Slashdot Mirror
Intel Mandates Universities Receiving Funds Not File Patents
sproketboy writes "Since January, four U.S. universities have agreed to host Intel Science and Technology Centers that will be funded at the rate of $2.5 million a year for five years. But wait, there's a catch: the company has made it a condition that in order to receive the millions, your university must open source any resulting software and inventions that come out of this research funding."
Intel NOT acting anticompetitive?
While I like this idea, doesn't it cause problems with first to file?
I just imagine a scenario where a university discovers something, doesn't file a patent, and megacorp comes along and patents it. With first to file, Megacorp gets the patent.
Maybe there's something I'm missing, but to me it would seem better that the university file the patent, but not be able to enforce it.
Surprisingly unproductive yet very expensive research coming from these universities.
Seven puppies were harmed during the making of this post.
...as long as Intel makes all their software and inventions open source as well.
Open-sourcing also prevents competitors, like AMD, from getting the benefit of said patents. With that, I assume Intel doesn't perceive any serious competition any time soon.
To-do List: Receive telemarketing call during a tornado warning. Check.
Universities make a lot of money from licensing patents. It's one of the only advantages they have over traditional businesses: universities are a pretty wide open R&D environment where ideas get explored that would never get money from VCs or other sources of funding.
I like bashing faceless mega corporations as much as the next guy, but this seems to be ... a benign act.
The usual condition to financing basic research would be that any resulting patents would go to the company that provided the financing, not that they'd be placed in the open source community. Someone has either had a stroke or needs their meds checked. Likely both.
It's just too bad it took a private company to make this requirement. My tax dollars should be enough already.
Intel simply doesn't want to pay for patents on ideas generates with its financial support. Here's the precedent they are trying to prevent from happening again: http://www.theinquirer.net/inquirer/news/1557536/intel-settles-university-wisconsin
Hope it sticks. Also should result in more of the money going to research instead of being used up on patent fees.
Don't get me wrong - I think this sounds like a fantastic idea by Intel. But is $2.5 mil a year (spit in the ocean for Intel) enough enticement to get a research university to forego any future revenues and other benefits of holding patents. Apparently it was at Carnegie Mellon . . . maybe i'm overestimating their annual research budgets . . . or perhaps overestimating the value of patents?
So why aren't we doing this with the national science foundation as well? Shouldn't research paid for "by the people" be available "to the people"?
Now if only the government would grow some balls and make the same condition for government research grants...
... someone else will. We have a first to file situation here. This is RIDICULOUSLY dumb on Intel's part. A nice sentiment, better executed by stating, "All fruits of this research must be patented by this foundation we've set up, which allows open, free licensing to anybody and everybody." Defensive patents are the only security you have; non-patent clauses just guarantee somebody other than your allies will patent! Ask Google, specifically whomever wrote the $12.1 billion check to acquire defensive patents from Motorola.
They're probably doing this because universities were selling the patents to companies that would then compete with Intel.
Just because it is prior art doesn't mean that you can't get a patent on it. As an example, slashdot did a story on LSI who patented the doubly-linked list, and that's just one example off the top of my head.
http://slashdot.org/story/06/11/23/1546218/LSI-Patents-the-Doubly-Linked-List
Sounds like Intel is basically crowdsourcing the universities for it's research. They can go back and apply for the patents themselves.
Join the Slashcott! Feb 10 thru Feb 17!
I wish the US government would take a similar approach -- any royalties a university receives should go back to the government in the proportion of the funding provided. If a university payed for research costs with 50% from the government then royalties from the patent should be split 50% with the government. If the government provided 100% funding, then 100% of the royalties should go beck to the government. In doing this, then the government is truly investing in research instead of just paying the bills.
I also would include corporations, too. If the government provides x% of funding for the creation of a new drug, then x% of the profits should come back to the government, since it is the taxpayer that footed the bill in the first place.
The other alternative is what Intel is proposing -- we will pay for the research, but everybody has the right to benefit from it.
I for one would like to say "thank you" to Intel. For once you've chosen not to go evil. Hopefully this "federal funding for state adoption of policy" style of coercion will catch on a bit more with respect to freeing up university research from patent encumbrance. Now if only a similar carrot could be invented and dangled in front of the rest of the corporate world. Tax breaks in proportion to the value of openly published not patented R&D...?
Two of my imaginary friends reproduced once
Nothing can be open with the incoming patent changes. The new first to file rule means I could file a patent on anything open, and they can't do jack.
I spent several hours learning the expected ramification on the law last weekend. You think it's borken now? ha!
You heard me!
The Kruger Dunning explains most post on
I wonder if this move is influenced from patent madness occurring in the mobile world, where Intel is currently behind on the curve and wants to catch up.
I went to eat some animal crackers and the box said, "Do not eat if seal is broken." I opened the box and sure enough..
The license of the code has absolutely no bearing on whether you patent it, in fact to patent something it must be open by definition since patents are public. Is there a less confused article about this somewhere?
Hopefully they will eventually do the same with all research that's financed with taxes.
The summary says they have to open source any resulting software and inventions.
The only way to do this is to patent it and THEN open source it....but this costs money. Since the patent goes to the first to file, even in the US now I believe, this is the only way to do it safely.
It's apparent from jobs.intel.com that Intel has a large appetite for employees who hold PhDs. Maybe they actually want more people to perform advanced research in semiconductors, computer science, and computer architecture, so they can hire those people? It certainly looks like they're willing to put their money where their appetite is. The "open source" provision is a no-brainier way of protecting themselves from having to pay royalties steaming from research they contribute to. At the very worst, it creates a barrier to entry (have to build their own lab) to other groups seeking to patent developments in those fields for exclusive use. I suppose it's possible that Intel is trying to limit the patentable research coming out of universities, but at least they're doing it with funding, and not political manipulations or lawsuits. Even in that case, research is funded and the fruits of that research become available to the public.
How does the new first to file law work into this? Sounds dangerous, do all that work and somebody else takes a monopoly on it.
If intel really wanted the crown for graphics performance they'd go and get it. Seriously.
But Nvidias closed source linux drivers are still better.
My preference for awesome open source drivers but if my choice is for awesome binary blobs or sucky open source drivers
ill take binary blobs for $100 alex.
He'll be missing out on the whole university experience.
The whole idea of attending a research University as oppose to a college is so you can take that cool cutting edge research, drop out, start a .dot com company, file a patent, get venture capital and cash out rich and buy a basketball team that'll never win an NBA championship.
It's decided that the advantages of patenting have started to flow less and less to companies like Intel, and more to patent trolls. Intel is not the bad guy here.
Therefore, it is in Intel's interest to fund research in areas it may want to commercialize, and simultaneously preclude patenting by insisting on open publication and no patenting.
In this scenario, the entity with the most money (i.e. somebody like Intel) wins if they have sufficient drive.
More realistically, they want to preclude the people funded by Intel to set up a startup on their own, one whose primary asset is the people and the patent estate. This way Intel can hire them as ordinary employees who are impoverished postdocs instead of having to first buy them out and then hire them.
Hurrah for Intel for open sourcing discoveries made with their money.
Too bad our taxpayer money is wasted on private patents. My ex-Kansas Senator Bob Dole ruined university basic research when he was outgoing Senate Majority Leader and pressed through a law that effectively gives away the patent rights to all university research done with government taxpayer money. Before this change, all government research was 'open source' and belonged to everyone. Now, the patents and all the monies that accrue belong to the universities. The effect has been to damp basic research which has low probability of profitable patents, and fund only applied science with the possibility of profit for the university from taxpayer dollars. It's the old 'something for nothing' where universities are corrupted by chasing after the money rather than chasing after the science. This law needs to be repealed. It's admirable that a private corporation sees the folly in granting patent rights to universities doing research with someone else's money.
It's not even that restrictive; it's only dictating what they can do with the results of the research that was funded by Intel., i.e. it wouldn't apply to a different research group down the hall funded by AMD.
Another storm in a pisspot. fuelled by hot air generated by amateur lawyers like AK Mark who think a law can be adequately described by a three word summary.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
Recently, we hosted a small-ish academic conference here at the university where I work, and I was one of the local organisers. Since we are in CS, potential sponsors are all the big name computing companies - Intel et. al.
Intel was very nice (it helped that we knew some researchers who work there, but still - everyone else was genuinely nice as well), and sponsored us. And interestingly, they have one non-negotiable condition for sponsoring academic conferences: the authors of presented papers *must* be allowed to put pre-prints of the papers (i.e. PDFs of the paper) on the web free of charge.
And that is a seriously cool think to ask for, because it prevents any sponsoring to go to the sort of conference which has papers disappear from general sight after publication, and only stores them behind a paywall of some sort. This is almost as important for research as the whole patents thing - *huge* kudos to Intel overall, someone has a major clue there!
A.
Open source is fine for intel. No patents is fine for intel. In terms of software, people still need hardware to run it on.
In terms of hardware to build for the software, intel has a huge fab advantage. Given an equal, level playing field, intel will come out in front in terms of fabrication.
However, if someone was to patent an invention they've funded, well... then they're at a disadvantage.
I run: Windows, OS X, Linux, FreeBSD. Just because you have a hammer, doesn't mean everything is a nail.
The US patent office has a little used mechanism to prevent this it is called Statutory Invention Registration. http://en.wikipedia.org/wiki/United_States_Statutory_Invention_Registration It effectively puts an invention into the public domain and provides a rock solid publication date. It's most common use is to force an invention into the public domain if an inventor feels that they can not get a meaningful patent on it to make sure what you suggest will not happen. It was probably created with academia and government agency in mind who's charter directs them to open source their work.
No words to thank Intel for this OpenSource on Patents Idea. Do the patent, but opensource it.
Leslie Satenstein Montreal Quebec Canada
Kudos to Intel for doing something right (even though it is also in their best interest).
I wish the U.S. government would specify exactly the same condition. Today, taxpayers fund research, but then resulting discoveries are locked behind patents and used to extract more money (sometimes, as with pharmaceuticals, at staggeringly hyperinflated prices) from the people who paid for them.
I'm aware of larrabee, however intel choose not to invest the money in the rasterising 3d acceleration business that would give them the top graphics card performance, simply because they do not see it as their core business, a core business that they like as it is thankyouverymuch.