Saturday, May 28, 2022

Your face is out there! Apparently as easy to find as your street address. How is this better or worse than Clearview?

https://www.pcmag.com/news/this-facial-recognition-site-is-creeping-everyone-out

This Facial Recognition Site Is Creeping Everyone Out

A facial recognition tool called PimEyes has recently gone from unknown to infamous.

PimEyes makes it easy to find pictures of people that are strewn across the internet. That isn't necessarily surprising—reverse image searches have been a thing for years—but it turns out PimEyes is astoundingly good at identifying people with naught but a single photograph.

The New York Times reports that it found years-old pictures even if the sample image featured people wearing sunglasses or face masks. Other factors such as different facial hair, new hair styles, or the passage of time didn't seem to make all that much of a difference either.

PimEyes has responded to the resulting scrutiny with a blog post in which it says:

PimEyes just provides a tool, and the user is obliged to use the tool with responsibility. Everyone can buy a hammer, and everyone can either craft with this tool, or kill. It is impossible to check if certain individuals use every tool in their possession in accordance with the law and it is unwise to oblige toolmakers to ensure that their product will be used as they are intended to be used. If PimEyes starts to verify every user and compare it to the data searched by the user, this might turn the company into a monster that stores not only the personal, biometric data of each subscriber, but the materials that in the most of cases they would like to leave confidential. Therefore, it is crucially important for company to have opened communication policy, cooperate with media and various information platforms to encourage ethical usage of service and internet.





Students are just statistics…

https://www.pogowasright.org/schools-urge-end-to-forced-disclosure-of-private-student-information-state-using-data-to-track-vaccinations/

Schools urge end to forced disclosure of private student information; state using data to track vaccinations

Sherrie Peif reports:

More than 20 charter schools across the state recently presented a petition to the Colorado Department of Public Health and Environment (CDPHE) requesting that it stop collecting and using private information of school children to publicize youth vaccination rates on its website.
Diana Herrero, deputy director of the division of disease control and public health response — a division of CDPHE — confirmed in an email obtained by Complete Colorado that it has been collecting identifying information about Colorado school children without their parents’ knowledge to combine with its COVID-19 vaccination records and create a public database that showcases the vaccination rates at every school in Colorado.

Read more at Complete Colorado.





Also good information if hacking is your side hustle…

https://www.makeuseof.com/how-to-start-career-in-ethical-hacking/

How to Make a Career Out of Ethical Hacking

To catch a criminal, you must think like one by understanding the tactics and patterns of the criminal.

Ethical hackers are the opposite of cybercriminals. As an ethical hacker, you'll be working around the clock to nullify the efforts of the bad buys trying to breach computer networks.

So, what skills do you need to become an ethical hacker? Where can you learn ethical hacking? And what are the entry-level jobs for ethical hacking? You are about to find out.





My AI suggests Ireland should have found an AI for that post.

https://analyticsindiamag.com/ireland-gets-its-first-ai-ambassador-will-other-countries-follow-suit/

Ireland gets its first AI ambassador. Will other countries follow suit?

Ireland has appointed Dr Patricia Scanlon as its first AI Ambassador to facilitate the Government’s AI adoption strategy launched last year.

She will work on demystifying AI and promoting the positive impacts it can have in areas such as transport, agriculture, health and education.





Perspective.

https://www.theatlantic.com/ideas/archive/2022/05/ukraine-russia-moskva-military-marine-corps/629930/

A Whole Age of Warfare Sank With the Moskva

On April 14, 2022, the Ukrainians sank the Russian cruiser Moskva with a pair of Neptune anti-ship missiles. And that success posed an urgent question to the world’s major militaries: Has another age of warfare just begun? After 20 years spent fighting the post-9/11 wars, the United States military’s attention is again focused on a peer-level adversary. The Pentagon hasn’t been thinking this way since the Cold War, and it is attempting a profound transformation. Today, fierce debate attends this transformation, and nowhere more acutely than in the Marine Corps.

In March 2020, the Marine commandant, General David Berger, published “Force Design 2030.” This controversial paper announced a significant restructuring based on the belief that “the Marine Corps is not organized, trained, equipped or postured to meet the demands of the rapidly evolving future operating environment.” That “future operating environment” is an imagined war with China in the South Pacific—but in many ways, that hypothetical conflict resembles the real war in Ukraine.

The military we have—an army built around tanks, a navy built around ships, and an air force built around planes, all of which are technologically advanced and astronomically expensive—is platform-centric. So far, in Ukraine, the signature land weapon hasn’t been a tank but an anti-tank missile: the Javelin. The signature air weapon hasn’t been an aircraft, but an anti-air missile: the Stinger. And as the sinking of the Moskva showed, the signature maritime weapon hasn’t been a ship but an anti-ship missile: the Neptune.

Berger believes a new age of war is upon us. In “Force Design 2030,” he puts the following sentence in bold: “We must acknowledge the impacts of proliferated precision long-range fires, mines, and other smart weapons, and seek innovative ways to overcome these threat capabilities.”





Perspective.

https://thenextweb.com/news/halseys-record-label-wont-release-a-new-song-until-it-goes-viral-on-tiktok-is-this-the-future-of-the-music-industry

Halsey’s record label won’t release a new song until it goes viral on TikTok. Is this the future of the music industry?

On Sunday, popular American singer songwriter Halsey shared a video on TikTok with tinny music in the background, the on-screen text reading:

Basically I have a song that I love that I wanna release ASAP but my record label won’t let me. I’ve been in this industry for 8 years and I’ve sold over 165 million records. And my record company is saying that I can’t release it unless they can fake a viral moment on TikTok. Everything is marketing. And they are doing this to basically every artist these days. I just wanna release music, man. And I deserve better tbh. I’m tired.



Friday, May 27, 2022

I dismissed this initially, thinking that no one would think the metaverse was real. Then I thought of all the people who think Trump’s version of a stolen election is real…

https://thenextweb.com/news/uae-ai-minister-wants-murder-metaverse-real-crime

The UAE’s AI minister wants ‘murder’ in the metaverse to be a real crime

Omar Sultan Al Olama, the United Arab Emirates minister of artificial intelligence, yesterday told an audience at the World Economic forum in Davos that it’s his belief that people who commit “serious crimes” in the metaverse should be punished with real-world criminal consequences.

Per an article by CNBC’s Sam Shead, the minister views this as a necessary measure to protect people’s mental health:

If I send you a text on WhatsApp, it’s text right? It might terrorize you but to a certain degree it will not create the memories that you will have PTSD (post-traumatic stress disorder) from it.
But if I come into the metaverse and it’s a realistic world that we’re talking about in the future and I actually murder you, and you see it … it actually takes you to a certain extreme where you need to enforce aggressively across the world because everyone agrees that certain things are unacceptable.





Unintended consequence, but something to keep in mind going forward!

https://www.cpomagazine.com/cyber-security/nsa-sanctions-on-russia-having-a-positive-effect-on-ransomware-attacks-attempts-down-due-to-difficulty-collecting-ransom-payments/

NSA: Sanctions on Russia Having a Positive Effect on Ransomware Attacks, Attempts Down Due to Difficulty Collecting Ransom Payments





An automated navy?

https://seapowermagazine.org/us-5th-fleet-commander-explains-role-of-unmanned-ai-in-middle-east/

US 5th Fleet Commander Explains Role of Unmanned, AI in Middle East

The commander for U.S. naval forces in the Middle East discussed the role of unmanned systems and artificial intelligence in naval operations at an international security conference in the United Kingdom, May 24, NAVCENT Public Affairs said May 25.

Vice Adm. Brad Cooper, commander of U.S. Naval Forces Central Command, U.S. 5th Fleet and Combined Maritime Forces, spoke to an audience of nearly 800 international defense and industry leaders during the Combined Naval Event at the Farnborough International Exhibition and Conference Centre.

We are on a path to build the world’s first international unmanned surface vessel fleet,” Cooper said. “Three weeks ago, we surpassed 10,000 total sailing hours for unmanned surface vessels throughout the region. Additionally, two vessels each exceeded 100 consecutive operating days at sea.”



Thursday, May 26, 2022

War makes all logistics a target. How long would you continue fighting if you could not feed your children, let alone your army.

https://www.csoonline.com/article/3661434/remote-bricking-of-ukrainian-tractors-raises-agriculture-security-concerns.html#tk.rss_all

Remote bricking of Ukrainian tractors raises agriculture security concerns

Modern agriculture depends on internet-connected machinery that is centrally controlled and collects and analyzes massive amounts of data, making it an inviting target for threat actors.

Against the backdrop of horrific reports from Russia's Ukraine invasion, an encouraging story emerged earlier this month when unidentified Ukrainians remotely disabled tractors worth $5 million that Russian soldiers in the occupied city of Melitopol stole from Agrotek-Invest, an authorized John Deere dealer. The soldiers stole 27 pieces of farm machinery and shipped them primarily to Chechnya, 700 miles away, only to discover they had been rendered inoperable due to a "kill switch."

The dealership tracked the machinery using the tractors' embedded GPS technology. Although the equipment was reportedly languishing at a farm near Grozny on May 1, one source said the Russians had found consultants who would try to bypass the digital protection that bricked the machines.

Some observers fear that malicious actors could exploit the same technology Deere and other manufacturers use to update and monitor farm equipment. If successfully accomplished on a large-enough scale, a cyberattack could disrupt significant portions of what has become critical agricultural infrastructure.





Imagine a school district that “monitored” their students (and faculty, parents, and others?) perhaps once a week. Did they have a flawless AI algorithm? What did they promise parents this system would do?

https://www.dailymail.co.uk/news/article-10853241/Uvalde-school-district-monitoring-students-social-media-pages-threats-against-staff-site.html

REVEALED: Uvalde school district was part of AI program that rooted out potential mass killers and monitored social media for threats and potential shooters

Texas school officials had been monitoring students' social media prior to the deadly shooting in Uvalde Tuesday, it has been revealed - but still failed to pick up on concerning posts from the teenage gunman in the days leading up to the tragedy.

Now, Uvalde School officials say they had been monitoring its students' social media pages using an advanced AI-based service called Social Sentinel, designed to recognize signals of potential harm found in digital conversations.

The district revealed Monday it had been using the platform 'to monitor all social media with a connection to Uvalde as a measure to identify any possible threats that might be made against students and or staff within the school district.'

However, in this particular instance, the technology fell short - failing to spot Ramos' objectively concerning posts and notify district officials.

It is not immediately clear why the technology failed to flag Ramos' posts. DailyMail.com reached out to Social Sentinel and Uvalde district staffers for comment on the software's apparent failure Wednesday morning, but did not immediately hear back.





If consent is what you want, consent is what we’ll give you. (Will they copy images from the ‘consent’ database to the bigger database?)

https://www.businesswire.com/news/home/20220525005378/en/Clearview-AI-Launches-Clearview-Consent-Company%E2%80%99s-First-Consent-Based-Product-for-Commercial-Uses

Clearview AI Launches Clearview Consent; Company’s First Consent Based Product for Commercial Uses

For the first time, Clearview AI’s industry leading FRT algorithm will be available for a multitude of identity verification purposes across the commercial marketplace. Product will be offered separate and apart from Clearview AI’s database of facial images

Clearview Consent will offer Clearview AI’s state-of-the-art, world renowned, highly accurate and bias-free facial recognition technology (FRT) to companies for use in consent-based enterprise workflows. It will be sold as a licensed product, separate and apart from the company’s database of 20+ billion facial images, the largest such database in the world. This means private companies will be able to use Clearview AI’s industry leading FRT platform, powered by the most accurate and bias-free algorithm in the U.S. and Western World.



(Related) Nothing if not opportunistic.

https://gizmodo.com/clearview-ai-facial-recognition-privacy-1848975528

Clearview AI Says It's Bringing Facial Recognition to Schools

The company revealed it's working with a U.S. company selling visitor management systems to schools within hours of a horrific school shooting.

In a press release Wednesday, the company outlined a path toward an apparent one-to-one face match verification method that could be used in schools, banks, and other private firms as part of its new “Clearview Consent product. Clearview says it seeks to sell its facial recognition tool to enterprise companies decoupled from its massive database of faces. Theoretically, that means private companies could use Clearview’s system as a 1:1 identity verification tool before creating an online account, check-in a passenger at an airport, or protecting against financial fraud.

Clearview’s pivot towards a database-free version of its tech comes partly out of necessity. The walls have started to close in around the country in recent months, with new restrictions and government opposition threatening to upend their core product offerings.





Today the intelligence services, tomorrow the world...

https://www.cpomagazine.com/data-privacy/swiss-government-proposes-backdoor-in-its-banking-secrecy-laws-for-spy-agency-pressure-applied-by-un-over-human-rights-concerns/

Swiss Government Proposes Backdoor in Its Banking Secrecy Laws for Spy Agency; Pressure Applied by UN Over Human Rights Concerns

The private Swiss bank account has been one of the country’s distinguishing features for nearly 100 years. But these banking secrecy laws began to see significant erosion in 2018, and are facing another substantial blow as the country’s government is proposing to allow its spy agency to monitor financial transactions.

This comes amidst debate about doing away with the country’s famous banking secrecy laws entirely, as the United Nations applies pressure in the wake of the Credit Suisse scandal.



Wednesday, May 25, 2022

I disagree. There are too many organizations who want this technology for it to disappear.

https://www.technologyreview.com/2022/05/24/1052653/clearview-ai-data-privacy-uk/

The walls are closing in on Clearview AI

… “The company not only enables identification of those people, but effectively monitors their behaviour [New to me. Have I missed something? Bob] and offers it as a commercial service. That is unacceptable,” said John Edwards, the UK’s information commissioner, in a statement.

Europe is working on an AI law that could ban the use of “real-time” remote biometric identification systems, such as facial recognition, in public places. The current drafting of the text restricts the use of facial recognition by law enforcement unless it is to fight serious crimes, such as terrorism or kidnappings.



(Related) Is a face match “identification” or must other steps be taken? (A face is not a name.)

https://www.insideprivacy.com/data-privacy/general-court-of-the-eu-finds-that-individual-was-unable-to-prove-that-information-published-online-constitutes-personal-data/

General Court of the EU Finds that Individual was Unable to Prove that Information Published Online Constitutes “Personal Data”

On May 4, 2022, the General Court of the EU handed down a decision that helps clarify the standard of proof required to demonstrate that information that does not identify someone by name constitutes “personal data” under EU data protection law. The court also clarifies that the burden of proof falls on the entity alleging that the information is personal data.



(Related) If not Clearview, someone...

https://www.databreaches.net/israeli-ministry-illegally-shared-biometric-images-of-millions-with-unknown-agency/

Israeli Ministry Illegally Shared Biometric Images of Millions With Unknown Agency

Josh Breiner and Bar Peleg report:

The Population and Immigration Authority illegally shared in the past seven years the facial images of millions of Israelis with an unnamed government agency.

The actions of the Interior Ministry division were disclosed in an official report published last week by Roy Friedman, the head of the Israel National Cyber Directorate’s Identity and Biometric Applications Unit.

Read more at Haaretz.



(Related) ...and not just facial recognition.

https://www.bespacific.com/how-dhs-massive-biometrics-database-will-supercharge-surveillance-and-threaten-rights/

How DHS Massive Biometrics Database Will Supercharge Surveillance and Threaten Rights

Immigration Defense Project – HART Attack: “The US Department of Homeland Security (DHS) is building a $6.158 billion-dollar, next-wave biometric database that will vastly expand its surveillance capabilities and supercharge the deportation system. The Homeland Advanced Recognition Technology System (HART) will collect, organize, and share invasive data on over 270 million people (including juveniles), with that number projected to grow significantly. This data will come from federal agencies including DHS and the FBI, as well as local and state police, and foreign governments. Powered by military-grade technologies, HART will aggregate and compare biometrics data including facial recognition, DNA, iris scans, fingerprints, and voice prints—most often gathered without obtaining consent or a warrant. This will allow DHS to target immigrants for surveillance, raids, arrests, detention, and deportation. HART could be used to identify people in public spaces, creating chilling consequences for people’s rights to protest, assemble, associate, and to live their daily lives. HART threatens to violate human and privacy rights at an exponential rate, particularly in Black, brown, and immigrant communities already facing discriminatory policing and surveillance. Despite the terrifying risks, HART remains a black box—shrouded in secrecy with virtually no oversight and accountability mechanisms. Although only in phase one of its development, HART has become vastly more expensive than anticipated—generating massive revenues for first, Northrop Grumman (a military contractor), and now, Veritas Capital (a billionaire private equity firm). While troubling questions over its privacy and human rights violations remain, Congress continues to fund HART, even though it has failed to meet every milestone in its government contract.

Our report explains the dangers of HART by diving into the system’s mechanics, costs, and biometric and biographic data sources. We spotlight the companies profiting from HART’s development, and the agencies, private companies, and foreign governments that will contribute to and access its data. We outline the short- and long-term civil, privacy, and human rights risks. The underlying role and impact of HART will be to turbocharge DHS’ unchecked power—to approve or deny immigration benefits, assemble target lists for ICE raids, expand the tech border wall, and to facilitate surveillance, arrests, immigrant detention and deportation. For such reasons, we call on DHS to dismantle HART. We also call on Congress to freeze funds dedicated to HART as an interim step…”





Fear the legal AI!

https://abovethelaw.com/2022/05/the-legal-consequences-of-when-who-dun-it-becomes-what-dun-it/

The Legal Consequences Of When Who Dun It Becomes What Dun It

Elementary, my dear human.

You may have already been exposed to the fear that robots will take over the profession. This, of course, is a fringe concern for now. It is hard to give serious thought to the technological singularitys making human labor and thought obsolete when aging men who have difficulty opening Word documents still move and shake so much of the industry. Even if applications to law school are down ~10% compared to last year, its a far cry from 100%. As it stands, when pro se litigants are advised to consult a competent attorney, they usually ask around for a graduate of an accredited law school rather than Siri.

But not all fields are prepared for technological innovation, and one of them is patent law.





Isn’t this obvious?

https://www.brookings.edu/research/the-case-for-placing-ai-at-the-heart-of-digitally-robust-financial-regulation/

The case for placing AI at the heart of digitally robust financial regulation

… Thanks to digitization, regulators today have a chance to gather and analyze much more data and to see much of it in something close to real time.

The potential for peril arises from the concern that the regulators’ current technology framework lacks the capacity to synthesize the data. The irony is that this flood of information is too much for them to handle. Without digital improvements, the data fuel that financial regulators need to supervise the system will merely make them overheat.

Enter artificial intelligence.





Redefinition requires rethinking. Well, I (the non-lawyer) find it amusing…

https://abovethelaw.com/2022/05/will-fracking-advocates-and-pollutant-manufacturers-become-bedfellows-if-roe-is-overturned/

Will Fracking Advocates And Pollutant Manufacturers Become Bedfellows If Roe Is Overturned?

Oklahoma’s anti-abortion law that totally isn’t just a religious belief codified into law is the strictest abortion ban in the nation. And given that the whole separation of church and state thing isn’t enough, some folks have been decided to change strategies by playing chicken — you can define life at conception all you want, but are you actually ready for that?

will child support start At conception

can we take out life insurance on fetus

can you jail a pregnant person; fetus did not commit crime or get due process

does US citizenship start at conception

tax credit?

miscarriage requires death certificate

Can a person who has suffered recurrent miscarriages in Oklahoma sue companies that introduce chemicals into the environment with a known link to miscarriages? Can we look at, say, areas around manufacturing plants where chronic infertility is higher than the general population, conclude that the polluters are to blame, and file suit? I think we should.





Tools & Techniques.

https://techcrunch.com/2022/05/24/copilot-githubs-ai-powered-coding-tool-will-become-generally-available-this-summer/

Copilot, GitHub’s AI-powered coding tool, will be free for students

Last June, Microsoft-owned GitHub and OpenAI launched Copilot, a service that provides suggestions for whole lines of code inside development environments like Microsoft Visual Studio. Available as a downloadable extension, Copilot is powered by an AI model called Codex that’s trained on billions of lines of public code to suggest additional lines of code and functions given the context of existing code. Copilot can also surface an approach or solution in response to a description of what a developer wants to accomplish (e.g. “Say hello world”), drawing on its knowledge base and current context.

While Copilot was previously available in technical preview, it’ll become generally available starting sometime this summer, Microsoft announced at Build 2022. Copilot will also be available free for students as well as “verified” open source contributors. On the latter point, GitHub said it’ll share more at a later date.



Tuesday, May 24, 2022

We never said you had to, but we kinda think you should…

https://www.databreaches.net/ftc-blog-the-ftc-act-creates-a-de-facto-breach-disclosure-requirement/

FTC Blog: “The FTC Act Creates a De Facto Breach Disclosure Requirement”

Joseph Lazarrotti of JacksonLewis writes:

On May 20, 2022, the Federal Trade Commission’s Team CTO and the Division of Privacy and Identity Protection published a blog post entitled, “Security Beyond Prevention: The Importance of Effective Breach Disclosures.” In the post, the FTC takes the position that in some cases there may be a de facto data breach notification requirement, despite there currently being no section of the Federal Trade Commission Act or implementing regulation imposing an express, broadly applicable data breach notification requirement. Businesses should nonetheless take this de facto rule into account as part of their incident response plans.

Read more at Workplace Privacy, Data Management & Security Report.





Assumption: Fighting crime is more important than privacy.

https://www.cpomagazine.com/data-privacy/mexican-supreme-court-rules-in-favor-of-warrantless-access-to-bank-data-but-strikes-down-national-registry-of-cell-phone-biometric-data/

Mexican Supreme Court Rules in Favor of Warrantless Access to Bank Data but Strikes Down National Registry of Cell Phone Biometric Data

The Supreme Court of Mexico confounded privacy advocates in recent weeks by ruling in favor of greatly expanded government access to the bank data of private citizens, but also striking down a plan for a biometric data registry for cell phone users.

Both of the propositions were based on crime issues that are currently hot topics in the country. The warrantless search measure comes as part of the current presidential administration’s campaign to rein in big companies attempting to dodge taxes and stymie money laundering by criminal cartels, while the biometric data scheme was in response to an endemic kidnapping problem that saw overall numbers drop in 2021 but violence against kidnapped migrants increase.

The finance agencies of the nation will now have automatic access to the bank data of citizens, something that previously required a court-issued search warrant.

The Supreme Court ruled that citizen’s right to privacy was secondary to the need for access to bank data to fight money laundering and tax evasion. The decision was sparked by the defense of a businessman in a tax fraud case, who argued that he should not be compelled to hand over bank data as it is private information protected by the country’s constitution.





A call for papers…

https://www.pogowasright.org/ftc-calls-for-research-presentations-for-privacycon-2022/

FTC Calls for Research Presentations for PrivacyCon 2022

The Federal Trade Commission today called for research presentations on a wide range of privacy and data security topics such as commercial surveillance and automated decision making for its annual PrivacyCon event, which will take place virtually on November 1, 2022.

PrivacyCon 2022 will bring together a diverse group of stakeholders to discuss the latest research and trends related to consumer privacy and data security. As part of this event, the FTC is seeking empirical research and demonstrations, including rigorous economic analyses, on such topics as:

    • Algorithmic bias and ensuring fairness in the use of algorithms;

    • Commercial surveillance including workplace monitoring, surveillance advertising, and biometric surveillance;

    • Potential new remedies and approaches to improve privacy and security practices such as the deletion of algorithms or other products developed using data illegally collected from consumers; and

    • Children’s and teen’s privacy risks, harms, and vulnerabilities, particularly those presented by emerging technologies.

More details on other topics and information on how to submit presentations can be found in the Call for Presentations. The deadline for submitting a presentation for PrivacyCon is July 29, 2022.

The event is free, open to the public, and will be webcast on the FTC’s website at www.ftc.gov. The agenda will be posted to the event page prior to the event.

The Federal Trade Commission works to promote competition and protect and educate consumers. Learn more about consumer topics at consumer.ftc.gov, or report fraud, scams, and bad business practices at ReportFraud.ftc.gov. Follow the FTC on social media, read consumer alerts and the business blog, and sign up to get the latest FTC news and alerts.

Source: FTC





Perhaps my AI should not share it’s inventions with me? Or should I lie and claim that I invented it?

https://www.nature.com/articles/d41586-022-01391-x

Artificial intelligence is breaking patent law

In 2020, a machine-learning algorithm helped researchers to develop a potent antibiotic that works against many pathogens (see Nature https://doi.org/ggm2p4; 2020 ). Artificial intelligence (AI) is also being used to aid vaccine development, drug design, materials discovery, space technology and ship design. Within a few years, numerous inventions could involve AI. This is creating one of the biggest threats patent systems have faced.

Patent law is based on the assumption that inventors are human; it currently struggles to deal with an inventor that is a machine. Courts around the world are wrestling with this problem now as patent applications naming an AI system as the inventor have been lodged in more than 100 countries1. Several groups are conducting public consultations on AI and intellectual property (IP) law, including in the United States, United Kingdom and Europe.

If courts and governments decide that AI-made inventions cannot be patented, the implications could be huge. Funders and businesses would be less incentivized to pursue useful research using AI inventors when a return on their investment could be limited. Society could miss out on the development of worthwhile and life-saving inventions.

Rather than forcing old patent laws to accommodate new technology, we propose that national governments design bespoke IP law — AI-IP — that protects AI-generated inventions. Nations should also create an international treaty to ensure that these laws follow standardized principles, and that any disputes can be resolved efficiently. Researchers need to inform both steps.





The miracle of AI…

https://thenextweb.com/news/mits-ai-cant-determine-a-persons-race-from-medical-images

No, MIT’s new AI can’t determine a person’s race from medical images

That's not how AI works

… Prediction and identification are two entirely different things. When a prediction is wrong, it’s still a prediction. When an identification is wrong, it’s a misidentification. These are important distinctions.

AI models can be fine-tuned to predict anything, even concepts that aren’t real.

Here’s an old analogy I like to pull out in these situations:

I can predict with 100% accuracy how many lemons in a lemon tree are aliens from another planet.
Because I’m the only person who can see the aliens in the lemons, I’m what you call a “database.”
I could stand there, next to your AI, and point at all the lemons that have aliens in them. The AI would try to figure out what it is about the lemons I’m pointing at that makes me think there’s aliens in them.
Eventually the AI would look at a new lemon tree and try to guess which lemons I would think have aliens in them.
If it were 70% accurate at guessing that, it would still be 0% accurate at determining which lemons have aliens in them. Because lemons don’t have aliens in them.

In other words, you can train an AI to predict anything as long as you:

  • Don’t give it the option to say, “I don’t know.”

  • Continue tuning the model’s parameters until it gives you the answer you want.