“A Logical Argument for Faith In God”

The New God Argument [1] (NGA) purports to be a logical argument for faith in God. It was created by Lincoln Cannon, the founder of The Mormon Transhumanist Association.

The NGA owes much to a philosophical treatise known as The Simulation Argument [2] (SA). This argument has been popularly construed to suggest that we may be living in a computer simulation. It’s important to have a basic understanding of SA before we can fully and fairly analyze NGA, which leverages SA for foundation, assumptions and structure.

The Simulation Argument

From Wikipedia [3]:
In 2003, philosopher Nick Bostrom proposed a trilemma that he called “the simulation argument”. Despite the name, Bostrom’s “simulation argument” does not directly argue that we live in a simulation; instead, Bostrom’s trilemma argues that one of three unlikelyseeming propositions must be true. The trilemma points out that a technologically mature “posthuman” civilization would have enormous computing power; if even a tiny percentage of them were to run “ancestor simulations” (that is, “high-fidelity” simulations of ancestral life that would be indistinguishable from reality to the simulated ancestor), the total number of simulated ancestors, or “Sims”, in the universe (or multiverse, if it exists) would greatly exceed the total number of actual ancestors. Therefore, at least one of the following three
propositions is almost certainly true:

  1. “The fraction of human-level civilizations that reach a posthuman stage (that is, one capable of running high-fidelity ancestor simulations) is very close to zero”, or
  2. “The fraction of posthuman civilizations that are interested in running ancestorsimulations is very close to zero”, or
  3. “The fraction of all people with our kind of experiences that are living in a simulation is very close to one”

Bostrom is careful in how he discusses SA; he emphasizes that it does not prove that we are living in a simulation. It presents three alternatives, one of which is almost certainly true. Further, Bostrom does not claim to be able to assign any specific probabilities to the three disjunctives. When asked whether he believes we are living in such a simulation, he answers no. He has at various times given option (3) (that we are living in a computer simulation) a 20%
chance; but again, this is not based in critical analysis, but rather, a best guess, a qualification Bostrom frequently makes.

There is some debate as to what type of argument SA actually is. In a paper titled “A Critical Engagement of Bostrom’s Computer Simulation Hypothesis 4, author Norman Swazo, says:

“there is little reason to take Bostrom’s proposal as a scientific hypothesis, in which case it remains at best a philosophically speculative proposition only.”

Bostrom doesn’t seem to counter this.

This does not mean, however, that SA is simplistic; on the contrary, the formulations and arguments that underlie each proposition seem well reasoned and rigorous. Rooted in probability theory, and documented in mathematical notation, Bostrom and his supporters seem confident that each individual proposition in SA is not only valid, but sound.

The New God Argument: An Overview

Author Lincoln Cannon summarizes NGA:

“The New God Argument is a logical argument for faith in God. Given assumptions consistent with contemporary science and technological trends, the argument proves that if we trust in our own superhuman potential then we should also trust that superhumanity probably would be more compassionate than we are and created our world. Because a compassionate creator may qualify as God in some religions, trust in our own superhuman potential may entail faith in God, and atheism may entail distrust in our superhuman potential.”

Whereas Bostrom tends to tamp down any assertion of truth claims derived from SA, NGA makes strong and novel truth claims. “Given assumptions consistent with science and technology,” NGA asserts the following 8 Truth Claims:

  • TC1: It is a logical argument.
  • TC2: It is a logical argument for faith in God.
  • TC3: It will only use assumptions consistent with science and technology.
  • TC4: It proves that superhumanity is probably more compassionate than we are.
  • TC5: It proves that superhumanity created our world.
  • TC6: God = superhumanity + compassion + creation
  • TC7: Human potential for advancement may require faith in this God.
  • TC8: Atheists may be required to distrust human potential.

The intent of this analysis is to explore these claims, to analyze the soundness of the arguments, the reasonableness of the assumptions, the coherence of the premises and the validity of the conclusions. As we progress through each sub-proposition of NGA (F1, CO, CR, G1) I will discuss how each interact with the 8 Truth Claims.

Summary of NGA and Its Arguments

NGA is comprised of the following propositions:

  • The Faith Assumption (F1)
    • 1 Assumption
  • The Compassion Argument (CO)
    • 2 Assumptions, 1 Conclusion
  • The Creation Argument (CR)
    • 2 Assumptions, 1 Conclusion
  • The God Conclusion (G1)
    • 1 Conclusion (final)

NGA holds that we are living in something like a simulation of the lives of previous beings, that those beings are God, and that we are justified, even required, to have faith in those Gods.

Notes on Truth Claim 1: NGA is a Logical Argument

It’s important to understand what NGA claims to be so as to judge it through the proper lens. By claiming to be a logical argument, it must be structured according to rules of logical argumentation. SA, for example, seems not to make claims about being a logical argument. In contrast to NGA, SA does not assert more than very general claims; it makes no truth claims beyond that one of its three propositions may obtain. This is notable due in part to the
foundational reliance that NGA has on SA. And yet, they make significantly different types of claims.

Logical arguments are primarily of two kinds: deductive and inductive. A valid argument has on or more premises followed by a single conclusion.

Deductive Arguments

A deductive argument asserts that the truth of the conclusion is a logical consequence of the premises. If the premises are true, the conclusion must follow with certainty.

Deductive Example:

1. All men are mortal.
2. Socrates is a man.
3. Therefore, Socrates is mortal.

Deductive arguments can be:

  • Valid: premises are assumed to be true and the conclusion follows
    • Sound: premises are actually true and the conclusion follows
    • Unsound: premises are false even though the conclusion follows
  • Invalid: premises are assumed to be true and the conclusion does not follow
    • Unsound: premises are false and the conclusion does not follow

Inductive Arguments

An inductive argument asserts that the truth of the conclusion is supported to a high degree of probability by the premises. Inductive arguments require you to know the premises are true, but the conclusion is only probabilistic (with a high degree of certainty). An inductive argument is said to be cogent if and only if the truth of the argument’s premises would render the truth of the conclusion highly probable, and the argument’s premises are, in fact, true.

Inductive Example:

1. All biological life forms that we know of depend on water to exist.
2. Therefore, if we discover a new biological life form it will probably depend on water to exist.

Inductive arguments can be:

  • Strong: premises are assumed true, conclusion is probable.
    • cogent: premises are actually true, conclusion is probable.
    • uncogent: premises are false
  • Weak: premises are assumed true, conclusion is improbable.
    • uncogent: premises are false

The Faith Assumption (F1)

NGA starts by asking us to allow for an assumption that it terms “The Faith Assumption.” It is summarized this way:

The Faith Assumption is a proposition that humanity will not become extinct before evolving into superhumanity. The proposition may be false. However, to the extent we do not know it to be false, we may have practical or moral reasons to trust that it is true. In any case, the Faith Assumption is a common aspiration among secular advocates of technological evolution, and it may be consistent with the religious doctrine of theosis, also known as divinization or deification: the idea that humanity should become God.

The NGA Faith Assumption (F1) is:

F1: humanity will not become extinct before evolving into superhumanity

Note that F1 of NGA is the negation of disjunctive (1) of SA, which says:

(1) the human species is very likely to go extinct before reaching a “posthuman” stage;

With F1, NGA introduces its first premise in the structuring of the logical argument. It does so by asking us to assume that disjunctive (1) of SA obtains. This is no small ask. Recall that SA makes no claims as to the probability of which of its three disjuncts will obtain. Bostrom recognizes the explicit limits of SA to draw conclusions; the argument simply isn’t capable of making that leap. There are attempts to calculate the likelihood that humanity either will or will not go extinct. One well known argument is The Doomsday Argument(5) which claims there is a 95% chance of extinction after 9,120 years of humans on earth. Back to Bostrom(6):

“we must give a high credence to DOOM, the hypothesis that humankind will go extinct before reaching a posthuman level” But extinction is only one part of F1 that we are asked to assume; the other part is that humanity will obtain the state of Superhumanity (note here that Cannon substitutes the word Superhumanity in place of the word Posthuman used in SA). Again, Bostrom(7): Proposition (1) doesn’t by itself imply that we are likely to go extinct soon, only that we are unlikely to reach a posthuman stage. This possibility is compatible with us remaining at, or somewhat above, our current level of technological development for a long time before going extinct. Another way for (1) to be true is if it is likely that technological civilization will collapse. Primitive human societies might then remain on Earth indefinitely.

I mention this to impress upon the reader the magnitude and the complexity of what Cannon is asking us to assume in F1. In the introductory text of F1, Cannon makes an appeal as to why it may be practical for humans to have faith in F1. I’ll note here that there may be a problem of circularity in that F1, The Faith Assumption, is asking us to assume “faith” in a proposition which may be core to the conclusion of the argument; that we should have “faith” in God.

Additionally, There is an important distinction between how Bostrom discusses SA and how Cannon discusses NGA. Bostrom, when referring to humanity, is not necessarily speaking about “us” on this earth at this time. He’s speaking in more general terms about the progression of life wherever it may be found in the Universe. So, when Bostrom is talking about extinction, and the possibility of progression to posthuman, he’s considering the entire possible spectrum of life in the universe. Whereas billions of instances of life may go extinct, how many will make it to posthuman? When Bostrom posits that there’s perhaps a 20% chance that humanity won’t go extinct, what he’s really saying is that among the billions of possible forms of “humans” developing in different places throughout the Universe, some small subset of them may not go extinct.

Contrast this with Cannon who, best as I can tell, is speaking about we humans, here and now, on this planet. F1 asks us not simply to have faith that some lifeforms somewhere may bypass extinction; but that we, you and I, our specific race will bypass extinction. This matters because of the vast difference in probability that SA and NGA represent. Bostrom is potentially saying the odds are one in a billion that some race of human-like beings somewhere in the universe will make it. Cannon seems to be saying the odds are 1:1 that our race will make it; or at least to accept that on faith.

What matters here is that F1 is a significant part of the entire basis for NGA, and Cannon asks us to simply assume it is true. As he states in the summary of F1 “the proposition may be false.” This has implications for the claim that NGA is a logical argument.

Now, let’s evaluate the impact of F1 on the various truth claims:

F1:TC1: NGA is a Logical Argument

NGA claims to be a logical argument. F1 becomes the first premise of NGA, and as such, allows us to begin to examine TC1 more closely. F1 is assumed true as a matter of faith by Cannon, with weak probabilisitic certainty if not
assumed. F1 presents a significant problem. A deductive logical argument can be valid, even with a premise that is known to be false. For example:

• P1: All Little Boys Can Fly
• P2: Billy is a little boy
• Conclusion: Therefore, Billy Can Fly

This is considered a valid deductive logical argument, even though we know P1 and the conclusion are false. As a truth bearer, all we can say about this argument is that its structure is valid. But as a method to test the truth claim that Billy can fly, we’d call this argument false or unsound, because P1 is not true, and therefore the conclusion cannot be true. This would be so even were we to structure the above example as an inductive argument, concluding that it is “probable” that Billy can fly. Induction places similar weight on the veracity of the premises as does deduction.

F1 asks us to assume as true something that is possible, but likely improbable. And if F1 were known to be impossible, that would be the end of the truth claims of NGA. NGA’s purpose is not to convince us that it is a logically valid argument without consideration of the truth claims; NGA claims to prove truth claims, and must be judged on those claims. This likely triggers the Fallacy of Unwarranted Assumption, which can be characterized as arguments based upon questionable, although sometimes popular, principles or assumptions. That phrase “sometimes popular” jumps out due to what Cannon states in the summary when he describes F1 as “common” among both technologists and the religious. Additionally, F1 likely triggers the sub-set fallacy known as Ad Ignorantiam, or the argument from ignorance, which states that a specific belief is true because we don’t know that it isn’t true. Again, reference the summary where Cannon states “to the extent we do not know it to be false, we may have practical or moral reasons to trust that it is true.”

Logically, we can dismiss the entirety of the truth claims of NGA outright on the basis of F1. If we can’t know that F1 is likely, indeed if we know that F1 is probably unlikely, then we can’t know that any of the conclusions of NGA are likely. All logical arguments are only as strong as their weakest premise.

Take the following example, where I ask you to assume that A1 is true:

• A1: Assume we will cure cancer 50 years from today
• P2: Bob is a man living 60 years from today
• C: Therefore, Bob is cancer free.

This argument is logically valid because it is structured such that the conclusion follows necessarily from the premises. But If I were to call this the New Cancer Free Bob Argument and suggest that I am telling you something actual or likely to be true, based on the premises provided, you’d be forced to conclude that my argument is simply invalid relative to the truth claims. I can’t know A1, and therefore, I can’t make any claims about Bob (leaving aside for now the notion that I can’t even know that Bob exists, that he’s a man, etc.).

In the summary of F1, Cannon states:

“The proposition may be false. However, to the extent we do not know it to be false, we may have practical or moral reasons to trust that it is true.”

If we take the Doom Argument mentioned above, which posits a 95% chance that humanity will go extinct after approximately 9,120 years of humans existing on this earth, the “extent to which we do not know it to be false” would be 5%. In other words, we’d only have a 5% justification (exponentially less if we’re formulating just for this world we live on now) for trusting F1 is true “for practical and moral reasons.”

I do not know whether the Doom Argument presents us a valid measure of the probability of extinction (remember, Bostrom said we must give high credence to DOOM); the point is, nobody really knows or can in any way measure such a likelihood. So there is no logical “extent to which we do not know it to be false.”

I’m inclined to conclude that F1 is catastrophic to all of the truth claims of NGA, that this analysis could end here without need to analyze the remainder of the argument. However, to be charitable to the author, I will continue on, accepting for the sake of argument his request to simply assume F1 is true. I maintain the right, in my final analysis of NGA, to rely heavily on the concerns I am raising here.

To that end, F1 is assumed true, but not actually true:

Therefore, NGA, as a whole, cannot be:

• Deductive: Valid: Sound
• Inductive: Strong: Cogent

NGA still may be:

• Deductive: Valid: Unsound
• Deductive: Invalid: Unsound
• Inductive: Strong: Uncogent
• Inductive: Weak: Uncogent

After the first premise, F1, of NGA, NGA cannot be considered a sound or cogent logical argument.

• Status of TC1: Unconfirmed

F1:TC2: NGA is a Logical Argument for faith in God.

• F1 provides no information relevant to TC2
• Status of TC2: Unconfirmed

F1:TC3: NGA assumptions are consistent with science & technology

Science and technology can make little predictive analysis relative to the likelihood that humanity will achieve superhumanity. But F1 is also not inconsistent with a scientific view of the possibility that humanity may obtain superhumanity.

• Therefore, F1 is consistent with TC3.
• Status of TC3: 20% Confirmed (1 of 5 assumptions hold)

F1:TC4: Superhumanity is probably more compassionate than we are

• provides no information relevant to TC4
• Status of TC4: Unconfirmed

F1:TC5: Superhumanity created our world

• provides no information relevant to TC5
• Status of TC5: Unconfirmed

F1:TC6: God = superhumanity + compassion + creation

• provides no information relevant to TC6
• Status of TC6: Unconfirmed

F1:TC7: Human potential for advancement may require faith in this God.

• provides no information relevant to TC7
• Status of TC7: Unconfirmed

F1:TC8: Atheists may be required to distrust human potential.

• provides no information relevant to TC8
• Status of TC7: Unconfirmed

F1:Overall Analysis of NGA

• Does F1 increase or decrease confidence in the overall truth claims of NGA?
• F1 decreases confidence.

Next we will analyze The Compassion Argument.

The Compassion Argument (CO) of The New God Argument

After the Faith Assumption (F1) NGA presents its first sub-argument. The Compassion argument
(CO) is summarized this way:

The Compassion Argument is a logical argument for trust that superhumanity probably would be more compassionate than we are. The basic idea is that humanity probably will continue to increase in decentralized destructive capacity, so it probably will stagnate or destroy itself unless it increases in compassion. If we trust in our own superhuman potential, we should trust that superhumanity would be more compassionate than we are.

The NGA Compassion Argument (CO) is:

CO1 (Assumption)

  • P1: EITHER humanity probably will become extinct before evolving into superhumanity; OR
  • P2: superhumanity probably would not have more decentralized destructive capacity than humanity has; OR
  • P3: superhumanity probably would be more compassionate than we are

CO2 (Assumption)

  • (CO1:P2): superhumanity probably would have more decentralized destructive capacity than humanity has

CO3 (Conclusion: deduction from CO1, CO2, and F1)

  • superhumanity probably would be more compassionate than we are
  • Technical note: Cannon labels all is conclusions as “deductions.” It seems more accurate that he label them as inductions

Analysis:

We now enter into the structure of the logical argument of NGA. With F1 asserted as a faith position, CO is the first logical argument we encounter in NGA that is meant to logically cohere.

This is how CO is intended to be read:

CO1:P1 is negated by F1, CO1:P2 is negated by CO2, and therefore, CO1:P3 obtains, and becomes the Conclusion (CO3) of CO.

The Missing Argumentation

Immediately we encounter a difficulty: there is no supportive argument for CO. Are we to take CO as self-evident absent any argumentation? I reached out to Cannon to clarify how best to interpret CO absent an accompanying argument. (See my questions and his answers in the comments section on his blog [8])

Cannon states:

“CO1 is a logical disjunct with three possibilities, one of which must be true — or the assumption itself is false. The three disjuncts purport to be exhaustive of the possibility space they describe.”

Additionally, it wasn’t clear to me how CO1 and CO2 relate to each other. Does CO2 somehow follow from CO1? Again, from Cannon:

“there’s no logical path from CO1 to CO2. They are each assumptions, for which we may
make independent arguments.”

By not providing argumentation to accompany CO, Cannon introduces the logical fallacy known as Shifting the Burden of Proof. This, also known as an “argument from ignorance,” occurs when either a proposition is assumed to be true because it has not yet been proved false or a proposition is assumed to be false because it has not yet been proved true. This has the effect of shifting the burden of proof to the person criticizing the proposition. Cannon may consider CO1 so self evident as to not require argumentation and support. I disagree. CO1 raises significant questions which could challenge its veracity.

The first is whether CO1 is “exhaustive of the probability space it describes” as Cannon has asserted. I do not believe it is, as I can posit alternate explanations that are not covered in the disjunct. All that would be required to show that CO3 does not hold would be a single plausible alternative. It’s tempting here to list some, but doing so plays into the very fallacy we are discussing; Cannon, as the proponent for his argument, must provide his argument for the
coherence of CO1.

The next problem is another fallacy known as Affirming the Disjunct. This fallacy lies in concluding that one disjunct must be false because the other disjunct is true; in fact they may both be true. For example, it seems possible that CO:P2 and CO:P3 could both obtain. Or that CO:P2 and CO:P3 could both be negated independent of CO:P1. The relationship between CO:P2 and CO:P3 seems weak.

Part of the weakness of CO1 is that we are positing for a future instantiation of the human race. We have no way of knowing what that race is like, how they think and what their social constructions are. We may have completely erroneous conceptions about the future of life. By analogy, we may think of Cannon’s superhumanity like we once thought of Newtonian physics, but later come to realize that it’s more analogous to quantum physics. To put this into religious terms, “‘For My thoughts are not your thoughts, nor are your ways My ways,’ says the
Lord.” (Isaiah 58:8).

Speaking to a similar point, Bostrom states [9]:

“It’s hard to know what [posthumanity] might look like, because our human experience might be just a small little crumb of what’s possible. If you think of all the different modes of being, different kinds of feeling and experiencing, different ways of thinking and relating, it might be that human nature constrains us to a very narrow little corner of the space of possible modes of being. If we think of the space of possible modes of being as a large cathedral, then humanity in its current stage might be like a little cowering infant sitting in the corner of that cathedral having only the most limited sense of what is possible.”

Cannon may have persuasive argumentation and not just anecdotal evidence for CO1; but without seeing it, we can’t assume it exists. Any logical proof that requires us to guess a future subjective state is inherently weak. An actual argument would include a discussion as to why support for why the conclusions chosen by Cannon hold.

Links, in Lieu of Argument

In lieu of an actual argument, Cannon provides a list of links that variously support the theory that humanity is becoming less violent. Back to Cannon[10]:

“On the New God Argument website, I’m collecting information that relates to support for each of the assumptions. For example, on the Compassion Argument page, there’s a list of links below both CO1 and CO2.”

This is not an acceptable replacement for an actual argument. Cannon, for example, seems to largely rely on Steven Pinker, following themes from Pinker’s book “The Better Angels of our Nature” which argues that human violence has been trending toward peace in something of a natural arc. Pinker is a well-known cognitive scientist and popularizer of science. I’ve enjoyed reading Pinker. But Pinker and his theories are not without criticism. Pinker is widely criticized for what some see as his simplistic Hobbesian view of early man (Hobbes famously quipped that the life of early man was solitary, poor, nasty, brutish, and short). There are many theories of early man that are in direct contrast to Hobbes (and Pinker) on this topic. For example:

New Study of Foragers Undermines Claim That War Has Deep Evolutionary Roots
(Scientific American[11])
“A study published today in Science, “Lethal Aggression in Mobile Forager Bands and Implications for the Origins of War,” provides more counter-evidence to the Deep Roots Theory. The study’s authors, anthropologists Douglas Fry and Patrik Soderberg of Abo Akademi University in Finland, say their findings “contradict recent assertions that [mobile foragers] regularly engage in coalitionary war against other groups.”

There also abound many proponents of theories that directly counter the theories of Pinker, placing the blame for war on the rise of modern culture itself. Well known linguist and philosopher Noam Chomsky has an ongoing feud with Pinker. They were colleagues at one point. In this interview with Lawrence Krauss[12], in response to the assertion that peace is emerging, Chomsky states:

Chomsky: well that’s what [Pinker] claims, but the weight of claims is strongly against it. There’s very good work by serious anthropologists, Brian Fergusson is one who’s worked through [Pinker’s] data and shown that it’s probably wrong for 95% of human existence; namely the Hunter Gatherer period. It’s really with the emergence of state systems, the development of complex agriculture, that you begin to get wars that are very destructive. I mean there’s killing among hunter gatherers, but it’s pretty limited.

Krauss: but there has been moral progress, in fact you wrote…

Chomsky: well one thing that [Pinker] says is correct — and I think that’s about it — since the enlightenment, there has been moral progress. In our own lifetimes, there’s been moral progress. So consider for example the status of women’s rights today and in 1950. Very different. The status of civil rights, very different. Gay rights, very different. That’s progress. And it’s been going on slowly since the enlightenment. But that’s a pretty brief period of human history. And in fact, right at this same time have been the most murderous, destructive wars ever. And even seventy years ago, the creation by some of the smartest people in the world of a device that may destroy us all. And is on the verge of destroying us all

There is an ongoing debate between Nassim Nicholas Taleb (Author, The Black Swan) who contends that Pinker simply doesn’t apply proper statistical analysis[13]:

“Pinker doesn’t have a clear idea of the difference between science and journalism, or the one between rigorous empiricism and anecdotal statements. Science is not about making claims about a sample, but using a sample to make general claims and discuss properties that apply outside the sample.”

“The mistake appears to be commonly made in common inference about fat-tailed data in the literature. The very methodology of using concentration and changes in concentration is highly questionable. For instance, in the thesis by Steven Pinker that the world is becoming less violent, we note a fallacious inference about the concentration of damage from wars with minutely small population in relation to the fat-tailedness. Owing to the fat-tailedness of war casualties and consequences of violent conflicts, an adjustment would rapidly invalidate such claims that violence from war has statistically experienced a decline.”

And perhaps with direct relevance to the topic at hand, Taleb states:

“we do not estimate the severity of a future risk based on past in-sample historical data.”

Well known Philosopher John Gray is another of Pinker’s known critics (read his rebuttal to Pinker here: Steven Pinker is Wrong About Violence and War14). Again, with perhaps some relevance to Cannon’s argument for faith:

“Comparing Pinker’s statistical analyses to the obsidian mirrors with which Aztec soothsayers divined the future, Gray contends that Pinker’s vision of progress stems from faith rather than reason.”

The point here is not that I am trying to prove Pinker wrong. I’m illustrating that Pinker is not without his respected critics. Scientific American published an article in 2015 attempting to moderate the ongoing dispute between Pinker and Gray (Scientific American: Steven Pinker, John Gray and the End of War[15]).

Those debates should go on. The net result, however, is a decrease in the trust one might have in simply choosing sides in this debate and declaring that you have evidence for a logical claim about the future social and moral constructions of superhumans. Note what I am doing here: I am countering arguments that I assume Cannon intends in support of CO. But those arguments do not yet exist. Which brings us back to the fallacy of shifting burdens.

More Unwarranted Assumptions

After the significant ask to take F1 on faith, it is of concern to find that CO1 is specifically reliant on F1, which negates CO1. But more importantly, it is problematic that Cannon is again asking us to take CO2 simply as an assumption. CO2, and subsequently CO3, are core pillars to the entirety of NGA.

Assumptions are an integral part of all logical arguments. We frequently rely on assumptions for premises that are likely true, but may have a gap. The author of an argument asks the audience to grant the assumption. Those asks, however, are generally not meant to fill in blanks where the author has simply failed to provide a valid argument.

There’s an anecdote of a mathematician at a chalk board, who after writing out a lengthy proof but still not getting all the way to the conclusion says “and here’s where the magic happens.” By asking us to take CO2/CO3 on assumption without providing argumentation, Cannon is asking us to fill in the gaps with something akin to magic (faith?); he’s asking us to know things we can’t know.

If CO1 and CO2 were sufficiently persuasive on their own, this would be a reasonable ask. For example, if a premise said “all biological animals need oxygen” we could assume this is sufficiently true even though it’s possible that an animal may be found that doesn’t require oxygen. But turn that premise around to say “some animals don’t require oxygen” based on the assumption that it’s possible we may find such an animal in the future and the premise would be
considered an unwarranted assumption, and therefore rejected by the reader.

Is “Possible” Enough?

It is not enough for Cannon to claim that the truth claims in CO are possible, if he then intends this to be support for the truth claims of NGA. “Possible” does not provide sufficient evidence to be able to make the truth claims NGA makes.

There is no bright line here; but generally speaking, if an argument intends to prove truth claims, the gap in the assumptions should have a probability approaching 1. If the probability is significantly less than 1, our confidence in the conclusion is significantly diminished. I am not persuaded that CO2 has a probability approaching 1.

Analyzing the Conclusion: CO3

CO3 (Conclusion: deduction from CO1, CO2, and F1)

• superhumanity probably would be more compassionate than we are

What I don’t understand about CO is why Cannon has bothered with CO1 and CO2 at all. Both are premised on the reader simply accepting assumptions F1 and CO2. To assert that CO3 is logically derived from CO1, CO2 and F1 may prove to be a valid argument, but it fails to be a sound argument, because we have diminished faith in the premises.

Realistically, CO is really F2; another faith assumption. It is not logically derived from true premises, but simply asked to be assumed.

Therefore, while CO may be structurally valid, the conclusion is necessarily unsound.

Analyzing Status of Truth Claims of NGA Resulting from CO

CO:TC1: NGA is a Logical Argument

NGA claims to be a logical argument. CO is the first logical argument of NGA. Most of the concerns about unwarranted assumptions that I raised in F1 are repeated in CO. But this time, those concerns invalidate a core conclusion of NGA.

The form of NGA could still be valid; but it can no longer be sound.

Since I don’t agree that CO3 follows necessarily from CO1 and CO2 due to a lack of sufficient confidence in the premises, therefore:

NGA cannot be considered a Deductive argument, because CO3 claims to derive probabilistically, rather than necessarily, from its premises.

NGA cannot be considered “Inductive: Strong” due to our diminished trust in the premises of CO.

Therefore, NGA, as a whole, cannot be:

• Deductive: Valid: Sound
• Inductive: Strong: Cogent
• Deductive: Valid: Unsound
• Deductive: Invalid: Unsound
• Inductive: Strong: Uncogent

NGA still may be:

• Inductive: Weak: Uncogent
• Status of TC1: Unconfirmed. However, insofar as Cannon intends NGA to be not just a valid argument, but a cogent argument, then NGA has failed.

CO:TC2: NGA is a Logical Argument for faith in God.

• CO provides no support for the argument for faith in God.
• Status of TC2: Unconfirmed

CO:TC3: NGA assumptions are consistent with science & technology

CO2 posits that superhumanity will be more compassionate than we are. Compassion can be argued to be a subjective term, and it would be hard to understand how this premise can be tested, primarily since we are imposing it upon a future race of whom we know nothing other than they are biologically (or by other means?) related to us. I grant however that it may be possible to definitionally define compassion in a measurable way.

• Therefore, CO is consistent with TC3.
• Status of TC3: 60% Confirmed (3 of 5 assumptions hold)

CO:TC4: Superhumanity is probably more compassionate than we are

CO provides the core of the argument for TC4. So the question is whether it has provided a compelling argument that would satisfy the truth claim that superhumanity is probably more compassionate than we are. But as stated, Cannon has provided no argument for the claim other than presenting it and asking us to assume it holds. This is insufficient for a core truth claim.

Additionally, Cannon purports to show that superhumans are compassionate not just relative to their fellow beings, but to us specifically. I do not see how we can reach this conclusion with the provided arguments.

Therefore:

• Status of TC4: Failed due to insufficient evidence and argumentation

CO:TC5: Superhumanity created our world

• provides no information relevant to TC5
• Status of TC5: Unconfirmed

CO:TC6: God = superhumanity + compassion + creation

• provides no information relevant to TC6
• Status of TC6: Unconfirmed

CO:TC7: Human potential for advancement may require faith in this God.

• provides no information relevant to TC7
• Status of TC7: Unconfirmed

CO:TC8: Atheists may be required to distrust human potential.

• provides no information relevant to TC8
• Status of TC7: Unconfirmed

CO:Overall Analysis of NGA

• Due the concerns raised here about CO, and given that we cannot have confidence that CO3, the conclusion, follows directly from F1, CO1 and CO2, I judge CO3 to have failed. CO therefore fails to advance the NGA. With the failure of CO, NGA is thus invalidated.
• CO invalidates the conclusions of NGA.

Where to Go From Here?

I expressed doubts following F1, but opted to continue in an effort to be charitable. After CO, I find myself in an even deeper hole. CO, as currently presented, invalidates NGA. However, in the interest of being thorough, and in case I have misjudged CO, it seems prudent to commit to taking this analysis all the way to the end of NGA.

The Creation Argument (CR) of The New God Argument

Next in the progression of NGA is The Creation Argument (CR):

“The Creation Argument is a logical argument for trust that superhumanity probably created our world. The basic idea is that humanity probably would not be the only or first to create many worlds emulating its evolutionary history, so it probably will never create many such worlds unless it is already in such a world. If we trust in our own superhuman potential, we should trust that superhumanity created our world.”

The NGA Creation Argument (CR) is:

CR1 (Assumption)

• P1: EITHER humanity probably will become extinct before evolving into superhumanity; or
• P2: superhumanity probably would not create many worlds emulating its evolutionary history; or
• P3: superhumanity probably created our world

CR2 (Assumption)

• superhumanity probably would create many worlds emulating its evolutionary history

CR3 (Conclusion: deduction from CR1, CR2, and F1)

• superhumanity probably created our world

Analysis:

This is how CR is intended to be read:

CR1:P1 is negated by F1, CR1:P2 is negated by CR2, and therefore, CR1:P3 obtains, and becomes the Conclusion of CR.

The Simulation Argument, Imported…except…

CR is Bostrom’s Simulation Argument, with substitution of a few words. On the positive side, CR can lean on all the arguments for SA. But again, oddly, NGA does not import those arguments in support of CR. In my earlier conversation with Cannon, he states:

“The assumption for which the most argumentation has been done directly is probably CR1 in the form of the Simulation Argument, which can then be generalized for all feasible creation mechanisms (computation or otherwise) into CR1.”

Note here that Cannon, in following SA, is not necessarily arguing for the potential that we are living in a computer simulation. Cannon substitutes something like terraforming (the creation of physical worlds and realities) for the computer simulation argument contained in SA. Bostrom has stated that he thinks SA holds with such substitutions with caveats:

“Yes, the form of the simulation argument could in principle be applied more generally, but one would have to check in each case whether the necessary empirical presuppositions obtain and whether the conclusions are interesting and significant.” Bostrom (“The Simulation Argument FAQ.” The Simulation Argument. 2008. Web. 15 Aug. 2014[16])

The Catastrophic Argument

CR continues in the theme of asking us to assume the core elements of the argument. In this instance, such an ask is catastrophic to NGA.

As has been mentioned several times, Bostrom explicitly claims that SA does not and cannot be used to predict which of the three disjuncts would obtain. Bostrom claims that the argumentation of SA, rooted in probability theory, are exhaustive on this limitation of SA.

Cannon clearly respects Bostrom and SA, as he has built NGA directly on the work of SA. And yet, in CR, in contravention to SA, Cannon is asking us to assume that P3 of SA obtains.

In the prior analysis, of CO, we could at least hope that it might be possible for Cannon to present arguments that convince us that CO3 holds. But here, in CR, we have explicit proof, through the argumentation of SA, that CR3 cannot hold, due the limitations of the argument.

As discussed previously, logic arguments can and do rely on assumptions; but those assumptions must be close to 1 in probability in order for the conclusions to be sound. We have explicit proof of the inability of CR to hold, because we have explicit proof that this logical argument cannot predict which disjunct obtains. We can’t say, as the mathematician with the unfinished proof “here’s where the magic happens.”

Therefore, it is unreasonable of Cannon to ask us to assume that CR1:P3 obtains. And insofar as Cannon claims that NGA is both logically valid and logically sound, and that it presents supported truth claims, NGA fails.

We can now create a logically valid and sound argument for the failure of both CR and NGA:

The Failure of CR:

• Whereas SA provides proof of its inability to predict which of 3 disjuncts obtain; and
• whereas CR is built on the same premises as SA; and
• whereas CR relies on CR1:P3 obtaining, in contravention to evidence presented in SA,
• Therefore CR Fails

The Failure of NGA:

• Whereas CR Fails; and
• whereas NGA relies on the validity of CR,
• Therefore NGA Fails.

No further discussion of CR is necessary.

Analyzing Status of Truth Claims of NGA Resulting from CO

CR:TC1: NGA is a Logical Argument

NGA may still be logically valid, in that assuming the premises, the conclusions may follow. However, NGA can no longer be consider logically sound.

Therefore, NGA, as a whole, cannot be:

• Deductive: Valid: Sound
• Inductive: Strong: Cogent
• Deductive: Valid: Unsound
• Deductive: Invalid: Unsound
• Inductive: Strong: Uncogent

NGA still may be:

• Inductive: Weak: Uncogent
• Status of TC1: Unconfirmed.

CR:TC2: NGA is a Logical Argument for faith in God.

• TC2 relies on the soundness of NGA. Since CR has failed and since NGA has failed, therefore:
• Status of TC2: Failed

CR:TC3: NGA assumptions are consistent with science & technology

Whereas CR has failed explicitly due to conflicts with probability theory explained in the
argumentation of SA, therefore:

• Status of TC3: Failed

CR:TC4: Superhumanity is probably more compassionate than we are

Whereas TC4 failed due to analysis in CO, therefore:

• Status of TC4: Failed

CR:TC5: Superhumanity created our world

Whereas TC5 is explicitly reliant on CR, and whereas CR has failed, therefore:

• Status of TC5: Failed

CR:TC6: God = superhumanity + compassion + creation

Whereas TC6 relies explicitly on the validity and soundness of NGA, and whereas NGA fails, therefore:

• Status of TC6: Failed.

CR:TC7: Human potential for advancement may require faith in this God.

Whereas TC7 relies explicitly on the validity and soundness of NGA, and whereas NGA fails,
therefore:

• Status of TC7: Failed.

CR:TC8: Atheists may be required to distrust human potential.

Whereas TC8 relies explicitly on the validity and soundness of NGA, and whereas NGA fails, therefore:

• Status of TC8: Failed.

CR:Overall Analysis of NGA

• Whereas CR Fails; and
• whereas NGA relies on the validity of CR,
• Therefore NGA Fails.

Where to go from here

While both F1 and CO created conditionals under which I was able to continue to work through
NGA, albeit with skepticism, CR causes NGA to fail, and thus ends the need for further analysis.
However, I’ll proceed to an outline of G1, and list some questions and concerns for it that I think
would be necessary to address were we doing a full analysis of G1.

The God Conclusion (G1) of The New God Argument

Summary:

“The God Conclusion is a logical deduction for faith in God. Given assumptions consistent with contemporary science and technological trends, the deduction concludes that if we trust in our own superhuman potential then we should also trust that superhumanity probably would be more compassionate than we are and created our world. Because a compassionate creator may qualify as God in some religions, trust in our own superhuman potential may entail faith in God, and atheism may entail distrust in our superhuman potential.”

The NGA Creation Argument (CR) is:

G1: (deduction from CO3 and CR3)

• BOTH superhumanity probably would be more compassionate than we are; and
• superhumanity probably created our world

Analysis:

G1 fails due to the failure of CR3. Additionally, CO3 provides significant doubt due to an over reliance on unwarranted assumptions.

While that is the end of the formal arguments in NGA, there remain significant questions that arise from the overall Summary of NGA and the Summary of G1. These specifically involve the following Truth Claims:

• TC6: God = superhumanity + compassion + creation
• TC7: Human potential for advancement may require faith in this God.
• TC8: Atheists may be required to distrust human potential

These truth claims are seemingly the most important conclusions of NGA, as stated in the Summary. These are the logical end that Cannon wants to get to. Therefore, it’s concerning to see that none of these truth claims are included in any of the logical arguments. There is nothing about them in F1, CO, CR, or G1. The only place they are mentioned is in the summary. Had NGA not failed for prior reasons, this would be a significant flaw in NGA. It’s not clear how Cannon supports these conclusions without including logical arguments that lead to their conclusions.

A Naturalistic God? Problems with TC6

TC6: (God = superhumanity + compassion + creation) has the semblance of an argument for it included in the summary, where Cannon states: “Because a compassionate creator may qualify as God in some religions, …” but the argument is incomplete. I assume he means to imply “therefore, we can label the superhumans in NGA “God.”

This is the most significant claim of NGA, and there is scant argument for it. But let’s analyze the argument as it stands. The god described in NGA is a naturalistic god, which contrasts with a supernatural, universal and all-knowing god. Paraphrasing, Cannon says “because someone somewhere may consider this god, we can therefore label it God.”

Let’s turn that into a logical argument:

P1: Superhumans described in NGA are compassionate creators
P2: Some people somewhere would call such beings God
C: Therefore, we can call these beings God.

This argument may be logically valid in that the premises (if true) might support the conclusion. The problem is that few people would consider this a sound argument, and few would accept the definition. P2 is wholly lacking in probabilistic certainty; rather it is characterized as likely having very little general support throughout the religious community.

The vast majority of religious people reject such a definition of god. Here are common definitions of God held by major religious organizations, none of which are compatible with the definition of god posited by Cannon:

The Church of Jesus Christ of Latter-day Saints [17] (mormons):
God the Father is the Supreme Being in whom we believe and whom we worship. He is the
ultimate Creator, Ruler, and Preserver of all things. He is perfect, has all power, and knows
all things. He “has a body of flesh and bones as tangible as man’s”

Catholicism [18]:
The Catholic Church teaches that God is All-Perfect; this infinite Perfection is viewed,
successively, under various aspects, each of which is treated as a separate perfection and
characteristic inherent to the Divine Substance, or Essence. A certain group of these, of
paramount import, is called the Divine Attributes.

Judaism [19]:
God in Judaism is understood to be the absolute one, indivisible, and incomparable being
who is the ultimate cause of all existence.

Islam [20]:
In Islamic theology, God is the all-powerful and all-knowing creator, sustainer, ordainer and
judge of everything in existence.

Problems with TC6:

The Problem of Evil:

The problem of evil[21] is a significant hurdle for the theory of a Compassionate Naturalistic God. While all the traditional problems of the problem of evil hold (both logical and evidential) that would exist with an Omnipotent God, those problems are significantly amplified with a Naturalistic God. We don’t know anything about the motivations of “superhumans” and if we are going to rely on a logical argument, then a logical argument could be made to completely invalidate NGA; there’s significantly more evidence of extreme suffering (surpassing the arguments from free will on the problem of evil) in humanity that would likely weigh in stronger than NGA. Let’s craft a simple argument:

The Argument Against Superhuman Compassion

P1: Humans, including children, experience a great deal of suffering; and
P2: Any advanced race of intelligent beings capable of creating human simulations
would have the ability to either eliminate or significantly reduce human suffering; and
P3: Any being who would choose to instantiate severe suffering into a program meant to
simulate human life cannot be considered compassionate; therefore:
C: Any such beings cannot be considered compassionate.

Whereas logic problems require a high degree of probability, this argument likely obtains much easier than does NGA. If we can make an argument that is more readily obtainable than NGA, then NGA fails.

The Problem of Consciousness

A note on consciousness. I believe that Bostrom underplays the potential problem of consciousness. This problem manifests in two ways:

1. The problem of simulating actual consciousness in beings
2. The problem of knowing whether another being experiences consciousness

In SA, the first problem would likely mean that disjunct (1) obtains; by definition, SA says that posthumans are those who are able to recreate high-fidelity simulations, which would include consciousness. But the second problem, the ability to know whether a being is experiencing consciousness is a likely more difficult problem. In SA, this is not a problem because SA doesn’t make any claims as to why posthumans would create simulations; and they may create them without realizing that the “characters” in the simulations are experiencing consciousness.

This problem becomes much more acute in NGA, where Cannon asserts that not only are Superhumans more compassionate, but that their compassion extends toward the created beings. If Superhumans can’t know whether their creations are conscious, that’s a potential problem with the compassionate God argument.

Problems with TC7: Human potential for advancement may require faith in this God.

TC7 has no logical argumentation or support. It’s one of the primary conclusions of NGA, but it is not included in the logical arguments, and not explained in the summaries. Cannon does not explain how we get from G1 to the requirement of faith for advancement.

Problems with TC8: Atheists may be required to distrust human potential

As with TC7, there are no logical arguments for TC8. It seems simply tacked onto the end of the summary of NGA, with no explanation or argument as to how or why this is shown in NGA.

Conclusion

NGA simply does not hold. It is built upon unwarranted premises.

By analogy, imagine I were to create an argument that purported to prove that it’s possible to climb Mt. Everest in less than 1 minute, and then I ask you to assume “You can Fly” as a Faith Assumption.

NGA follows the form and structure of SA and then, without providing any additional argumentation or logic, claims to be able to conclude things that SA and Bostrom claim cannot be derived from the argument.

NGA set a high bar for itself, claiming to logically prove new and novel arguments related to God and Faith. The best theologians and philosophers in history have been unable to achieve such lofty goals. Neither has Cannon.

Cannon has built the foundation of NGA on the back of five unwarranted assumptions; that is a significant clue to the careful reader that NGA is unsound. NGA offers no new facts or argumentation beyond what is contained in Bostrom’s SA, and yet Cannon claims to be able to reach grandiose conclusions previously not known to be achieved by any prior philosophers or theologians.

In the final analysis, NGA tells us nothing more interesting than this:

  • Assume that P3 of SA Obtains.
  • Call the resulting species God.

NGA does not hold.

References

  1. https://new-god-argument.com
  2. http://www.simulation-argument.com
  3. https://en.wikipedia.org/wiki/Simulation_hypothesis#Bostrom.27s_trilemma:_.22the_simulation_argument.
  4. philsci-archive.pitt.edu/11537/1/Computer_Simulation_paper_revised.pdf
  5. https://en.wikipedia.org/wiki/Doomsday_argument
  6. www.simulation-argument.com/simulation.html
  7. www.simulation-argument.com/simulation.html
  8. lincoln.metacannon.net/2016/06/new-god-argument-version-33.html
  9. http://www.theatlantic.com/technology/archive/2012/03/were-underestimating-the-risk-of-human-extinction/253821/
  10. lincoln.metacannon.net/2016/06/new-god-argument-version-33.html
  11. http://blogs.scientificamerican.com/cross-check/new-study-of-foragers-undermines-claim-that-war-has-deepevolutionary-
    roots/
  12. https://www.youtube.com/watch?v=zy0n4dbHbdA
  13. http://www.fooledbyrandomness.com/pinker.pdf
  14. https://www.theguardian.com/books/2015/mar/13/john-gray-steven-pinker-wrong-violence-war-declining
  15. http://blogs.scientificamerican.com/cross-check/steven-pinker-john-gray-and-the-end-of-war/
  16. www.simulation-argument.com/faq.html
  17. https://www.lds.org/topics/god-the-father?lang=eng
  18. https://en.wikipedia.org/wiki/Catholic_views_on_God
  19. https://en.wikipedia.org/wiki/God_in_Judaism
  20. https://en.wikipedia.org/wiki/God_in_Islam
  21. https://en.wikipedia.org/wiki/Problem_of_evil