work from this project – rather than the cloak-and-dagger fire sale that was alluded to in the newspaper piece.
You can find the whole discussion still on my blog as I feel that the ongoing and evolving discussion illustrates one of the key advantages of the new technological models that we have: the ability to exchange ideas, update our published text and construct more accurate representations of knowledge. I would like to thank Professor Heiser for his responses, especially as it would have been very easy to either ignore them or be very dismissive. Instead, he’s provided a great deal more information that has certainly informed me as to how these decisions are made and what they mean.
I hope that you have enjoyed reading this as well, and that you have found the whole exchange useful.
(Update: Gernot has responded to this blog and has found fault with both it and the original article. I have responded to him. You can read his article here and my comment below that, or just look at the comments on this post. Thanks again, Gernot, for the clarifications.)
(Update 2: Gernot has put a further discussion of the points raised both in the previous post and this one, which you can find here. In this one, Gernot clearly explains why approaches were taken the way they were, how NICTA is benefiting from the ongoing work (as are we) and further identifies that the original article didn’t manage to capture a lot of the detail of what had happened. My thanks again to Professor Heiser for taking the time to respond to this so thoroughly and so patiently!
As I noted on his blog post, the article took a tone that I responded to and, with additional information, I can clearly see both the benefit as expressed and the reasons behind such a decision. I have left this and the follow-up posts intact, with these updates, to show the evolution of the discussion. Please make sure that you read both Parts 1 and 2 of Gernot’s response if you’re going to read this!)
I stumbled across this article in the Australian (Australia’s national newspaper) inside their AustralianIT section. In it, it was announced that the Australian research body National ICT Australia had sold “groundbreaking technologies” to a US company, for virtualisation security software that was used on 1.6 billion mobile devices worldwide. The spun-off company that was sold, Open Kernel (OK) Labs, was sold in its entirety and with no provision of royalties back to NICTA. Now, before we go any further, let’s talk about NICTA. NICTA is Australia’s Information and Communication Technology Research Centre of Excellence, employing about 700 people and funded by the Australian government. One of NICTA’s primary goals is to apply the high-impact research it develops to create national benefit and wealth for Australia. Remember this, it’s important.
Now let’s go back to the sale of OK Labs and, if you read the article carefully, you’ll see that there is some serious non-discussion of how much money changed hands and whether the Australian government, or NICTA, would receive any payment back at all from the sale. The former CTO and co-founder, Professor Gernot Heiser, has stated that while he couldn’t reveal the cost of the technology, it was about 25 person years of development. He then goes on to point out that the original micro-kernel was open source and hence no royalties accrued, but they had received some payments for it. (In the past, I think?) The second kernel was developed after the original OK Labs had been spun off, with NICTA retaining a minority share, but that NICTA didn’t have any share or role in its development, hence that had transferred wholesale to the new US owner and, again, no royalties. The third micro-kernel was a research outcome from NICTA but hadn’t been deployed commercially – but this was moot as OK Labs had received an exclusive licence to use it, then purchased it outright and NICTA had obtained some equity without cash in OK Labs as a result.
Got that? Now let’s get to the profit sharing. Firstly, there has been no indication whether NICTA would receive any payment back from the sale to balance against the initial investment of taxpayer funds.
Any profit from the deal went to OK Labs investors initially and “anything left” is distributed to shareholders, which included NICTA. (Remember that they traded valuable and NICTA developed research for a greater stake of the pie, which will be valuable if “anything is left”.)
Let me add the final paragraph of the article here, because I can’t do it any better justice:
Professor Heiser said professional bankers were engaged to make the sale “and they didn’t do it for free”. He said the sale of OK Labs enhanced the reputation of Australian IT research.
I can only hope that this is the worst-written, hatchet-job of an article because, otherwise, I’m flabbergasted. It appears that a government funded body has managed to develop and deploy a technology while systematically ensuring that any actual benefit from IP developed on these monies was distributed to everyone else before a single dollar flowed back in to turn over the research cycle once more. The investors are making money, NICTA traded some valuable IP for magic beans and may not get any money, the bankers are making money and, somehow, in the scope of this operatically complex financial dance, where the private benefit is enormous, Professor Heiser then turns around and sticks a public benefit statement on the end. We’ve enhanced the reputation of Australian IT research.
How does this … situation enhance anyone’s opinion of our research? Who is going to know in a year’s time where that research came from and why will they ever have to know?
The standard shining light in Australian IT from public funding is the CSIRO WiFi patent which is scheduled to attract royalty payments of roughly $1 billion over the next 5-10 years. This is the model that everyone explains to you when you first get into University research and, if you have anything commercialisable, expect a knock on the door from your local research innovation group because everyone wants another CSIRO patent. A billion dollars buys a lot of research.
I don’t know how you can possible slice up 25 person years of time and trade that for a peppercorn in potentia, with federal funding and the dominant position of NICTA on the Australian academic research scene, and possibly call this enhancing the reputation of Australian IT Research. Why, yes, I’m sure investors will want to come back, get us to pay for it, trade it away, sell it to them with no hope of recouping our investment and then not require royalties. I have no doubt that this may bring more investors but in the same way that a wounded fish attracts sharks. The enhanced reputation of the fish is a fleeting experience and is hardly enjoyable.
If Professor Heiser is reading this, then I welcome any clarification that he can make and, in the Australian have miscast this, then I welcome and will publish any supported correction. I sincerely hope that this is merely a miscommunication because the alternative is really rather embarrassing for all concerned.