Prior Artist
Alexander Reben is automatically generating all possible prior art. Which will probably take a while.
Instead, why not stop examining patents altogether?
(via Boing Boing)
Paresh Dave in Wired writes about TDCommons.org, a Google funded but bepress operated site. The idea is to publish technical disclosures as prior art that might invalidate future patents. It's an interesting overview of the subject, including a USPTO attempt to do the same thing (I covered this here) and a commercial competitor, IP.Com. Apparently USPTO is looking for help with this problem:
"Google is hoping TDCommons has a chance to be embraced as Kathi Vidal, a tech patent attorney who was sworn in as director of the USPTO almost two years ago, settles into her role. Deciding that generative AI programs can’t be patent holders has been a higher priority, she says, but creating a better search tool for prior art is an issue she’s discussed with a lot of organizations. Vidal says she’s open to the USPTO administering and funding its own prior art repository, offering up her email, [email protected], for feedback on how to do so."
I'm not super-convinced that she's providing her actual email address, but when I have a few minutes I might suggest my own plan - issue all patent applications and shift the examination to the start of any litigation or enforcement attempt.
(Published to the Fediverse as: TDCommons and the Future of Patent Law #politics #patents #uspto Prior art attempts like TDCommons, IP.com and even the USPTO's failed scheme should be replaced with a fundamental overhaul of the patent system. )
Alexander Reben is automatically generating all possible prior art. Which will probably take a while.
Instead, why not stop examining patents altogether?
(via Boing Boing)
I've supported the SHIELD Act before, which would force patent trolls to pay legal bills for unsuccessful shakedown attempts, but a TechCrunch article today made me think this through some more.
SHIELD would be a serious deterrent for trolls who have their eye on large companies with the means to defend themselves. But trolls eat startups first and a startup is often unable to fight through the courts and get to the point where SHIELD would help. If the troll is after something like $1,000 from every company using a scanner then not many businesses are going to risk going to court. And if the troll isn't interested in any reasonable settlement then the legal fees and management distraction can kill you.
SHIELD is well intentioned and would certainly help. But we need to stop examining patents before issuing them and do the job properly for the few that ever get used in anger.
(Published to the Fediverse as: Bringing a SHIELD to a conker fight #politics #patents #uspto #shield #trolls SHIELD might help patent reform but what we really need to do is stop examining patents altogether. )
In 1790 Thomas Jefferson became the first patent examiner and issued a grand total of three patents. 220 years later the US Patent Office has a backlog of over 1.2 million applications. Internet and software industries are created and destroyed in the time it takes the average patent to issue.
Here's how a startup can easily turn to the dark side:
Software patents are especially frustrating as it's the idea — usually the easiest part of the business — that gets patented. Because patent law doesn't require actual implementation, let alone success, it's as if you could patent “cancer drug” and then sue pharmaceutical companies each time a new treatment rolled out.
It's tempting to call for an outright ban on software patents, as Vivek Wadhwa did recently. This doesn't work because so much technology is software based and because somewhere in that 1.2 million backlog there probably are a few genuinely novel ideas.
The first part of the fix is to stop examining the patents at all.
This may sound crazy, but think about copyright. You don't need to pay a bunch of lawyers to represent you before the copyright office and prove that this really is the first time a particular novel has been written, or that you were in fact holding the camera when you snapped that photo. Copyright is automatic.
A patentable invention isn't a specific work of art and so some registration system is required. My suggestion is that you upload a PDF and pay a registration fee of around $1,000. It's small enough not to discourage startups and large enough to prevent abuse. USPTO would timestamp the PDF, store it for eighteen months and then publish it.
The obvious flaw would seem to be opening the floodgates to even more patent trolls. That's where the second part of the fix comes in — shift the burden of proof of validity to the company that owns the patent. You can still sue, but the first step is a rigorous and expensive exam process.
This system makes it easy to obtain a patent but changes the nature of the patent to an insurance policy in case your idea really is as clever as you think it is. It also makes it much harder to use a patent offensively. Resources are more efficiently used to evaluate the novelty of a patent when it is enforced, rather than to do the impossible task of evaluating millions of ideas that never will be.
Full disclosure: I'm co-inventor of several software patents that I hope will never be used against you. I also recently lost a product (Cucku Backup) as a result of settling a patent infringement lawsuit and I hope that never happens to you either!
(Published to the Fediverse as: How to fix software patents #politics #patents #jefferson #uspto #trolls #copyright A fix for the patent system - scrap up front examination and remove the assumption that any patent is valid. Easier to get a patent, harder to use if offensively. )
Export Google Fit Daily Steps, Weight and Distance to a Google Sheet
Accessing Printer Press ESC to cancel
Monitor page index status with Google Sheets, Apps Script and the Google Search Console API
Download a Sharepoint File with GraphServiceClient (Microsoft Graph API)
Upgrading from word2vec to OpenAI
Scanning from the ADF using WIA in C#
International Date Line Longitude, Latitude Coordinates
Automate Google PageSpeed Insights and Core Web Vitals Logging with Apps Script
Enable GZIP compression for Amazon S3 hosted website in CloudFront
User scoped custom dimensions in Google Analytics 4 using gtag