End Software Patents!

Here’s a fun video on the problem with patents brought to you by myself and the great folks at Idea Rocket Animation! Read Launching the Innovation Renaissance (amz) for more (TED book app).


Given your views, I have taken the opportunity to download your video, copy your format, instructions and code, and music, and substitute my own content, and distribute my new product.

Thank you for making a thousand flowers bloom.

Also, while your at it, next time could you add some links and make some format changes so it is easier to copy your materials. It takes way too long--5 minutes--to copy your work effort dude.

I think Alex is OK with weak protection like copyright--it's patents that's the bugaboo. Now I would favor a law saying that software patents must rise to a high level of innovation, novelty and unobviousness, before they would be granted. But I would not be against granting them in theory. After all, the difference between hardware, software and firmware has been blurred for decades now. Why just patent widgets?

"the difference between hardware, software and firmware has [always been] blurred"

That's the real problem. There's no way to allow patents for electronic circuits but not for "software." Microprocessors have used microcode (http://en.wikipedia.org/wiki/Microcode) for as long as I can remember to handle the most basic, hardware-esque functions.

Even what you would consider pure hardware is often "coded" in a hardware description language (e.g., http://en.wikipedia.org/wiki/VHDL) that looks just like software.

And even a physical instance of what one might consider pure software can be just an arrangement of transistors (e.g., ROM).

In other words, when it comes to what the invention really is, what is the difference between using a hard-coded full adder (which is a handful of transistors) or compiling the code "a = b + c;"

Mathematically, they are equivalent.

To this point, I have not seen a debate on what the proper definition of a software patent would be under this doctrine. I think that a serious discussion of "what is in, and what is out" would bring out the difficulties in the execution of this idea of dropping "software" patents.

And yet every day some bureaucrats effectively manages the idea of not dropping them.

Andrew: The patent office is following the direction of the courts. Congress has failed to provide clarity thus giving the courts wide latitude in overruling the patent officers decisions, and the patent office has chosen to not take a position and lobby Congress on specifics on a well known policy controversy.

Perhaps you believe Congress should eliminate the bureaucrats and pass laws granting patents like in the olden days because the Constitution gets it wrong in putting bureaucrats in the patent office and in the third branch, courts, in charge?

But merely rendering existing hardware in an fpga and firmware is hardly innovative and non-obvious, and clearly not worthy of a patent.

One patent application was for doing what businesses have been doing for centuries - combining multiple orders into one shipment and one bill of lading and one invoice. One-click shopping is hardly innovation worthy of patent either.

The court direction to the patent office to issue software patents was vague. The court should have ruled "Congress puts the authority in the patent office, and if the patent office is in error on software patents, then petition Congress to provide specific directions to the patent office on software patents." Before that court ruling, software innovations were rendered into rube goldberg machines which were much more obviously rendered in software instead of a machine, but Congress had ended the requirement that a working model be presented to the patent office.

However, we have had the experiment with software patents under the vague terms which have led to patents on the obvious, so Congress can make an informed revision of patent law. In fact, for Congress to fail to act would be a breach of duty of oversight.

"But merely rendering existing hardware in an fpga and firmware is hardly innovative and non-obvious, and clearly not worthy of a patent."

You miss the point. There is no way to draw a line between hardware and software.

If any hardware is "innovative, non-obvious, and worthy of a patent," then your ban on software patents would also cover almost any real-world implementation of that hardware.

Or conversely, any attempt to carve out a hardware exception in your ban would also carve out a software exception.

The problem is that there is no mathematical or inventive difference between hardware and software. How you implement any solution (or invention) is an implementation detail, and whether you call that implementation "hardware" or "software" is really a matter of taste. There is no coherent line between the two.


ALL general purpose computer out ther DO JUST ONE ALGORITHM. So limit software patents to any algorithms run on any device complete in turring sense (mathematica concept thus easily testable), but this one algorith. And you are good to go.

Yes all general purpose computers out there do "program" algorith. 1) Get instruction 2) Decode instruction 3) Execute instruction 4) Go to 1). That is it.

Well yes some software is like it (in that it is 2nd (or more) nested algorithm in it), but hardware is not closing gap to software :P And not the other way round.

*Firmware can be only TWO things: software run on general purpose hardware, or software run on special purpose hardware.

PS To execute software you need turring machine or equvalent (or special purpose machine), that machine is called hardware (it can be even set of LEGO ;) ).

Anyway Bill replay is SILLY. Copyright is not Patents and vice versa. To defend your right to this mathereial you would sue based on Copyright not Patents.

Yeah but you didn't. And you couldn't. Alex is now the IP guy. Had he copied some other IP guy for (all) his content (everything is remix as they say) he'd likely lose reputational capital.

It still trips me out that economists publish working papers sometimes 5 years before they publish them. You could never get away with this in some other sciences and I think about the reasons why.

I've been involved with expert testimony and I've found there's no 'reputation capital'--experts are whores. The Third Law of Litigation is: for every expert there is an equal and opposite expert. Bork became "anti-monopoly" when Microsoft paid his fee. Alex also has a price. These freebie positions economists take on blogs are just talking points, nothing serious.

Correction: when DOJ paid his fee--his "Antitrust Paradox" book was the bible for lax antitrust enforcement back in the days of Reagan's administration.

What's your point? Trust no one?

I would consider litigation a special case.

When dueling experts are rampant in determining the "non-obvious", then there is clearly a problem with the relevant law.

To change the law to require patents be obviously non-obvious is clearly a real heavy lift, and with lifting feathers is too hard....

The enumerated power of Congress to promote progress has become a power to promote monopoly profits by blocking progress, justified by arguing that freezing Mickey Mouse in 1950 is progress because Mickey in the 21st century would be destructive of progress, and to require progress in the physical world instead of defining progress in engineering to be writing about the past - I haven't seen a software patent without an obvious physical world analogue, and 99% of the real world analogues are centuries old.

Have not seen any software patent without analogous mathematical publication(s) or simple implementation of their thesis.

Michael Ward (also an econblogger) turned down an extremely lucrative contract with a certain company to be an expert witness when he disagreed with them. I bet it was a hard thing to do, but in the end he showed he had integrity. I wonder how much, if any, the integrity carries over to other cases.

You have just described in detail how the stunning growth of the internet happened.

That's a good point. And Alex has made it before. In these super-easy to copy highly dynamic environments where the innovations are flying around making sure you get paid would probably make you worse off, and not just in the aggregate.

Getting paid is not my problem. If you start a business and don't know how you are going to get paid, you fully deserve to go out of business. If you have something you really don't want others to have, don't distribute it. You can fully enjoy every aspect of it in the privacy of your home. The basis of copyright and patents isn't ownership; it is a granted monopoly to encourage development.

Anyone who says that intellectual property regimes are necessary for the development of new technology must have slept through the last couple of decades.

The intellectual property regime of the US is one of the millstones that it has tied around it's collective neck that will keep it in stagnation.

All I'm saying is Alex would be outraged like the guy on Network if everyone started saying the same things he's saying ;)

Nice work, Bill. Do you have a link so we can see what you've done? I love innovation!

My insider's perspective on how the sausage is made:

Mostly wrong, some kernels of truth

"mostly wrong"

Like where? Australian engineer in protest to change of Australian Patent Law, patentet round means of transportations. Yes. He patented Wheel, and got that Patent.

And lets think about it! Minimize time and cash spent on patent aplication and you MAXIMIZE potential benefits.

That who claim that your idea need to be conformat to the law are PLAIN wrong. Patent aplication JUST have to pass examination. Its easy. Add "over Internet". Done.
Then go to any victim who just sound to a) be relatively poor b) producting/servicing something similar to your patent.
Then demand LOW licensing fees. So LOW that litigation would cost them MORE.
Repeat at nausea.

Lots of cash. No risk of loosing patent if you choose your victims carefuly. Only risk is of time/money invested in patent aplication if it fails. But then you can resubbmit mildly modified version.

Heck you can even use already granted patent and fill in some blank box, and claim improvement over already existing innovation. You do not need to thinka bout invention then :P

Yes. Patent System where Patents do have presumption of validity and no CHEAP way of invaliditing them are DREAM COME TRUE for any leach who are willing to explioit it.

Instituting 6mth of after granting re-evaluation process for patents in recent US litigation is probably only significant change (ok there is also rule that loosing party pays the costs). US system is BAD.

Yes, filing patents is "super-easy" (but expensive if you pay for it yourself). Getting them accepted by the USPTO is not supereasy and will take more than 15 work-hours.

Examiners are not as clueless as this guy says. In my experience it's best if you get an examiner who puts all his eggs in one basket by pointing to a particular piece of prior art. Then if you can show your invention is different you pretty much have the patent.

It's not worth doing patents just to pad your CV. There are however people and corporations who have made successful businesses out of "legitimate" IP licensing (eg. IDCC).

I favor software patents on selfish grounds: they give American companies an edge over the Chinese, and increasingly provide lucrative jobs for engineers who are willing to work for the Chinese.

“super-easy” (but expensive if you pay for it yourself)

I'm SHOCKED that the government gets the record company (and grad school, and lots of other stuff) cross-subsidization financing model exactly backwards!

> but expensive if you pay for it yourself
Yes, but as mentioned in the article, the provisional patent can be useful, at it lasts a year and costs only $110. That's only relevant to independent "inventors"; for employed engineers, they pay nothing out of pocket, of course.

> It’s not worth doing patents just to pad your CV.
As described, you do these patents because it is part of your work and indeed is often considered to be above the call of duty. Your employer ( a committee or manager) approves all patents before filing. And I suppose that patent filings (we are talking about real, official patent filings) are a significant asset on a CV, though it depends on who is reading it.

> Examiners are not as clueless as this guy says

I said what I did from experience. YMMV.

> Getting them accepted by the USPTO is not supereasy and will take
> more than 15 work- hours.

In my experience, in almost all cases, the process of getting the patentss accepted by the USPTO (after filing) was handled by lawyers almost completely without my help, except perhaps answering an email query. YMMV.

Certainly, it may have taken more than 15 hours of a lawyer, but this essay was written for the engineer in the trenches.

In UK patent aplication whole fee is 3 000 of their currency (or was it euro?).

So not all countries put brakes to small innovators.

I think that small entrance fee and progressive reneval fees are best.

If (When) patent works.

Software Patents do not. Innovation happened EVERYWHERE even in US pre-software patents at amazing speed, and explosion of technical universities with CS departments guarantee that such peace will be sustained. Software is also HIGHLY COMPETETIVE market. Innovation give wage over competition. With or withou patents innovation will happen.
Spread of information and knowlege is SO easy in software industry. So again Software Patent as means of DISCLOSURE are not needed. Hecke they thereaten it. Software industry/community developed SOUND and ECONOMICALY VIABLE means to share and disseminate that knowlege withou patents (look at FLOSS). Patents by their virtue of being monopolies actually can SLOW DOWN adoption of that knowlege. (Software Patents are granted for TOO long. Its not a joke that years == eons in software space. Look at mobile. Look at PCs. Look at your toasters :P )

So two major reasons of supporting Patents DO NOT stand against reality of software industry/community.

There is also third. "Innovation" as "Property". But then again. Software as "property" is protected by Copyright. That is enough.

Is the gene sequencing that Genentech et al are doing considered software or medical?

Is it the gene sequencing or the gene sequences?

There is a lot of hardware and technique in the processes of sequencing genes quickly. It really is an amazing field with a lot of interesting innovation.

There are significant software innovations, including algorithms that detect errors in the sequencing techniques that make the results much, much more accurate.

I do agree that this is a field that would be impacted. It is a highly competitive market, and if you throw out all software patents, some of the money spent researching better algorithms might be refocused on better hardware platforms.

The patent system was originally to provide protection for individual inventors and to encourage innovation by disseminating information and ideas. It has been perverted so that it does neither. Those with the most lawyers ($) win. There have been inventors (i.e. Baxandall) that have forgone patenting their invention and put them in the public domain to keep big companies (i.e. RCA) from controlling them.

As for the drug companies, many of the drugs that are patented are designed to replace those whose patents are expiring. The intent is to maintain the revenue stream. They are only comparably "safe" and "effective" as determined by the approval process of the FDA and actually many times are less safe (i.e. VIOXX). If the drug companies do not want to invest money in drug development, they should find another business. They make a fair amount of money.

Just because something gets a patent does not mean that it is innovative, useful or beneficial. It just means that it was different than the patents that preceeded it.

"The patent system was originally to provide protection for individual inventors and to encourage innovation by disseminating information and ideas. It has been perverted so that it does neither."

This statement is commonly made. My experience working with the research and development areas of companies leads me to believe otherwise.

Major corporations do keep track of what is disclosed in patent applications. It does have an impact on the development process. I believe that it does disseminate information regarding innovation direction. There is often times a lot of noise in the applications process, but there are significantly large departments that spend a significant amount of time looking at these filings.

I know it's not what you mean, but certainly the individual inventor is not protected. It's pretty hard to file a patent without corporate backing.

I'm guessing you don't work in the software industry. I've worked at a number of big names, and not a single one of them has ever tracked software patents. Not only does doing so subject you to increased liability (it's willful at that point, meaning triple damages, and if your company has a history of looking at new patents, it's much easier for whoever is suing you to prove), software patents are unusably vague. Almost none of them that aren't trivially obvious include enough information to implement them. This is done on purpose, since the less actual detail you put in the patent, the more different implementations of one idea you can cover.

All of which means that, for software patents, looking at them only has downside potential. Well, except for lawyer-only patent troll companies.

It is not even a sure thing that we are better off by having drug patents.

This terribly misinformed. Big pharma isn't innovative. Drug companies spend much more on marketing/sales than they do on R&D. They were given special patent privileges extending to them protection not only in years, but also by just slightly changing the chemistry. (ref. 1984 Hatch-Waxman Act) So instead of the sick getting cheaper off-patent drugs, these companies continue to rake in profits by only slightly changing the formula. Further they are not required to show any additional efficacy. They're worse than the "bad guys" in this story
(worse because health is a moral issue). Even the US GAO said the drug industry has become "stagnant" years ago! Big pharam is the main cause of the rising costs of US healthcare. Start by reading this book, "The Truth
About the Drug Companies" and catch up on the history.

Why does everyone mindlessly accept the "necessity" of patents for drugs. How can you deny the existence of penicillin or the polio vaccine?

Comments for this post are closed