Ethical issues in moral and social enhancement

CHAPTER 7. Freedom, autonomy and the God Machine 
7.1. 
Introduction 
Proposals suggesting the use of biomedical emotion modulation for achieving better 
moral outcomes (Persson and Savulescu, 2008) were met with criticism (e.g. Harris 
2010, 2014a, 2014b; Sparrow, 2014). An important objection originally raised by 
Harris (2010) highlights the potential negative impact of MB on freedom. Harris 
argues that even if the interventions Persson and Savulescu (2008) propose were to 
be effective (i.e. achieve the behavioural end), they would come at an unacceptable 
cost to the kind of freedom that is the foundation of moral agency while grounding 
moral responsibility. In response to Harris and other critics, Savulescu and Persson 
(2012a) develop a thought experiment in which ‘the most powerful bioquantum 
computer’, which they nickname ‘the God Machine’, uses direct modulation of 
intentions to prevent citizens from doing great evils. The thought experiment 
purports to demonstrate a technological intervention that, while it directly modifies 
the roots of morally relevant actions, is still desirable.
In this chapter, I look closer at the God Machine and its impact on moral life and 
freedom, while examining the criticisms raised by Harris (2014a) and Sparrow 
(2014). I will examine the God Machine thought experiment as well as introduce a 
several similar thought experiments in order to tease out what is an important 
problem giving raise to an ethical worry in the God Machine scenario. I will use the 
Moral Luck Machine thought experiment in order to argue that there is a relevant 
difference for moral agency between the two scenarios: Moral Luck Machine 
allows for greater engagement with reasons by the agent. I will use the Rational 
Persuader Machine thought experiment to further support that point. It seems that 
even though an outcome (agents actions in the world) do not change, there is a 
difference in how we arrived at the outcome. Finally, I will use Halls Brian Implant 
thought experiment to provide an example closer to the real world, to argue that it is 
not overdetermination of agent’s actions (resulting in the comparable outcomes in 
the world) that is the main problem raised by the God Machine thought experiment.
Instead, I will argue that the most significant impact of the God Machine’s 
intervention lies in the fact that it diminishes the agent's appropriate engagement 

---

134 
with reasons and freedom of thought, rather than the fact that it has an impact on 
freedom of action or on freedom from political domination. 
The analysis presented in this chapter is aimed at a) addressing the issues raised by 
God Machine thought experiment, the interpretation of which has been used as a 
basis for arguing both for and against the desirability of MB and, b) using the 
analysis to tease out the important factors influencing moral agency and freedom in 
cases of overdetermination. However, as I argue in the last section of this chapter, 
the utility of those arguments in discussing the real-world MB is limited. Similarly 
to other instances of using this philosophical method in discussions regarding real-
world applications of technology, thought experiments of this level of abstraction 
are useful tools in teasing out the aspects that are relevant, drawing attention to 
some aspects of reality and helping us in building conceptualizations that then can 
be of help when applied to real-world cases. However, those gains do not translate 
straightforwardly into moral assessment of real-world interventions. Thus, the 
analysis in this chapter aims at providing additional conceptual clarity and 
providing further support for the importance of engagement with reasons for moral 
agency but not as an argument about real world moral enhancement. Chapter 8 will 
address this limitation and build on the analysis presented in this chapter.
After introducing Savulescu and Persson’s (2012a) thought experiment in section 
7.2, section 7.3 examines the God Machine’s impact on ascription of praise and 
blame. I conclude that the problems the God Machine brings to the ascription of 
moral praise and blame do not amount to a serious argument against its desirability. 
In section 7.4, I analyse the God Machine’s impact in light of the distinctions 
proposed by Frankfurt and argue that the God Machine preserves freedom of action 
and even free will, but affects the ability to form ‘a will of one’s own’ and impacts 
freedom of thought. Section 7.5 responds to Harris’ argument in which he sees the 
problematic impact of the God Machine as relating to the divorcing of thought and 
action: ‘Decisions to no effect are pointless from the moral perspective; for what is 
a good state of mind worth, if it makes no difference to the world?’ (2014a, p. 249)
In response, I argue that the agent’s lack of appropriate engagement with reasons is 
the crucial factor, rather than the fact of changing the behaviour – that the impact 
for freedom of action related to overdetermination and the ‘direct’ mode of 
changing behaviour is less important than the lack of awareness and control of the 

---

135 
agent over those changes and that a good state of mind is worth quite a lot. It may 
even ground meaningful ‘alternative possibilities’.
Section 7.6 examines Sparrow’s (2014) argument according to which the God 
Machine scenario is undesirable because the God Machine dominates its subjects. I 
argue that Sparrow’s argument depends on an inappropriate personification of the 
God Machine and that the God Machine is better understood as an analogy for the 
law. Pettit’s (1997) theory of freedom as non-domination is an unfortunate choice 
as the basis of Sparrow’s (2014) critique because it focuses on the relation of 
persons
and is limited in the analysis of the freedom-impairing impacts of 
structures. Insofar as we treat the God Machine as a structure similar to law, there is 
no strong reason to suppose that it would enable the exercise of others' arbitrary and 
unchecked power in 
the idealized conditions
of the thought experiment.
However, as I argue in section 7.8, worries about the arbitrary and unchecked use of 
power become warranted if we consider the application of the God Machine in a 
possible world more like ours. Moreover, Savulescu and Persson’s (2012a) appeal 
to Mill’s harm principle is weak because it relies on a misinterpretation of the 
application of the harm principle. In fact, the harm principle would provide a 
stronger argument 
against
the God Machine’s changing people’s intentions. 
Moreover, the spirit of Mill’s use of the harm principle is to protect the individual’s 
private sphere from state and societal encroachment, and there is nothing more 
central and private than the ‘inner citadel’ of our thoughts. 
In the final section of this chapter I outline the limitations of the application of the 
God Machine to the analysis of the impact of MB on moral agency.