Hello!

Thesis: As artificial intelligence gets more powerful and useful in the research and design process for new inventions and innovations, many new products are being found, many that don’t even exist yet. Currently, the rights to ownership of these new inventions are unknown.

Welcome to the Insights, Innovation, and Economics blog. If you’re new here, feel free to read my general Introduction to the Blog to understand more about the blog. If you’re returning, thank you, and hope you have a great read!

If you haven’t read my Introduction to Intellectual Property, I’d highly recommend it before reading this article as some of the terminology associated with this subject may be difficult to understand.

Credit Case Western Reserve University

Artificial Intelligence

If you’ve been anywhere in the last year, you’ve probably heard the words “artificial intelligence” or “AI”.

Credit Statista

At this point, artificial intelligence has become pretty much a meaningless buzzword for most. Who actually knows what artificial intelligence is?

Here’s a quick primer on artificial intelligence (not super in-depth, but enough to understand and be able to comprehend the rest of this article):

IBM writes:

Artificial intelligence leverages computers and machines to mimic the problem-solving and decision-making capabilities of the human mind.

They continue to cite a 2004 paper by John McCarthy:

It is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable.

Artificial intelligence, or AI, refers to complex computer systems that can perform tasks that typically require human intelligence, such as visual perception, speech recognition, decision-making, and language translation. The key aspect of AI systems (what makes them revolutionary) is that they learn from experience rather than being directly programmed.

In essence, artificial intelligence enables computers and machines to handle cognitive tasks and make decisions the way humans would, by “learning” from the world around them rather than solely following static program instructions. The goal of AI is to create intelligent software and machines that are more helpful, effective, and intuitive.

The artificial intelligence discussion is generally understood to have started with Alan Turing’s “Computing Machinery and Intelligence”, published in 1950. Turing asks the fundamental question “Can machines think?” From there, he defined the “Turing Test” where a human would try to distinguish between a computer and human text response.

Following this, Stuart Russell and Peter Norvig published “Artificial Intelligence: A Modern Approach”, one of the leading textbooks on artificial intelligence. They describe 4 potential definitions of AI, which differentiates computer systems based on rationality and thinking vs. acting:

Human Approach:

1. Systems that think like humans (e.g., image recognition software, chatbots)
2. Systems that act like humans (e.g., video game NPCs, robot receptionists)

Ideal Approach:

3. Systems that think rationally (e.g., stock trading algorithms, chess AIs)
4. Systems that act rationally (e.g., smart grid infrastructure, industrial automation)

In its simplest form, artificial intelligence is a field that combines computer science and robust datasets to enable problem-solving.

Over the years, AI has gone through many hype cycles (including the one happening right now). The current AI hype cycle has largely been dominated by generative AI (primarily developed by OpenAI and Anthropic).

Credit Towards Data Science

Generative AI

Generative AI refers to artificial intelligence systems that are capable of creating new content on their own, rather than simply analyzing existing data sets. These models take raw data (e.g., all of Wikipedia, etc.) and “learn” from it to generate outputs when prompted.

The “generative” part means these AI models can generate brand new images, sounds, text, and other outputs from scratch (this work is similar, but not identical to the original data it was “trained” with).

The key technique underlying many of these systems is machine learning, allowing the AI model to study vast data sets to discern subtle patterns. By learning the relationships between elements, the AI then attempts to create never-before-seen content that plausibly fits together.

If an artificial intelligence system is producing new content rather than merely recognizing existing data, it falls into the area of generative AI.

Credit VICE

Artificial Intelligence Discoveries

As artificial intelligence becomes more and more sophisticated, people continually find more and more uses, more ways for it to produce new content, new materials, and new ideas. Throughout these many uses, people have been discovering more and more new technologies that had previously been unknown.

In November 2023, Google highlighted its new DeepMind program, the builder of the artificial intelligence tool GNoME, which found 2.2 million new crystals, 380,000 of which are stable materials that could be used to power future technologies.

They cite these discoveries as equating to nearly 800 years’ worth of knowledge.

So who owns these artificial intelligence discoveries?

The artificial intelligence company? The person using the AI to discover the new technology?

The American Enterprise Institute writes the following introduction:

While the artificial intelligence (AI) application ChatGPT going mainstream has engendered thousands of column inches and terabytes of content in the physical and digital worlds respectively, in the back offices of academia and the courts, a new flurry of activity has emerged concerning the question of who—if anyone—owns the rights to outputs produced by these revolutionary new technologies.

As it currently stands, it’s only a matter of time before a tsunami of legal cases land in the courts as rights holders seek to determine the extent to which artificial intelligence infringes upon existing content.

Current cases are against artificial intelligence companies for training their models with billions of copyrighted data - more focused on the inputs to the AI process.

However, new issues are arising from the outputs of these AI models. Is the artificial intelligence company responsible or the user? Who owns the output, the artificial intelligence company or the used?

At one end of the scale, we see small AI-driven innovations (e.g., where someone uses text editing software to help them write an article). At the other end of the scale, we see largely AI-driven innovations (a human simply provides training data to an AI and presses go).

Credit SCC Online

Consider this example:

Professor Ryan Abbott of the University of Surrey in New Zealand applied for a patent, citing an artificial intelligence platform (DABUS) as the “inventor” (That is, the company that owns an inventive AI is entitled to the patents for its creations. More generally, the company owning a creative AI is entitled to the intellectual property in the outputs it creates).

The New Zealand Intellectual Property Office declined to issue the patent, ruling that “inventor” only refers to a natural person, which an artificial intelligence platform is not.

The American Enterprise Institute cites the following:

So far, only South Africa has recognized [artificial intelligence platforms] as an inventor. The UK court came to a similar finding as New Zealand, but its ruling—along with 11 other cases from the experiment, including the US—is either under appeal or still awaiting a decision.

The rights to artificial intelligence discoveries need careful consideration because this concept goes to the heart of intellectual property rights, as the World Trade Organization explains in their brief on intellectual property:

The protection and enforcement of intellectual property rights should contribute to the promotion of technological innovation and to the transfer and dissemination of technology, to the mutual advantage of producers and users of technological knowledge and in a manner conducive to social and economic welfare, and to a balance of rights and obligations.

Artificial intelligence machines are not humans. As such, they don’t need incentives, such as intellectual property rights, to create works currently protectable by intellectual property rights.

In other words, artificial intelligence would create these works with or without intellectual property incentives. However, the absence of intellectual property protection may disincentivize humans from creating works.

So, how should the balance be struck?

Currently, the system exists to prevent artificial intelligence companies from protecting outputs generated by AI, leaving that to the human running the AI (inputting prompts). In most countries, if no human can be identified for the invention, it will be unable to be patented and therefore it will fall into the public domain.

This leaves us with 2 very important questions: 1) Does the law need to be revised? and, 2) Will the law be revised?

Credit London Law Student

Question #1: Does the law need to be revised?

As it currently stands, artificial intelligence remains a tool to expand human capabilities but is not yet capable of independent inventorship or conception.

The level of debate surrounding ownership of the works created by artificial intelligence would suggest that current laws need to be looked at to determine suitability, with an eye particularly to future developments.

Changing laws to include artificial intelligence differs by country. The United States government has approached this problem as a public policy issue, rather than relying on courts to interpret the wording of laws that were created significantly before AI emerged.

Part of this reason is because the involvement of artificial intelligence in protected works is only partially compatible with the foundation of the intellectual property system.

Many proponents throughout the intellectual property community are clear about their stance on AI inventions and their subsequent protections. Consider one excerpt from the Center for Strategic and International Studies:

As AI becomes more ubiquitous and makes greater contributions to the invention process, the role of protecting and promoting human inventorship becomes critical. Humans should continue to be recognized as the actual inventors, even as these inventors use AI as a “sophisticated tool” of experimentation and discovery.

Question #2: Will the law be revised?

This may seem like a similar question to the one above, but it approaches the problem from a different angle. In the above section, I highlight how intellectual property proponents are discouraging artificial intelligence from being able to protect its intellectual property as it isn’t human.

Yet, that doesn’t mean that they will prevail. It’s a classic case of just because something shouldn’t be done doesn’t mean it won’t be done.

As countries like the United States approach the problem of artificial intelligence intellectual property protection from a policy standpoint instead of a judicial one, they leave the door open to lobbying.

For these increasingly large artificial intelligence entities (OpenAI recently valued at $100+ billion), being able to control the property of the output of your artificial intelligence entity would be massively economically important, as you could own the rights to the next big drug or the next big product.

In this case, if the law were to change and artificial intelligence entities were the owners of intellectual property from their outputs, we could see many scenarios where an artificial intelligence company becomes a true monopoly due to its massive intellectual property portfolio.

But, is this realistic? Would the law actually change even though most people wouldn’t want it to change?

Consider the examples of many of the climate change legislation. This would be the opposite example, where most people would want the laws and policies to change, but certain powerful entities (in this case fossil fuel companies) petition and fight to have the laws continue as they are.

Credit True North Coaching

Lack of Clear Direction

The rise of artificial intelligence (especially generative AI) raises many key questions relating to independent creativity and the originality of ideas (in addition to their ownership).

Is our society better off when artificial intelligence is creating most things? In a world dominated by AI, will the creativity and originality of ideas generated by humans be significantly impacted? Will humans ever stop being able to compete with AI systems?

Artificial intelligence may end up being an example of “too much of a good thing” in the coming decades. Who knows?

Either way, artificial intelligence looks like it is here to stay, yet the policies and regulations haven’t caught up.

Many government authorities have given the public a lack of clear direction when it comes to artificial intelligence and intellectual property rights. People understand how the system works currently, but there’s the threat that that could change at any moment.

Key questions such as the following lack clear answers: “Are you able to use Artificial Intelligence to discover new inventions and still be able to protect them yourself?” or “Is it even worth using artificial intelligence for research?”

As the Center for Strategy and International Studies puts it:

Investors and innovators need clarity that the use of AI in the invention creation process will not lead to the invalidation of patents down the line.

For artificial intelligence to become more widely used in the invention process, policies and regulations need to be clear on the effects of using artificial intelligence and subsequent intellectual property claims on any outputs generated.

Without this, innovation is being unfairly curtailed.

Anywho, that’s all for today.

-Drew Jackson