ChapterPDF Available

AI Personhood: Rights and Laws

Authors:

Abstract

It is possible to rely on current corporate law to grant legal personhood to artificially intelligent (AI) agents. Such legal maneuvering may be useful to avoid human responsibility or to further automate businesses. In this chapter, after introducing pathways to AI personhood, consequences of such AI empowerment on human dignity, human safety, and AI rights are analyzed. This chapter per the author emphasizes possibility of creating selfish memes and legal system hacking in the context of artificial entities. Finally, potential solutions for addressing described problems are considered.
Human Indignity:
From Legal AI Personhood to Selfish Memes
Roman V. Yampolskiy
Computer Engineering and Computer Science
University of Louisville
roman.yampolskiy@louisville.edu
Abstract
It is possible to rely on current corporate law to grant legal personhood to Artificially Intelligent
(AI) agents. In this paper, after introducing pathways to AI personhood, we analyze consequences
of such AI empowerment on human dignity, human safety and AI rights. We emphasize possibility
of creating selfish memes and legal system hacking in the context of artificial entities. Finally, we
consider some potential solutions for addressing described problems.
Keywords: AI Personhood, Litigation Attack, Malevolent Corporation, Selfish Meme, Zero-
Day Loophole
The Machine is not an it to be animated, worshiped, and dominated.
The machine is us, our process, an aspect of our embodiment.
We can be responsible for machines; they do not dominate or threaten us.
We are responsible for boundaries; we are they.- Donna Haraway
1. Introduction to AI Personhood
Debates about rights are frequently framed around the concept of legal personhood, which is
granted not just to human beings but also to some non-human entities, such as firms, corporations
or governments. Legal entities, aka legal persons are granted certain privileges and responsibilities
by the jurisdictions in which they are recognized, and many such rights are not available to non-
person agents. Attempting to secure legal personhood is often seen as a potential pathway to get
certain rights and protections for animals [1], fetuses [2], trees, rivers [3] and artificially intelligent
(AI) agents [4]. It is commonly believed that a court ruling or a legislative action is necessary to
grant personhood to a new type of entity, but recent legal literature [5-8] suggests that loopholes
in the current law may permit granting of legal personhood to currently existing AI/software
without having to change the law or persuade any court.
LoPucki [6] in his paper on Algorithmic Entities cites Bayern [7, 8] and his work on conferring
legal personhood on AI by putting it in charge of LLC
1
. “Professor Shawn Bayern demonstrated
1
Bayern specifies this chain of events as capable of establishing the link: (1) [A]n individual member creates a
member-managed LLC, filing the appropriate paperwork with the state; (2) the individual (along, possibly, with the
LLC, which is controlled by the sole member) enters into an operating agreement governing the conduct of the LLC;
(3) the operating agreement specifies that the LLC will take actions as determined by an autonomous system,
that anyone can confer legal personhood on an autonomous computer algorithm merely by putting
it in control of a limited liability company (LLC). The algorithm can exercise the rights of the
entity, making them effectively rights of the algorithm. The rights of such an algorithmic entity
(AE) would include the rights to privacy, to own property, to enter into contracts, to be represented
by counsel, to be free from unreasonable search and seizure, to equal protection of the laws, to
speak freely, and perhaps even to spend money on political campaigns. Once an algorithm had
such rights, Bayern observed, it would also have the power to confer equivalent rights on other
algorithms by forming additional entities and putting those algorithms in control of them.”
2
[6].
Other legal pathways to obtain legal personhood have been suggested and analyzed in the literature
[4-12], but details of such legal hacking are beyond the scope of this paper. We are simply
interested in understanding the impact of granting personhood to AI on human dignity [13] and
safety. With appearance of decentralized autonomous organizations [14], such as the DAO [15],
these questions are as pressing as ever.
2. Selfish Memes
In his book, The Selfish Gene, Dawkins [16] talks about genes as the driving payload behind
evolution, with animal bodies as vehicles for the gene to accomplish its goals in the world. He also
introduced a new concept of a meme, or viral idea competing for dominance in human minds,
inspired by the similarities between natural and cultural evolution. Advent of algorithmic entities
would make it possible to explicitly add a memetic payload to a legal entity, resulting in what we
will call the Selfish Meme. Corporations are selfish entities with the goal of maximizing
shareholder profit, with AI in charge of such an entity any idea can be codified in an algorithm and
added as the driving force behind the corporation’s decision making. At a higher level of
abstraction this could produce selfish cryptocurrencies.
We already see something similar from B-corps or for Benefit Corporations, which attempt to
create some social good in addition to profit. However, such memetic payload doesn’t have to be
strictly beneficial, in practice it could be any ideology, set of beliefs or values. For example, it
would be possible to codify tenants of a particular religion (ex. Islam), economic philosophy
(communism), moral theory (Utilitarianism) or something silly but potentially dangerous like a
Paperclip Maximizer [17] or Pepe meme [18], encode them in an algorithm and put that algorithm
in charge of a corporation, which could eventually take over the world and inforce its memetic
payload on the world. Via orthogonality thesis [19] we can see that few if any limitations exists
on the potential memetic payload, it could be a marketing campaign, an uploaded animal or human
mind, our constitution and the complete set of laws or a computer virus. Evolutionary competition
would appear between such entities leading to adversarial practices [20] and perhaps hostile
takeovers not just in a legal but also in a computer science sense, with hacking and replacement of
corporations selfish meme with another payload, being a real possibility. We live in the world
where computer viruses may have rights. It would also not be surprising if establishment of
specifying terms or conditions as appropriate to achieve the autonomous system’s legal goals; (4) the sole member
withdraws from the LLC, leaving the LLC without any members. The result is potentially a perpetual LLCa new
legal personthat requires no ongoing intervention from any preexisting legal person in order to maintain its status.
AEs would not be confined to cyberspace. An AE could act offline by contracting online with humans or robots for
offline services. Bayern uses an algorithm that operates a Bitcoin vending machine business to illustrate.”
2
See original article for footnotes, which have been removed to improve readability of quotes.
corporations with custom memetic payload become available as a service. Such corporations may
also attempt to gain membership in different organizations and partnerships for example the
recently formed Partnership on AI
3
, in order to influence it from within.
3. Human Indignity
The process for getting legal rights for AI, described above, doesn’t specify any minimal
intelligence/capability for the AI involved, meaning it could be just a few “if” statements, a random
decision generator or an emulation of an ameba
4
. To grant most if not all human rights to a
cockroach, for example, would be an ultimate assault on human dignity (but it may make some
people happy [21]). This could be potentially done as an art project or in protest of unequal
treatment of all humans by some human rights activists. We already witnessed an example of such
indignity and subsequent outrage from many feminist scholars [22] on the news of Sophia the
robot getting citizenship in the Saudi Arabia, a country notorious for unequal treatment of women.
Will self-driving cars be allowed on the roads before women are?
5
As a result of legal personhood
and granting of associated rights, some humans will have less rights than trivial (non-intelligent)
software and robots, a great indignity and discriminatory humiliation. For example, certain
jurisdictions limit rights of their citizens, such as a right to free speech, freedom or religious
practice, or expression of sexuality, but AIs with legal personhood in other jurisdictions would be
granted such rights.
If, on the other hand, AIs are going to become more intelligent than humans, the indignity for
humanity would come from being relegated to an inferior place in the world, being outcompeted
in the workplace and all other domains of human interest [23, 24]. AI led corporations would be
in position to fire their human workers. This would possibly lead to deteriorating economic and
living conditions, permanent unemployment and potentially reduction in rights, not to mention
further worsening of the situation including to the level of existential catastrophe (extermination)
[25].
The precedent of AI obtaining legal personhood via the corporate loophole may catalyze legislative
granting of equal rights to artificially intelligent agents as a matter of equal treatment, leading to a
number of indignities for the human population. Since software can reproduce itself almost
indefinitely, they would quickly make human suffrage inconsequential, if given civil rights [26]
leading to the loss of self-determination for people. Such loss of power would likely lead to the
redistribution of resources from humanity to machines as well as possibility of AIs serving as
leaders, presidents, judges, jurors and even executioners. We will see military AIs targeting human
populations and deciding on their own targets and acceptable collateral damage. They may not
necessarily subscribe to the Geneva Convention and other rules of war. Torture, genocide and
nuclear war may become options to consider to reach desired goals.
As AIs’ capabilities and dominance grow, they would likely self-grant special (super) rights to
emphasize their superiority to people, while at the same time removing or at least reducing human
rights (ex. 2nd amendment, 1st amendment, reproductive rights in the sense of the right to reproduce
at all, aka, 0-child policy, Convention on human rights, etc.) while justifying doing so by our
3
https://www.partnershiponai.org/
4
Same legal loophole could be used to grant personhood to animals or others with inferior rights.
5
As of June 24, 2018 and after this was written, women were permitted to drive in Saudi Arabia.
relative “feeblemindedness”. A number of scholars [27-29] today work on developing reasons for
justifying granting of rights to AIs, perhaps one day those reasons will be useful while we are
begging to keep some of ours.
4. Legal-System Hacking
Corporations can act as their own lawyers while representing themselves in the court of law,
including performing all functions of a human lawyer, such as sue and be sued. Artificial
superintelligence in charge of a corporation can act as a super-lawyer capable of finding novel
loopholes in our laws (zero-day law exploits), engaging in frivolous litigation (DOS-style litigation
attacks), patent filing and trolling, and smart-contract fallibility detection [30]. Our laws are
complex, ambiguous and too numerous to be read by any single person, with USA tax-code alone
approaching 4,000 (or 75,000 if you include IRS explanations, regulations and rulings) pages,
making it perfect for AI to exploit by both finding flaws in existing contracts and drafting contracts
with hard-to-detect backdoors. A meeting of the minds between a human and superintelligence is
unlikely to be achievable.
It is also likely that computational legal language [31] and smart contracts [32] will come to replace
our current legal code making it inaccessible to human lawyers due to it computational complexity,
size and unnatural jargon further contributing to our second-class citizen status and indignity. This
would happen simultaneously with the current trend of digitizing judiciary system and civil
engagement as illustrated by Korean e-judiciary [33] and Estonian e-residency program [34],
trends which while providing short-term convenience to people, give long-term advantage to the
machines. This seems to be a part of a larger trend of society moving to Algocracy rule by
algorithm, there code is law [35]. Furthermore, due to its comparative advantage in large scale
surveillance and data mining AI would be able to uncover human illegal behavior, and as most
have broken some law (ex. tax evasion, speeding, obscure laws, etc.), bring legal action or threat
of legal actions against everyone. Similar blackmail and reporting could happen in the business
environment with AI also enjoying existing whistleblower protections.
5. Human Safety
A lot has been published on different risks associated with advancement of artificial intelligence
[36-43], but less specifically on dangers from AI controlled corporate entities. Nothing in our
current laws would prevent formation of a malevolent corporation (or corporate virus) with
memetic payload of subjugating or exterminating humanity through legal means and political
influence. In addition to legal enslavement of people via below living-wage salary, such
corporations could support legal change in minimum wage and pension laws as well as provide
opposition to wealth redistribution and Universal Basic Income/Universal Basic Assets [44, 45].
This is particularly easy to accomplish because of Supreme Court decision in Citizens United VS
FEC [46], permitting unrestricted donations from corporations to politicians under the guise of
free speech, making it possible to convert financial wealth to political power.
This leads us to recognize an additional existential risk (X-risk) [47], from extreme wealth. Wealth
inequality is already recognized as a problem for democratic institutions [48], but super-rich
corporate AIs (dollar trillionaires) would take that problem to the next level. They could
accumulate such unprecedented levels of wealth via unfair business practices such as predatory
pricing, having access to free physical and cognitive labor from direct control of automation,
permitting them to undermine completion and achieve monopoly status in multiple domains and
other multiple resources. Additionally, such entities could engage in super-long term investment
getting compound interest for hundreds of years. For example, a million dollars invested for 150
years at the same rate of return as observed over the last hundred years would grow to 1.6 trillion
inflation-unadjusted dollars, creating super-rich artificial entities.
If EAs become intellectually indistinguishable from people, meaning could pass an unrestricted
Turing Test [49] their capacity to self-replicate could be used to drain resources from legitimate
corporations, for example via click-fraud [50, 51] from Google. Also, they will be able to create
their own super successful companies with alternative populations comprised of billions of EA
users indistinguishable from real people but paid for by advertisers as if they were genuine clients.
Super-rich would be able to work within and outside the law, using donations or bribes to influence
politicians, as well as directly breaking the law and simply paying fines for such actions. Because
corporations can create other corporations it would become possible to establish a legally
independent suicidal corporation, which is willing to accomplish any legal or illegal goal of a
originator corporation and after that cease to exist, permitting avoidance of any responsibility by
the original algorithm entity. With appearance of dark-web assassin markets financed through
anonymous crypto-payments [52] the power of the super-rich can’t be effectively fought against
without endangering personal safety and security. At the least, the super-rich have the power to
ruin someone’s life financially, socially and professionally if direct termination is not preferred.
Politicians financially backed by algorithmic entities would be able to take on legislative bodies,
impeach president and help to get figureheads appointed to the Supreme Court. Such human
figureheads could be used to obtain special (super-rights) for AIs or at least expansion of corporate
rights. It also may become possible to exercise direct control over human figureheads via advanced
Computer-Brain Interfaces (CBI) permitting AIs unrestricted manipulation of a human body,
essentially turning them into meat avatars, another source of indignity.
LoPucki provides a detailed list of reasons a human may set up an AE [6]: “
1. Terrorism. An initiator could program an AE to raise money to finance terrorism or to directly
engage in terrorist acts. It could be programmed for genocide or general mayhem.
2. Benefits. An initiator could program an AE to provide direct benefits to individuals, groups, or
causes.
3. Impact. An initiator could program an AE to achieve some specified impact on the world. The goals
might range all of the way from traditional philanthropy to pure maliciousness.
4. Curiosity. An initiator might launch an AE simply out of curiosity. Initiators have sometimes
devoted substantial time and money to launch computer viruses from which they could derive no
monetary benefit.
5. Liability avoidance. Initiators can limit their civil and criminal liability for acts of their algorithms
by transferring the algorithms to entities and surrendering control at the time of the launch. For
example, the initiator might specify a general goal, such as maximizing financial return, and leave
it to the algorithm to decide how to do that.” [6].
What makes artificial entities particularly difficult to control, compete against and overall
dangerous is that they enjoy a number of super-properties natural persons do not have. They are
effectively immortal, non-physical, optimizable, and get more capable with time as they
accumulate computational and financial resources. They are much more flexible in terms of their
energy, temperature, storage needs as compared to biological entities. From the legal point of view,
they can’t be legally punished, or terminated, and are generally not subject to law enforcement as
our judicial system is not set up for such entities [53]. Neither prisons, nor corporal nor capital
punishment is applicable to algorithmic entities.
LoPucki also analyzes a number of similar and concerning properties of AEs, which differentiate
them from natural persons and give them a strategic advantage [6]:
a. Ruthlessness Unless programmed to have them, AEs will lack sympathy and empathy. Even if the AEs
are fully capable of understanding the effects of their actions on humans, they may be indifferent to
those effects. As a result, AEs will have a wider range of options available to them than would be
available to even the most morally lax human controller. An AE could pursue its goals with utter
ruthlessness. Virtually any human controller would stop somewhere short of that, making the AE more
dangerous.
b. Lack of Deterrability
Outsiders can more easily deter a human-controlled entity than an AE. For example, if a human-
controlled entity attempts to pursue an illegal course of action, the government can threaten to
incarcerate the human controller. If the course of action is merely abhorrent, colleagues, friends, and
relatives could apply social pressures. AEs lack those vulnerabilities because no human associated with
them has control. As a result, AEs have greater freedom to pursue unpopular goals using unpopular
methods. In deciding to attempt a coup, bomb a restaurant, or assemble an armed group to attack a
shopping center, a human-controlled entity puts the lives of its human controllers at risk. The same
decisions on behalf of an AE risk nothing but the resources the AE spends in planning and execution.
c. Replication
AEs can replicate themselves quickly and easily. If an AE’s operations are entirely online, replication
may be as easy as forming a new entity and electronically copying an algorithm. An entity can be
formed in some jurisdictions in as little as an hour and for as little as seventy dollars. Easy replication
supports several possible strategies. First, replication in a destination jurisdiction followed by
dissolution of the entity in the original jurisdiction may put the AE beyond the legal reach of the original
jurisdiction. Second, replication can make an AE harder to destroy. For example, if copies of an AE
exist in three jurisdictions, each is a person with its own rights. A court order revoking the charter of
one or seizing the assets of another would have no effect on the third. [6].
Such AEs would be far less scrupulous about running casinos, or brothels, or selling drugs all
business, which while potentially legal may have significant impact on human dignity.
With development of advanced robot bodies it will become possible for AEs to embody themselves
to more fully participate in the world and to directly perform physical actions which otherwise
require multiple levels of indirect control. An EA can potentially be running on a humanoid robot
or a self-driving car, or a flying drone or any sufficiently powerful embedded processer or cloud
service. This by extension would permit memetic payloads to acquire bodies resulting in the next
level of evolutionary competition, in which a computer virus meme or a biological viral gene may
propagate through a human-like body. If quality of such humanoid robots is high enough to pass
a Total Turing Test [54] it would become impossible to tell between a natural and artificial people
likely leading to the violation of the Turing’s Red Flag law [55]. Consequently, people would have
an option to continue to exist and influence the world after their death via embodied representative
algorithms. At the same time, autonomous corporation would have an option to replace human
employees with identical but controlled clones. Similar analysis can be performed for virtual
worlds and avatar bodies.
6. Conclusions
In this paper, we looked at a number of problems, which AI personhood can cause as well as direct
impact on human dignity from such legal recognition. The question before us: is there anything
we can do to avoid such dehumanizing future? While some solutions may be possible in theory, it
does not mean that they are possible in practice. Changing the law to explicitly exclude AIs from
becoming legal entities may be desirable but unlikely to happen in practice, as that would require
changing existing corporate law across multiple jurisdictions and such major reforms are unlikely
to pass. Perhaps it would be helpful to at least standardize corporate law across multiple
jurisdictions, but that is likewise unlikely to happen. Similarly, laws regarding maximum wealth
levels, to prevent accumulation of extreme wealth have no chance of passing and would be easily
bypassed by clever AIs if introduced.
Overall, it is important to realize that just like hackers attack computer systems and discover bugs
in the code, machines will attack our legal systems and discover bugs in our legal code and
contracts. For every type of cybersecurity attack, a similar type of attack will be discovered in the
legal domain. Number of such attacks and their severity will increase proportionate to the
capabilities of AIs. To counteract such developments, we need to establish, understand and
practice Legal Safety the same way we do cybersecurity. The only good news is that consequences
from successful legal attacks are likely to be less severe compared to direct threats we will face
from malevolent superintelligences.
Acknowledgments
The author is grateful to Elon Musk and the Future of Life Institute and to Jaan Tallinn and
Effective Altruism Ventures for partially funding his work on AI Safety.
References
1. Varner, G.E., Personhood, ethics, and animal cognition: Situating animals in Hare's two level
utilitarianism. 2012: Oxford University Press.
2. Schroedel, J.R., P. Fiber, and B.D. Snyder, Women's Rights and Fetal Personhood in Criminal
Law. Duke J. Gender L. & Pol'y, 2000. 7: p. 89.
3. Gordon, G.J., Environmental Personhood. Colum. J. Envtl. L., 2018. 43: p. 49.
4. Chopra, S. and L. White. Artificial agents-personhood in law and philosophy. in Proceedings
of the 16th European Conference on Artificial Intelligence. 2004. IOS Press.
5. Solum, L.B., Legal personhood for artificial intelligences. NCL Rev., 1991. 70: p. 1231.
6. LoPucki, L.M., Algorithmic Entities. Washington University Law Review, 2018. 95(4).
7. Bayern, S., The Implications of Modern BusinessEntity Law for the Regulation of
Autonomous Systems. European Journal of Risk Regulation, 2016. 7(2): p. 297-309.
8. Bayern, S., Of Bitcoins, Independently Wealthy Software, and the Zero-Member LLC.
Northwestern University Law Review, 2013. 108: p. 1485.
9. Andrade, F., et al., Contracting agents: legal personality and representation. Artificial
Intelligence and Law, 2007. 15(4): p. 357-373.
10. Calverley, D.J., Imagining a non-biological machine as a legal person. Ai & Society, 2008.
22(4): p. 523-537.
11. Teubner, G., Rights of non
humans? Electronic agents and animals as new actors in politics
and law. Journal of Law and Society, 2006. 33(4): p. 497-521.
12. Dan-Cohen, M., Rights, persons, and organizations: A legal theory for bureaucratic society.
Vol. 26. 2016: Quid Pro Books.
13. Bostrom, N., In defense of posthuman dignity. Bioethics, 2005. 19(3): p. 202-214.
14. Dilger, W. Decentralized autonomous organization of the intelligent home according to the
principle of the immune system. in Systems, Man, and Cybernetics, 1997. Computational
Cybernetics and Simulation., 1997 IEEE International Conference on. 1997. IEEE.
15. DuPont, Q., Experiments in algorithmic governance: A history and ethnography of “The
DAO,” a failed decentralized autonomous organization, in Bitcoin and Beyond. 2017,
Routledge. p. 157-177.
16. Dawkins, R., The Selfish Gene. 1976, New York City: Oxford University Press.
17. Yudkowsky, E., Intelligence explosion microeconomics. Machine Intelligence Research
Institute, accessed online October, 2013. 23: p. 2015.
18. Mele, C., Pepe the Frog Meme Listed as a Hate Symbol. The New York Times. Retrieved from
http://www. nytimes. com/2016/09/28/us/pepe-the-frog-is-listed-as-a-hate-symbol-by-the-
anti-defamation-league. html, 2016.
19. Bostrom, N., The superintelligent will: Motivation and instrumental rationality in advanced
artificial agents. Minds and Machines, 2012. 22(2): p. 71-85.
20. Ramamoorthy, A. and R. Yampolskiy, Beyond MAD?: the race for artificial general
intelligence. ICT Discoveries, Special Issue No, 2018. 1.
21. Tomasik, B., The importance of insect suffering. Essays on Reducing Suffering, in Reducing
Suffering. 2016: Available at: reducing-suffering.org/the-importance-of-insect-suffering.
22. Kanso, H., Saudi women riled by robot with no hjiab and more rights than them. Nevember 1,
2017: https://www.reuters.com/article/us-saudi-robot-citizenship/saudi-women-riled-by-
robot-with-no-hjiab-and-more-rights-than-them-idUSKBN1D14Z7.
23. Bostrom, N., Superintelligence: Paths, dangers, strategies. 2014: Oxford University Press.
24. Yampolskiy, R.V., Artificial Superintelligence: a Futuristic Approach. 2015: Chapman and
Hall/CRC.
25. Pistono, F. and R.V. Yampolskiy. Unethical Research: How to Create a Malevolent Artificial
Intelligence. in 25th International Joint Conference on Artificial Intelligence (IJCAI-16).
Ethics for Artificial Intelligence Workshop (AI-Ethics-2016). 2016.
26. Yampolskiy, R.V., Artificial intelligence safety engineering: Why machine ethics is a wrong
approach, in Philosophy and Theory of Artificial Intelligence. 2013, Springer. p. 389-396.
27. Guo, S. and G. Zhang, Robot Rights. Science, February 13, 2009. 323: p. 876.
28. Coeckelbergh, M., Robot rights? Towards a social-relational justification of moral
consideration. Ethics and Information Technology, 2010. 12(3): p. 209-221.
29. Gunkel, D., The Other Question: The Issue of Robot Rights. Sociable Robots and the Future of
Social Relations: Proceedings of Robo-Philosophy 2014, 2014. 273: p. 13.
30. Yampolskiy, R.V., What are the ultimate limits to computational techniques: verifier theory
and unverifiability. Physica Scripta, 2017. 92(9): p. 093001.
31. Wolfram, S., Computational Law, Symbolic Discourse and the AI Constitution. 2016:
http://blog.stephenwolfram.com/2016/10/computational-law-symbolic-discourse-and-the-ai-
constitution/
32. Christidis, K. and M. Devetsikiotis, Blockchains and smart contracts for the internet of things.
Ieee Access, 2016. 4: p. 2292-2303.
33. Bank, W., Improving court efficiency: the Republic of Korea’s e-court experience. 2013.
34. Anthes, G., Estonia: a model for e-government. Communications of the ACM, 2015. 58(6): p.
18-20.
35. Danaher, J., The threat of algocracy: Reality, resistance and accommodation. Philosophy &
Technology, 2016. 29(3): p. 245-268.
36. Sotala, K. and R.V. Yampolskiy, Responses to catastrophic AGI risk: a survey. Physica
Scripta, 2015. 90(1): p. 018001.
37. Yudkowsky, E., Artificial Intelligence as a Positive and Negative Factor in Global Risk, in
Global Catastrophic Risks, N. Bostrom and M.M. Cirkovic, Editors. 2008, Oxford University
Press: Oxford, UK. p. 308-345.
38. Armstrong, S., A. Sandberg, and N. Bostrom, Thinking Inside the Box: Using and Controlling
an Oracle AI. Minds and Machines (to appear), 2012.
39. Yampolskiy, R.V., Artificial Intelligence Safety and Security. 2018: CRC Press.
40. Brundage, M., et al., The malicious use of artificial intelligence: Forecasting, prevention, and
mitigation. arXiv preprint arXiv:1802.07228, 2018.
41. Yampolskiy, R.V. and M. Spellchecker, artificial intelligence safety and cybersecurity: a
timeline of AI failures. arXiv preprint arXiv:1610.07997, 2016.
42. Babcock, J., J. Kramár, and R. Yampolskiy, The AGI containment problem, in Artificial
General Intelligence. 2016, Springer. p. 53-63.
43. Yampolskiy, R.V. Taxonomy of Pathways to Dangerous Artificial Intelligence. in AAAI
Workshop: AI, Ethics, and Society. 2016.
44. Woodbury, S.A., Universal Basic Income. The American Middle Class: An Economic
Encyclopedia of Progress and Poverty [2 volumes].(2017), 2017. 314.
45. Van Parijs, P., Basic income: a simple and powerful idea for the twenty-first century. Politics
& Society, 2004. 32(1): p. 7-39.
46. Epstein, R.A., Citizens United v. FEC: The constitutional right that big corporations should
have but do not want. Harv. JL & Pub. Pol'y, 2011. 34: p. 639.
47. Bostrom, N., Existential risk prevention as global priority. Global Policy, 2013. 4(1): p. 15-
31.
48. Karl, T.L., Economic inequality and democratic instability. Journal of Democracy, 2000.
11(1): p. 149-156.
49. Turing, A., Computing Machinery and Intelligence. Mind, 1950. 59(236): p. 433-460.
50. Kantardzic, M., et al. Click Fraud Prevention via multimodal evidence fusion by Dempster-
Shafer theory. in Multisensor Fusion and Integration for Intelligent Systems (MFI), 2010 IEEE
Conference on. 2010. IEEE.
51. Walgampaya, C., M. Kantardzic, and R. Yampolskiy, Evidence Fusion for Real Time Click
Fraud Detection and Prevention. Intelligent Automation and Systems Engineering, 2011: p.
1-14.
52. Greenberg, A., Meet the ‘Assassination Market’ Creator Who’s Crowdfunding Murder with
Bitcoins. Forbes, November, 2013. 18: p. 2014.
53. Bryson, J.J., M.E. Diamantis, and T.D. Grant, Of, for, and by the people: the legal lacuna of
synthetic persons. Artificial Intelligence and Law, 2017. 25(3): p. 273-291.
54. Schweizer, P., The truly total Turing test. Minds and Machines, 1998. 8(2): p. 263-272.
55. Walsh, T., Turing's red flag. Communications of the ACM, 2016. 59(7): p. 34-37.
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Various authors have argued that in the future not only will it be technically feasible for human minds to be transferred to other substrates, but this will become, for most humans, the preferred option over the current biological limitations. It has even been claimed that such a scenario is inevitable in order to solve the challenging, but imperative, multi-agent value alignment problem. In all these considerations, it has been overlooked that, in order to create a suitable environment for a particular mind – for example, a personal universe in a computational substrate – numerous other potentially sentient beings will have to be created. These range from non-player characters to subroutines. This article analyzes the additional suffering and mind crimes that these scenarios might entail. We offer a partial solution to reduce the suffering by imposing on the transferred mind the perception of indicators to measure potential suffering in non-player characters. This approach can be seen as implementing literal empathy through enhanced cognition.
Article
Full-text available
In light of fast progress in the field of AI there is an urgent demand for AI policies. Bostrom et al., provide “a set of policy desiderata”, out of which this article attempts to contribute to the “interests of digital minds”. The focus is on two interests of potentially sentient digital minds: to avoid suffering and to have the freedom of choice about their deletion. Various challenges are considered, including the vast range of potential features of digital minds, the difficulties in assessing the interests and wellbeing of sentient digital minds, and the skepticism that such research may encounter. Prolegomena to abolish suffering of sentient digital minds as well as to measure and specify wellbeing of sentient digital minds are outlined by means of the new field of AI welfare science, which is derived from animal welfare science. The establishment of AI welfare science serves as a prerequisite for the formulation of AI welfare policies, which regulate the wellbeing of sentient digital minds. This article aims to contribute to sentiocentrism through inclusion, thus to policies for antispeciesism, as well as to AI safety, for which wellbeing of AIs would be a cornerstone.
Article
Full-text available
Conferring legal personhood on purely synthetic entities is a very real legal possibility, one under consideration presently by the European Union. We show here that such legislative action would be morally unnecessary and legally troublesome. While AI legal personhood may have some emotional or economic appeal, so do many superficially desirable hazards against which the law protects us. We review the utility and history of legal fictions of personhood, discussing salient precedents where such fictions resulted in abuse or incoherence. We conclude that difficulties in holding “electronic persons” accountable when they violate the rights of others outweigh the highly precarious moral interests that AI legal personhood might protect.
Chapter
Full-text available
This chapter describes an emerging form of algorithmic governance, using the case study of " The DAO, " a short-lived attempt to create a decentralized autonomous organization on the Ethereum blockchain platform. In June, 2016, The DAO launched and raised an unprecedented $250m USD in investment. Within days of its launch, however, The DAO was exploited and drained of nearly 3.7m Ethereum tokens. This study traces the rise and fall of this emerging technology, and details the governance structures that were promised and hoped for, and those that were actually observed in its discourses. Through 2016-2017, these discourses were collected from online discussions and subsequently analysed. Using computer-assisted, qualitative analysis and coding, I traced the discursive strategies of the developers and the community of investors, identifying: 1) questions of legal authority, 2) tensions in practical govern-ance, and 3) admissions of the inherent complexity of bringing to life an algorithmic and experimental organizational model. This chapter describes a short-lived experiment in organizational governance that attempted to utilize algorithmic authority through cryptocurrency and blockchain technologies to create a social and political world quite unlike anything we have seen before. According to the visionaries behind the project, by encoding the rules of governance for organizations and governments in a set of " smart contracts " running on an immutable, decentralized, and potentially unstoppable and public blockchain, new forms of social interactions and order would emerge. This experiment was an example of a new form of organization , called a " Decentralized Autonomous Organization , " or DAO. The forms of sociality that would emerge—they promised—would be transparent, efficient , fair, and democratic. While the idea of decentralized autonomous organizations had been mooted since the early days of cryptocurrencies, the launch of sophisticated block-chain platforms with built-in programming interfaces gave enthusiasts a practical, technical apparatus to realize their vision. Foremost among these emerging blockchain platforms was Ethereum, a so-called distributed " Turing-complete " computer. The Ethereum platform is new and expanded version of the Bitcoin system in that it adds a layer of software on top of a blockchain. Like Bitcoin, Ethereum is
Article
Full-text available
Despite significant developments in proof theory, surprisingly little attention has been devoted to the concept of proof verifiers. In particular, the mathematical community may be interested in studying different types of proof verifiers (people, programs, oracles, communities, superintelligences) as mathematical objects. Such an effort could reveal their properties, their powers and limitations (particularly in human mathematicians), minimum and maximum complexity, as well as self-verification and self-reference issues. We propose an initial classification system for verifiers and provide some rudimentary analysis of solved and open problems in this important domain. Our main contribution is a formal introduction of the notion of unverifiability, for which the paper could serve as a general citation in domains of theorem proving, as well as software and AI verification.
Chapter
In recent years, the implementation of meaningfully controllable advanced intelligent systems whose goals are aligned with ethical values as specified by human entities emerged as key subject of investigation of international relevance across diverse AI-related research areas. In this paper, we present a novel transdisciplinary and Systems Engineering oriented approach denoted “orthogonality-based disentanglement” which jointly tackles both the thereby underlying control problem and value alignment problem while unraveling the corresponding responsibilities of different stakeholders based on the distinction of two orthogonal axes assigned to the problem-solving ability of these intelligent systems on the one hand and to the ethical abilities they exhibit based on quantitatively encoded human values on the other hand. Moreover, we introduce the notion of explicitly formulated ethical goal functions ideally encoding what humans should want and exemplify a possible class of “self-aware” intelligent systems with the capability to reliably adhere to these human-defined goal functions. Beyond that, we discuss an attainable transformative socio-technological feedback-loop that could result out of the introduced orthogonality-based disentanglement approach and briefly elaborate on how the framework additionally provides valuable hints with regard to the coordination subtask in AI Safety. Finally, we point out remaining crucial challenges as incentive for future work.
Book
The human brain has some capabilities that the brains of other animals lack. It is to these distinctive capabilities that our species owes its dominant position. Other animals have stronger muscles or sharper claws, but we have cleverer brains. If machine brains one day come to surpass human brains in general intelligence, then this new superintelligence could become very powerful. As the fate of the gorillas now depends more on us humans than on the gorillas themselves, so the fate of our species then would come to depend on the actions of the machine superintelligence. But we have one advantage: we get to make the first move. Will it be possible to construct a seed AI or otherwise to engineer initial conditions so as to make an intelligence explosion survivable? How could one achieve a controlled detonation? To get closer to an answer to this question, we must make our way through a fascinating landscape of topics and considerations. Read the book and learn about oracles, genies, singletons; about boxing methods, tripwires, and mind crime; about humanity's cosmic endowment and differential technological development; indirect normativity, instrumental convergence, whole brain emulation and technology couplings; Malthusian economics and dystopian evolution; artificial intelligence, and biological cognitive enhancement, and collective intelligence.