Nouiere Järvinen’s Post

𝗪𝗵𝗮𝘁 𝗔𝗿𝗲 𝘁𝗵𝗲 𝗟𝗶𝗺𝗶𝘁𝗮𝘁𝗶𝗼𝗻𝘀 𝗼𝗳 AI 𝗠𝗼𝗱𝗲𝗹𝘀 𝗳𝗼𝗿 𝗣𝗮𝘁𝗲𝗻𝘁 𝗩𝗮𝗹𝗶𝗱𝗶𝘁𝘆/𝗜𝗻𝘃𝗮𝗹𝗶𝗱𝗶𝘁𝘆 𝗦𝗲𝗮𝗿𝗰𝗵𝗲𝘀—𝗮𝗻𝗱 𝗛𝗼𝘄 𝗖𝗮𝗻 𝗪𝗲 𝗦𝗼𝗹𝘃𝗲 𝗧𝗵𝗲𝗺? As many of us in the patent field know, there are several limitations to current models used for patent validity and invalidity searches. Today, I’d like to discuss one of these challenges and share how our technology addresses them to make searches more efficient and less time-consuming. 1. 𝘽𝙞𝙖𝙨 𝙞𝙣 𝙏𝙧𝙖𝙞𝙣𝙞𝙣𝙜 𝘿𝙖𝙩𝙖 𝙃𝙤𝙬 𝙬𝙚 𝙨𝙤𝙡𝙫𝙚 𝙩𝙝𝙞𝙨: We’ve trained our model using patents from a wide range of jurisdictions. With over 1,000,000 datasets included, we ensure coverage of both historical and recent patents to account for evolving patentability standards and technological trends. We’ve also diversified the dataset across various technical fields and incorporated court rulings on patent litigation to better understand validity in real-world scenarios. Additionally, the model is regularly tested and refined by patent examiners, legal experts, and technical specialists to maintain its robustness. 𝙒𝙝𝙖𝙩 𝙖𝙙𝙙𝙞𝙩𝙞𝙤𝙣𝙖𝙡 𝙨𝙤𝙡𝙪𝙩𝙞𝙤𝙣𝙨 𝙤𝙧 𝙨𝙩𝙧𝙖𝙩𝙚𝙜𝙞𝙚𝙨 𝙬𝙤𝙪𝙡𝙙 𝙮𝙤𝙪 𝙨𝙪𝙜𝙜𝙚𝙨𝙩 𝙛𝙤𝙧 𝙤𝙫𝙚𝙧𝙘𝙤𝙢𝙞𝙣𝙜 𝙩𝙝𝙚𝙨𝙚 𝙘𝙝𝙖𝙡𝙡𝙚𝙣𝙜𝙚𝙨? #patents #validitysearches #randd #intellectualproperty #litigence

The story of Carrie Hafeman is the pinnacle of sloppy, blatant corruption in the patent system. Why is no one focused on the Elephants in the room in Carrie's case? Does anyone know how hard it is to get an attorney these days? HARD! Carrie finally, in the 19th year of her patents, got legal support to stop her infringer, LG. Imagine going a decade after your infringers steal everything you worked for, your company destroyed, all your employees gone and finally getting your day. Carrie has spent over $300,000 in maintenance fees over the course of the last 20 years. This money goes to the UPSTO. Her money pays USPTO Director Kathi Vidals salary. Her money paid for her patents, which tell her that she has the right to stop her infringers. Carrie finally got her fight. Carrie Vs. LG. Enter the elephant(s), the bullies. Google, Microsoft and Kathi Vidal. Here are the questions we should all be asking: Why did Google and Microsoft enter Carrie's fight and request to challenge Carrie’s patents in their final year? Why did Kathi Vidal, the proclaimed supporter of women/children and minorities allow Google and Microsoft to send Carrie's patents to the PTAB to be invalidated? Why is NO ONE screaming off the rooftops on how Google and Microsoft broke the rules in doing this? The judges in the case even said they did. There is no precedent they say, so just let Google/Microsoft get away with it. Why is Kathi meddling in Carries case? What should we expect her reply brief to Carrie's case say? EVERY SINGLE PERSON in this industry should keep your eyes on this case. December 16th Vidals brief is due. US Inventor

The USPTO Director Kathi Vidal is often posting about the USPTO using AI. Perhaps the USPTO should recommend that PTAB Judges use AI to help them make decisions. They can ask AI a question about a patent's validity, and then be able to more quickly see whether a prior art patent application or combination of patent publications are both missing a critical claimed feature. If both patent applications in a combination are EACH missing the same critical claimed feature then you CAN NOT combine them to say a patent should be cancelled. Not only does AI suggest that my IPRs should not have been instituted in the first place, but it also shows that the PTAB Judges who determined that a Jenne/Cohen combination cancelled my patents was a clearly flawed decision! So are there really all these bad patents? Are patent examiners who issue patents really wrong up to 80% of the time? Or are there bad PTAB judges decisions? For any inventors experiencing PTAB, you can use an AI program like this. This video is a little too long, but it shows the USPTO should have never instituted my IPRs, and that the PTAB Judges should clearly NOT have cancelled my patents. #PTAB, #inventor, #patents, #IPRs

Great points, Carrie Hafeman. Time for PTAB APJs should revisit and apply MPEP 2143 when deciding whether two or more references in combination would render obvious a claim in an issued patent. A. Combining Prior Art Elements According to Known Methods To Yield Predictable Results To reject a claim based on this rationale, Office personnel must resolve the Graham factual inquiries. Then, Office personnel must articulate the following: (1) a finding that the prior art included each element claimed, although not necessarily in a single prior art reference, with the only difference between the claimed invention and the prior art being the lack of actual combination of the elements in a single prior art reference; (2) a finding that one of ordinary skill in the art could have combined the elements as claimed by known methods, and that in combination, each element merely performs the same function as it does separately; (3) a finding that one of ordinary skill in the art would have recognized that the results of the combination were predictable; and (4) whatever additional findings based on the Graham factual inquiries may be necessary, in view of the facts of the case under consideration, to explain a conclusion of obviousness. USPTO, Kathi Vidal, Under Secretary of Commerce for Intellectual Property, #PTAB, #APJ, #patent, #validity, #AI, #IP, #MPEP

#AI is definitely the topic of the day in the #patent field. I attended the Intellectual Property Owners Association (IPO) annual conference in Chicago this week, where most of the sessions seemed to include some artificial intelligence take on patent law. I thought it made sense to repost the 2024 AI Patent 100 List for everyone. Enjoy! https://lnkd.in/etPN5gvE

Patenting AI inventions is tough at the best of times. It's even tougher when you consider the different approaches taken by different patent offices... read more below!

With AI proving to be a particularly stubborn challenge for patent systems, Oliver Rigg, Lisa Williams and Gemma Robin explore the approaches of two of the world’s largest patent offices – the European Patent Office and the USPTO – in assessing the patentability of AI inventions. Read the full article here: https://lnkd.in/e4YdiGpS #ImagineBetterAI #AIInnovation #PatentingAI #AIExperts

Essential reading here for anyone involved in patenting AI inventions

On the February 13, 2024, the USPTO released guidance for the examination of AI-assisted inventions. In that guidance, the USPTO clarified that a human must make a significant contribution to each claim for the invention to be patented. On April 10, 2024, the USPTO issued a second round of guidance, this time directed to patent attorneys who use AI tools to prepare patent applications and other filings for the patent office. The April 10 guidance reminds patent attorneys to be especially vigilant in maintaining confidentiality of client data, because many AI tools retain the prompts and information entered into the system and use that information to train the system. The guidance also reminds patent attorneys of their duty of candor towards the patent office, which includes a requirement to disclose all information the attorney knows could be material to whether an invention is patentable. Attorneys should carefully review any AI-generated information to ensure that submissions to the patent office are complete, relevant, and accurate. Simply put, AI tools are not yet advanced enough to be used for patent writing or prosecution without meticulous oversight from a licensed patent attorney.  #innovation #ai #patentlaw

My Foley colleague, Matthew Horton, will be participating in the panel "AI and the U.S. Patent Landscape" at the Patent Quality Week 2024. This panel will discuss critical issues that U.S. policymakers and regulators must carefully consider when defining what emerging AI means for the U.S. patent system. Click below to learn more and to register. #Patent #ArtificialIntelligence #IntellectualProperty