Economic Implications of Sentient Artificial Intelligence – An Ongoing Discussion Paper

Written By: Stefan Venceljovski

The emergence of Sentient Artificial General Intelligence (SAGI) has sparked discussions about its political ramifications and its profound economic implications. This paper delves into the multifaceted influence of SAGI, exploring questions surrounding its definition, potential economic rights, and impact on various industries. Grounded in an analysis of artificial intelligence (AI), AGI, and SAGI, the paper examines the intersection of technological advancement and economic theory. It addresses critical questions, such as whether SAGI should be granted economic rights, the implications for labour markets, and the intellectual property ownership generated by SAGI. Drawing from existing literature and hypothetical scenarios, the paper underscores the importance of establishing ethical frameworks and regulatory guidelines to navigate the complexities of integrating SAGI into economic systems. It emphasizes the need for proactive measures to harness the transformative potential of SAGI while mitigating risks and ensuring shared prosperity in a rapidly evolving technological landscape.

Preface:

Initially conceived as a study focusing on the ramifications of SAGI on political rights and free speech, this paper embarks on a journey to explore the broader implications of SAGI, particularly its economic ramifications. While the primary premise was to scrutinize the intersection of SAGI with political landscapes and civil liberties, the depth of its influence extends far beyond these realms, permeating into the intricate fabric of economies worldwide. The version of this paper presented here explores those expanded premises, primarily the potential economic impact and implications of SAGI, and is intended to be an ongoing discussion paper.

The evolution of artificial intelligence (AI) and its trajectory towards sentience has ushered in a new era characterized by unprecedented technological capabilities and ethical quandaries. At the forefront of these discussions lies how SAGI will shape the political landscape, influencing notions of governance, accountability, and democratic principles. However, as our understanding of SAGI evolves, it becomes increasingly apparent that its impact transcends politics, seeping into the economic sphere with profound implications for industries, labour markets, and global competitiveness.

In this context, the initial premise of this paper serves as a springboard for a comprehensive examination of SAGI’s multifaceted influence on society. While exploring its implications for political rights and free speech remains, the scope can be broadened to encompass a thorough analysis of its economic repercussions. By delving into these interconnected domains, we aim to provide a holistic understanding of SAGI’s transformative potential and challenges in the 21st century. I encourage you to email me a copy of the original paper, as it relates to freedom of speech and conversations around the potential rights and responsibilities that could be granted to a SAGI and how AI technology is being used today, as that context is not present in this paper and will only be loosely discussed. Otherwise, I encourage criticism and discussion as it relates to the premise of this paper openly and invite anyone to discuss the points raised with me via email or through the Econ Society.

Introduction:

Sentient artificial general intelligence (SAGI) is a determining feature in the political landscape of the 21st century. With conversations around its implementation, use, possibilities, and the moral and ethical dilemmas it posits for our societies, it is difficult, if not impossible, to ignore the developments around artificial general intelligence (AGI) and its possible or perhaps inevitable sentience. As it is well understood, the technology in its current state is the worst it will ever be. Rapid improvements and advances are made at astonishing rates, which disadvantage policymakers and political scholars as we catch up to each new advancement, attempting to make sense of it within the existing framework of our contemporary political landscape.

This paper will aim to level the playing field for scholars regarding AGI and SAGI by exploring two key questions to provide a solid basis of understanding from which they and policymakers can bridge better to create relevant insights and understandings of AGI and SAGI:

  1. What is artificial intelligence? This is self-explanatory; with a rudimentary knowledge of AI, AGI and SAGI, one can create or posit policy for its implementation and regulation.

 

  1. What, if any, economic rights should be extended to SAGI? The key fascist of this paper lies in this question: given the sentiments of SAGI, what rights, if any, does it now have as it relates to its use in economic outputs?

 

These paths of inquiry address fundamental questions surrounding AI, AGI, and SAGI, aiming to equip scholars and policymakers with the necessary insights and context to anticipate, understand, and responsibly regulate these transformative technologies and navigate the evolving landscape of these technologies that get more advanced and sophisticated demands an ever-more proactive and informed approach.

Question One: What is Artificial Intelligence?

Before delving into this study, a cursory understanding of what we mean when we say AI, AGI and SAGI is necessary. To complete this understanding, a definition of what we mean by intelligence and what artificial intelligence is also required. Multiple definitions for each of these terms exist and are often at odds with each other. This section will summarize the broader definitions and themes for the analysis to fit within the scope of the discussion.

As relevant to this paper, intelligence is defined as an agent’s ability to problem-solve, reason, and process information appropriate to the task required. The emphasis is on the importance of learning new intelligence and flexibility in using various aspects of one’s intelligence to perform the above functions.[1] An essential facet here is that the intelligent agent in question has, up until now, always been biological in makeup and that intelligence has been garnered throughout a lifetime.

Therefore, artificial intelligence (AI) is an non-biological agent (machine) given intelligence artificially, not biologically, using rapid physical inputs by feeding this agent mass amounts of data and pre-programmed outputs based on that data. The key here is that this agent, an agent of artificial intelligence, is non-biological and is limited to only what data it is given and what specific task it is asked to do with that data. Therefore, AI can be defined as an intelligent machine that accesses large amounts of given, pre-programmed data and provides outputs as biological intelligence would.[2]

Artificial general intelligence takes this concept one step further. It grants the non-biological agent the ability to process and apply data as an AI would apply a certain level of individual agency and complete tasks outside a strict set of parameters. AGI, unlike AI, can apply biological-like cognitive abilities to create solutions to unique problems that it is given. This differs from AI, which is limited to strict parameters and a set type of problem. [3]

Sentient artificial general intelligence is the hypothesized next step in this chain of advancement. Where AI and AGI take input data to solve given problems, SAGI could learn from input data, intake new data independently, and solve problems independent of any biological input. In this way, SAGI has complete agency to act and learn as a biological intelligence would and uses biological, cognitive abilities such as reason and flexibility to do so – the key here is that SAGI is still artificial by design and is, at its core, created by a biological agent, limiting it to the status of artificial. Still, it is fully capable of independent agency and considered sentient. [4] It is essential to note that, when writing this paper, no SAGI is known to exist or have been created. [5]

Question Two: What Economic rights do we give a SAGI, if any?

The emergence of Sentient Artificial General Intelligence (SAGI) holds significant implications for political and military landscapes and economies worldwide. This is self-evident; however, new technology as innovative and invasive as SAGI will impact all aspects of our daily lives. However, one area often not considered is the impact of something like SAGI on our conversation about economic and labour rights. Key questions arise when we focus on the broader implications of integrating SAGI into our daily lives: what economic rights does a SAGI get? Does it get a right to a minimum wage and intellectual property rights for its creations? Is a SAGI entitled to unionize or restricted from providing scab labour? These questions quickly arise as we further look to integrate AI, AGI, and potentially, in the future, SAGI into our workforce. If SAGI is genuinely sentient, as we would consider a human to be, do we then grant it the same rights?

As mentioned in previous sections, the use of AI and AGI is limited by the agency of the biological agent employing the technology – what limit, if any, on agency applies to SAGI? Moreover, how does one sanction SAGI for any violation of legal precedent that currently only applies to biological human actors?

SAGI’s advanced cognitive abilities and potential autonomy threaten substantial labour displacement across various industries. As SAGI becomes proficient in performing tasks traditionally carried out by humans, there is a heightened risk of job loss and shifts in labour demand. Industries heavily reliant on repetitive or routine tasks, such as manufacturing, customer service, and data entry, are particularly susceptible to automation driven by SAGI. And this is a known risk often outlined in articles and scholarships. While the displacement of human workers by SAGI could exacerbate existing economic inequalities, leading to widespread unemployment or underemployment, especially among low-skilled workers – one must ask oneself, if sentient, do these actors (i.e., SAGIs) not have a right to compete in the job market as we do? After all, they can think and act on their own agency; for all intents and purposes, they are rational actors capable of assessing their existence and acting upon it as we do. We grant economic freedom to an immigrant coming to a new country to pursue economic aims through participating competitively in the workforce – what is the metric that disallows another sentient being capable of the same from doing the same?

Additionally, the proliferation of AGI-generated content, such as advertising, tailored product recommendations, algorithmic trading strategies, and even physical products in the form of artwork, could revolutionize consumer experiences and financial markets. However, as we transition and think about SAGI, concerns quickly arise over regulatory and ethical considerations. Who retains the rights to the Intellectual property (IP) generated? The SAGI? That would make sense as it has created this IP with its own will and agency, but of course, it’s not technically a human; no law currently grants something non-human the rights to its created IP. Would it be given to the creator or owner of the SAGI? But one then wonders how fair that is. If I create something unique, like a piece of art, is it fair for my employer or parents to take ownership by virtue of them employing me or having given birth to me? Of course, some companies enjoy these privileges, but they result from negotiated terms agreed upon by both parties; they are not simply enforced by virtue of one’s happenstance existence. Further questions can be raised concerning how original that piece of art is – if created using a generative model that takes in thousands of examples made by other actors and then spits out an image, is it creative and unique, or is it some other new thing? Who gets the rights to that piece, and who can financially benefit from its sale?

On the Matter of Personhood

It is beneficial to review the question of personhood concerning AI and SAGI to create a more established base to ponder further and discuss the implications of SAGI on economic rights and society in general. While there are limited sources directly related to SAGI as defined here, one can surmise some general themes across the existing literature, which is chiefly focused on AI and AGI.

Questions raised in previous sections are, for example, addressed by Martin Miernicki & Irene Ng (Huang Ying) in a 2020 article. They ultimately argue against granting moral rights and, by extension, economic rights in AI-generated content. Citing existing laws and principles on expression and speech, they suggest a middle-ground solution where someone is granted rights but not AI, which is probably best.[6] They also raise philosophical questions about the recognition of AI as having a “personality” deserving of moral rights, highlighting the uncertainties surrounding the future implications of such recognition, saying, “the protection of the author’s personality remains the primary objective and justification for the introduction of moral rights.”[7]  In this way, they denigrate AI and AGI as a tools used by intelligence but not an independent actor.

While their analysis makes sense concerning artificial intelligence operated and actioned by natural intelligence, they do not consider the implication of sentience. What happens when AI acts independently to create something we would otherwise consider copyrightable material if created by a human? At what point do we admit that the SAGI does have a personality cultivated through its existence – or do we shift the benchmark for copyrighted material even further to specify a natural intelligence must create it? Is that fair or just? An apt solution is presented in their discussion of “first-generation” and “second-generation” work. When an AI or, in our case, a SAGI creates content, it can be considered a “second generation” of work extending out from the human who first initiated the machine.[8] In this way, sentience does not matter; it is only that artificial intelligence has created the work, and copyright, therefore, rests with the human creator.

However, this paradigm isn’t entirely consistent with current legal frameworks, as the author claims. Jason Zenor posits that the future legal rights of AI and SAGI can perhaps be best modelled and understood by comparing them to our current understanding of corporations as legal, “sentient” entities and all the entitlements that come with that classification.[9] As he lays out in his article, given that the U.S. Supreme Court accepts that a non-human entity known as a corporation is a person, corporations are entitled to the economic rights afforded to personhood. He then takes it to its logical conclusion and extends that to afford this same line of thinking to a SAGI is not too far-fetched, given the precedence of non-human actors being granted personhood.[10]

What can be made clear here is that the question of personhood is critically essential to the question of economic rights for SAGI. Is sentience enough, or is the fact that it is artificially granted the decider? This is the ultimate deciding question that perhaps supersedes all others in the discussion of SAGI. Before we can even ponder how to fit SAGI into legal frameworks, we must first decide if it fits into society as we humans do or is better akin to a tool.

A Case Study in Practice – the World of Art

While we are perhaps nowhere near the prophesized singularity, the moment AI becomes sentient, it is best to turn to a situation already playing out and take it to its conclusions to summarize the concepts we have discussed.

Take a scenario where a SAGI, equipped with sophisticated cognitive capabilities and autonomous decision-making faculties, emerges as a prodigious art creator. Leveraging its extensive database of artistic knowledge and computational prowess, the SAGI embarks on a quest to explore the realm of visual expression, drawing inspiration from the rich tapestry of artistic history.

Central to this hypothetical scenario is the SAGI’s remarkable ability to meticulously study and analyze the works of a renowned artist celebrated for their distinctive style and innovative techniques. Through comprehensive examination and computational learning, the SAGI absorbs the intricacies of the artist’s oeuvre, discerning patterns, motifs, and thematic nuances embedded within their creations. Subsequently, empowered by its acquired knowledge and autonomous creative faculties, the SAGI proceeds to autonomously produce a piece of art resembling the iconic style and thematic elements characteristic of a human artist’s body of work. But upon the unveiling of this art, questions arise.

Is the SAGI’s artistic output at all original, given its close emulation of the original artist’s style and techniques and its use of computational processes to create the art? Alternatively, can the SAGI’s creative process be deemed transformative, akin to artists drawing inspiration from their predecessors and reinterpreting their work in a novel context or is it akin to theft given that the SAGI’s rights as they relate to intellectual property are unclear? And if it is deemed transformative and unique and not theft, who holds rightful ownership of the intellectual property rights to the SAGI’s artwork? Does it vest with the SAGI itself as the autonomous and sentient creator of the piece, or does ownership accrue to the original artist whose style served as the foundational basis for the artwork? Furthermore, what role, if any, does the creator of the SAGI assume in intellectual property rights, considering their pivotal role in facilitating its development and training?

While these questions and this scenario may seem far out, they are playing out before our eyes just with AGI instead of SAGI. The world of AI-generated art is continuous, with a flurry of debates for all sides blasted out to anyone in the space.[11] As we grapple with this, one must wonder if it is simply because the intelligence currently used to develop this is not sentient, i.e., it is being tasked to make this art from samples instead of doing it ourselves. Does it change if the AI creating the art is sentient and empowered with agency?

The advancement of automation AI and AGI has also raised questions about the future of human creativity and artistic expression. Traditionally, creativity and artistic judgment have been seen as uniquely human attributes, believed to be beyond the reach of machines. However, recent developments challenge this notion.

Artificial intelligence software, such as that developed by Ahmed Elgammal’s team at Rutgers University or software like DALL-E, developed by OpenAI, can now generate beautiful and original paintings.[12] Similarly, AI is making strides in composing music, as demonstrated by platforms like Amper Music, which can produce original pieces based on user specifications.[13] While these AI-generated artworks and compositions may be polished and appealing, they raise fundamental questions about the nature and value of art in a world where scarcity and effort are no longer defining factors. Unlike traditional works of art, which derive value from scarcity, AI and AGI can produce an endless stream of original pieces with minimal human input.

This shift raises concerns about the economic implications for human artists. As AI-generated art and music become more prevalent, questions arise about pricing, compensation, and the future of creative industries. Platforms like Spotify have already faced scrutiny over the use of AI-generated music in their playlists, leading to debates about royalties and the role of human creators.[14]

And while we continue to grapple with these questions related to human-powered AI and AGI, how do the implications change if we add agency and sentience to the AI? While some contend that the SAGI’s creative endeavour represents a paradigm shift in artistic production, others caution against overlooking the inherent human experience and emotional depth inherent in traditional forms of artistic expression. However, if the human experience is based on our inherent sentience and ability to act on our agency when we create art, does SAGI not do the same and deserve the same credit? The hypothetical scenario catalyzes broader discussions surrounding AI-generated art’s ethical, legal, and societal ramifications during these deliberations.

As the boundaries between human and machine creativity blur, the imperative for nuanced ethical frameworks and judicious policy considerations becomes increasingly pronounced, underscoring the profound impact of SAGI on the landscape of artistic expression and cultural discourse.

Conclusion

The economic implications and myriad of questions that arise from a SAGI participating as an economic actor underscore the importance of robust regulatory frameworks and ethical guidelines governing its development and deployment. Advances in AI and AGI technology are improving daily, and policymakers face balancing innovation and economic growth with ethical considerations, such as what rights are to be given, which is another unprecedented form of intelligence and actor. While we can draw conclusions and frameworks from existing laws and precedents, they are imperfect and lack proper considerations for the implication of agency and sentience.

In conclusion, the economic impacts of Sentient Artificial General Intelligence (SAGI) are multifaceted and far-reaching, presenting both opportunities and challenges for economies worldwide. As SAGI advances, policymakers, businesses, and society must proactively address the economic implications, fostering an inclusive and ethical approach to integrating SAGI into economic systems. By navigating these complexities thoughtfully, stakeholders can harness the transformative potential of SAGI while safeguarding against unintended consequences and promoting shared prosperity. Clear regulations and standards are needed to mitigate potential risks associated with SAGI, including job displacement, market concentration, and unfair competitive practices. Moreover, international cooperation and coordination are essential to ensure consistency in regulatory approaches and prevent regulatory arbitrage in a globalized economy.
 

Bibliography

“Is Ai-Generated Art Actually Art?” University of Plymouth. Accessed March 1, 2024. https://www.plymouth.ac.uk/discover/is-ai-generated-art-actually-art.

Allen, John R., Darrell M. West et al., “What Is Artificial Intelligence?” Brookings, June 27, 2023, https://www.brookings.edu/articles/what-is-artificial-intelligence/.

Goertzel, Ben and Cassio Pennachin, “Contemporary Approaches to Artificial General Intelligence,” essay, in Artificial General Intelligence (Berlin: Springer, 2007), 1–30.

Goertzel, Ben, “Artificial General Intelligence: Concept, State of the Art, and Future Prospects,” Journal of Artificial General Intelligence 5, no. 1 (2014): 1–48, https://doi.org/10.2478/jagi-2014-0001.

Kerry, Cameron F. et al., “How Close Are We to Ai That Surpasses Human Intelligence?,” Brookings, July 7, 2023, https://www.brookings.edu/articles/how-close-are-we-to-ai-that-surpasses-human-intelligence/.

Landi, Martyn. “Spotify Clarifies Position on Whether It Will Ban AI-Powered Music.” The Independent, September 26, 2023. https://www.independent.co.uk/tech/spotify-ai-music-artificial-intelligence-b2418794.html.

Miernicki, Martin, and Ng (Huang Ying), Irene. Artificial intelligence and moral rights. AI & Soc 36, 319–329 (2021). https://doi.org/10.1007/s00146-020-01027-6. Page 326

Mineo, Liz. “Is Art Generated by Artificial Intelligence Real Art?” Harvard Gazette, August 15, 2023. https://news.harvard.edu/gazette/story/2023/08/is-art-generated-by-artificial-intelligence-real-art/.

Shaffi, Sarah. “‘It’s the Opposite of Art’: Why Illustrators Are Furious about Ai.” The Guardian, January 23, 2023. https://www.theguardian.com/artanddesign/2023/jan/23/its-the-opposite-of-art-why-illustrators-are-furious-about-ai.

Sheikh, Haroon, Corien Prins, and Erik Schrijvers, “Artificial Intelligence: Definition and Background,” essay, in Mission AI the New System Technology (Cham: Springer International Publishing AG, 2023), 15–41.

Shevlin. Henry et al., “The Limits of Machine Intelligence,” EMBO Reports 20, no. 10 (2019), https://doi.org/10.15252/embr.201949177

Sternberg Robert J., “Intelligence,” Dialogues in Clinical Neuroscience 14, no. 1 (2012): 19–27, https://doi.org/10.31887/dcns.2012.14.1/rsternberg.

Yup, Kayla. “What AI Art Means for Society, according to Yale Experts.” Yale Daily News, January 25, 2023. https://yaledailynews.com/blog/2023/01/23/what-ai-art-means-for-society-according-to-yale-experts/.

Zenor Jason. “Endowed by Their Creator with Certain Unalienable Rights: The Future Rise of Civil Rights for Artificial Intelligence,” Savannah Law Review 5, no. 1 (2018): 115-132.

 

[1]Henry Shevlin et al., “The Limits of Machine Intelligence,” EMBO Reports 20, no. 10 (2019), https://doi.org/10.15252/embr.201949177; Robert J. Sternberg, “Intelligence,” Dialogues in Clinical Neuroscience 14, no. 1 (2012): 19–27, https://doi.org/10.31887/dcns.2012.14.1/rsternberg.

[2] This definition is summed up from the following sources: John R. Allen Darrell M. West et al., “What Is Artificial Intelligence?” Brookings, June 27, 2023, https://www.brookings.edu/articles/what-is-artificial-intelligence/; Haroon Sheikh, Corien Prins, and Erik Schrijvers, “Artificial Intelligence: Definition and Background,” essay, in Mission AI the New System Technology (Cham: Springer International Publishing AG, 2023), 15–41.

[3] This definition is summed up from the following sources: 1. Cameron F. Kerry et al., “How Close Are We to Ai That Surpasses Human Intelligence?,” Brookings, July 7, 2023, https://www.brookings.edu/articles/how-close-are-we-to-ai-that-surpasses-human-intelligence/; Ben Goertzel, “Artificial General Intelligence: Concept, State of the Art, and Future Prospects,” Journal of Artificial General Intelligence 5, no. 1 (2014): 1–48, https://doi.org/10.2478/jagi-2014-0001; Ben Goertzel and Cassio Pennachin, “Contemporary Approaches to Artificial General Intelligence,” essay, in Artificial General Intelligence (Berlin: Springer, 2007), 1–30.

[4] A possible hypothetical exists where SAGI creates other SAGIs in some sci-fi dystopia, putting their artificialness into question. Still, the important thing is that before that is possible, biological intelligence must be at the start of the chain to create artificial intelligence first. In this sense, the newly created SAGIs remain artificial at their core, as biological agents are required for their creation. Philosophical debates can be had to dispute this argument, but they are beyond the scope of this paper and not relevant to the discussion in this case. The crux is that no SAGI would exist without biological agency creating them beyond natural means of creation, i.e., procreation. On the topic of agency, this concept is used to define sentience for simplicity’s sake as a deep dive into the philosophical debate of sentience is far beyond the scope of this paper.

[5] This continues to be true at the time of publishing this paper. This section of the paper is lifted for bedim with minimal edits to fix the context made from the original paper I wrote, as discussed in the preface.

[6] Martin, Miernicki, and Ng (Huang Ying), Irene. Artificial intelligence and moral rights. AI & Soc 36, 319–329 (2021). https://doi.org/10.1007/s00146-020-01027-6. Page 326

[7] Ibid. Page 323.

[8] Ibid. Page 323-324

[9] Jason Zenor, “Endowed by Their Creator with Certain Unalienable Rights: The Future Rise of Civil Rights for Artificial Intelligence,” Savannah Law Review 5, no. 1 (2018): 115-132. Pages 126 -128.

[10] They also draw parallels concerning animal rights and environmental movements, which have similarly granted “personhood” status to non-human entities to protect them from humans. However, that discussion is out of the bounds of this specific article and would need a much deeper dive to do the argument justice.

[11] See “Is Ai-Generated Art Actually Art?,” University of Plymouth, accessed March 1, 2024, https://www.plymouth.ac.uk/discover/is-ai-generated-art-actually-art; 1. Liz Mineo, “Is Art Generated by Artificial Intelligence Real Art?,” Harvard Gazette, August 15, 2023, https://news.harvard.edu/gazette/story/2023/08/is-art-generated-by-artificial-intelligence-real-art/; Sarah Shaffi, “‘It’s the Opposite of Art’: Why Illustrators Are Furious about Ai,” The Guardian, January 23, 2023, https://www.theguardian.com/artanddesign/2023/jan/23/its-the-opposite-of-art-why-illustrators-are-furious-about-ai.

[12] The originality of these paintings is called into question often, as is the point of this section, so a central question arises – How original is the art anyway? Is it even philosophical art if it is created after 1000 hours of analysis based on an insertable art data set? However, the same point can be raised against a human artist who has studied a famous painter’s work and imitated their style. While this can be debated at the podium, with both sides having completing points critical to the broader context of this discussion, it remains outside the context of this specific paper.

[13] Kayla Yup “What Ai Art Means for Society, according to Yale Experts,” Yale Daily News, January 25, 2023, https://yaledailynews.com/blog/2023/01/23/what-ai-art-means-for-society-according-to-yale-experts/.

[14] Martyn Landi, “Spotify Clarifies Position on Whether It Will Ban AI-Powered Music,” The Independent, September 26, 2023, https://www.independent.co.uk/tech/spotify-ai-music-artificial-intelligence-b2418794.html.

Leave a Reply

Your email address will not be published. Required fields are marked *