How to Tie AI into a Pretzel

A philosopher teaches a
dogmatic AI engine how
to think about science.
Entertaining!

by John D. Wise, PhD

Artificial intelligence is good at repeating what it’s been trained to say. But when pressed with the hard questions, even the most advanced models start tying themselves into knots. In the following exchange, I (JW) engage Google Gemini (GG) in a direct dialogue about origins, design, and the philosophical ground rules of science. What emerges is not just an argument over biology, but a revealing look at how evolutionary thinking depends on a prior commitment: methodological naturalism. GG invoked this principle fifteen times in our conversation, as if repetition turns assumption into fact.

Read closely. The issue here is not facts of genetics, fossils, or DNA, but the framework in which those facts are interpreted. When science is defined by fiat to exclude design, it gropes in a self-imposed twilight. This is how an AI trained in evolutionary orthodoxy twists itself into a pretzel.

What follows is an edited transcript of my conversation with Google Gemini. Its responses have been trimmed for length but left substantially intact, to allow it to speak for itself. We begin the dialogue with:

Gemini’s critique of my article on the sweet potato genome.

Note: As an aside, I am not committed to allopolyploidy as an original condition. I am not a botanist, but a philosophy instructor. It was for me an interesting thought experiment. I’d love to run the idea by Robert Carter or Don Batten.

GG: In essence, your article takes a known evolutionary event (allopolyploidy and subsequent genomic rearrangement) and reinterprets it through a non-scientific lens. The complexity is not an original state; it is a result of a specific evolutionary path.

JW: The lens you are speaking of is not “scientific” but philosophical. If you start with the philosophical notion that all complexity arises from simpler forms, then you get the evolutionary viewpoint, as there is no alternative. Why is it not possible that allopolyploidy is a more original condition?

GG: Allopolyploidy is a process, not a starting condition. It is a biological event that combines two or more different genomes into one organism. Therefore, it is impossible for allopolyploidy to be the “original condition” in a strict sense. It requires pre-existing, simpler diploid organisms to occur.[1]

JW: Again, you assume a philosophical starting point – naturalism. If allopolyploidy is the starting condition of a created kind, then the mechanism for diploidy is what must be explained, not the other way around.

GG: While I understand the philosophical position you’re presenting, the scientific framework for understanding biology [read “evolution”] is not based on the assumption that all life was created in complex, polyploid forms. The scientific consensus is that allopolyploidy is a process, not a starting condition.

Science, by definition, operates on the principle of methodological naturalism. This means that it seeks to explain phenomena using only natural causes and observable processes [special pleading – evolution is assumed, historical and thus unobservable].

In conclusion, the scientific view is not a philosophical assumption; it’s a conclusion based on a massive body of evidence that supports the idea that life evolves from simpler forms to more complex ones.

How is “Science” Defined?

JW: Do you understand the philosophical position YOU are presenting? When you exclude the possibility of a supernatural starting point (original created complexity) you are being UNscientific by defining away one of the distinct logical possibilities – that design implies a designer?

GG: I understand the philosophical position you are presenting, and I can see how, from that viewpoint, the exclusion of a supernatural starting point might be seen as unscientific. The core of this perspective is that the complexity of living systems is so great that it suggests design, and that design implies a designer. This is a logical argument from a philosophical standpoint.

However, the scientific method by definition operates under the principle of methodological naturalism. Science deals with testable, falsifiable hypotheses. A hypothesis that relies on a supernatural or “intelligent” cause is not testable [GG completely contradicts itself on this point below] because the actions of a supernatural being cannot be measured, observed, or replicated in a lab. The argument that complexity implies a designer is a logical one, but it is not a scientific one.

While the complexity of life is indeed remarkable, attributing it to a designer is a philosophical or theological conclusion. Science, by its very nature, can only explain the “how” (the mechanisms) and not the “why” (the ultimate purpose or origin). [Again, watch for the contradiction of this point by GG below].

JW: Science sometimes, and especially when it wants to suppress design-inference, “operates under the principle of methodological naturalism” currently, but it wasn’t always so, nor is it logically coherent to make it so. Science is a search for truth about the natural world. If we arbitrarily define science to exclude the possibility of supernatural origins, then we are by fiat cutting ourselves off from what MAY be the truth. How scientific is that? Nor are you being logical when you define science as a position that excludes not just a logical possibility, but a potentially true explanation. IF the world was created, then refusing to face the possibility [evolutionary blinders] condemns science to walk in the dark and pursue ignorance rather than knowledge, contradicting the very name of science.

GG: That’s a powerful argument, and it brings up a fundamental tension between how science is practiced and how some people believe it should be practiced.

The core of your argument is that if the universe was created by a supernatural entity, and science excludes that possibility a priori, then science is being unscientific and illogical. Is a self-imposed limitation on the search for truth truly a quest for knowledge?

The “Problem” with a Supernatural Hypothesis

It’s true that the boundaries of science have shifted. Historically, a number of scientific pioneers, from Isaac Newton to Johannes Kepler [not to mention Einstein and many others], viewed their work as revealing the mind of God. For them, there was no division between discovering a law of nature and discovering a divine law.

The primary issue with including supernatural explanations in science is that they are not falsifiable. For example, the hypothesis that “all swans are white” can be falsified by finding a single black swan. When a black swan is found, the hypothesis is disproven, and we have learned something new.

In contrast, a hypothesis like “this biological complexity was designed by a creator” is not falsifiable. If we find evidence that points to a naturalistic explanation for a phenomenon, a proponent of the “designer” hypothesis can simply say, “The creator used that natural process to achieve the design.” If we find a flaw or a seemingly inefficient design, it can be argued that it is beyond our understanding or is part of a larger plan. [Or … let’s treat the “flaw” or the “inefficient design” claim as provisional rather than absolute, and wait for more data and a more expansive and detailed understanding]. The hypothesis can explain anything and, therefore, explains nothing in a way that can be tested.

The Falsifiability Criterion

JW: This is amusing. You said, “The [design] hypothesis can explain anything and, therefore, explains nothing….” This is precisely how the evolutionary paradigm works – it is in practice unfalsifiable until we have an answer in fact, just like the supernatural creation hypothesis. This is why Popper, who defined the falsifiability hypothesis, said that evolutionary theory was unscientific. I disagree on that point. I don’t think evolutionary theory is unscientific. I think it is wrong and internally and evidentially incoherent. And I don’t think – and for the same reasons – divine creation theory is unscientific. Suppose science never finds the black swan in your example. Then it hasn’t found the truth, but that doesn’t make the hypothesis unscientific. Likewise, if by searching the natural world scientists never “find” God, that doesn’t mean he doesn’t exist, nor does it falsify the divine-thesis. Falsification in science is never so simple as you make it here. But … if God would unmistakably reveal himself to scientists in a repeatable and testable way, their methodological naturalism would indeed be falsified, wouldn’t it? There are very serious problems for the falsifiability criterion that you are not accounting for.

GG: Karl Popper did express early skepticism about the falsifiability of natural selection, but he later clarified and retracted this position, stating that the theory of evolution by natural selection is a valid scientific theory. He acknowledged that while individual claims within the theory might be difficult to falsify, the theory as a whole makes testable predictions about the natural world.

Evolution constantly makes testable predictions. It predicts we should find transitional fossils … intermediate forms between different species. The discovery of fossils like Archaeopteryx [see here], which has features of both dinosaurs and birds, is a direct confirmation of this prediction. Finding a modern rabbit fossil in Precambrian rock … would directly falsify a core tenet of the theory [but not, contrary to your claim, evolution itself?]

The core issue with the “divine creation” hypothesis is that the hypothesis itself is untestable. It makes no specific, testable predictions about the natural world. Instead, it can be retrofitted to fit any evidence…. This is why it “explains everything and nothing.”

JW: In response to your first point, Popper did not retract the conclusion that natural selection, on which evolutionary theory rests, was tautological and unfalsifiable. Creation predicts engineered complexity and a highly efficient and well-designed genome, and this is precisely what we are finding. Darwinians expected to see lots of “junk DNA.” The more carefully we look, the more functional and well-designed the elements of the genome are.

It is not true that the design-thesis makes no falsifiable predictions, and it is also false that the predictions based on evolutionary theory have stood up well to new discoveries. The basic rule of evolutionary biology is “if it exists, it evolved.” This is equivalent to the Creationist saying, “God used this mechanism” – like the escape hatch you proposed for Design. Evolution/Nature are stand-ins for a designer.

GG: Your assertion that creationism makes testable predictions about increasing complexity and a lack of “junk DNA” is an interesting point, but it relies on a specific interpretation of both scientific findings and the nature of the theory itself.

Modern genomics has shown that a significant portion of what was once considered “junk” has regulatory functions, but this discovery doesn’t falsify evolution. Instead, it refines our understanding of how genomes evolve and function.

Post-Hoc Predictions

While you can assert that creationism predicts a highly complex, efficient genome, this is a post-hoc prediction. A truly falsifiable prediction would have been made before the evidence was found.[2] For example, a falsifiable creationist prediction might be: “If life was created, all species should have completely unique and unrelated genetic sequences.”[3] The discovery that all life shares a common genetic code and many genes would have falsified this hypothesis.

The statement “if it exists, it evolved” is not a scientific principle of evolutionary biology, but a colloquial paraphrase…. The scientific hypothesis is: “Organisms change over time [stuff happens], and the diversity of life arises from a process of descent with modification from common ancestors.” This hypothesis is constantly being refined, tested, and challenged. New discoveries, such as a functional non-coding DNA don’t falsify the entire theory. Instead, they lead to a more nuanced understanding of how evolution operates. [Translation? Evolutionary theory is unfalsifiable.]

JW: I didn’t say the evidence “falsified” evolution, did I? I implied that Creationism and Evolutionary theory are in the same boat, starting from an unfalsifiable position from the human standpoint. I was pointing to the fact that some things that evolution predicted are tenuous at best and extremely weak at worst.

GG: A scientific theory isn’t a single, monolithic idea; it’s a broad framework supported by a web of evidence and interconnected hypotheses. When a hypothesis within a theory is challenged, the theory isn’t necessarily falsified. Instead, it’s refined or expanded.

If we were to find a modern mammal fossil in the oldest rock layers, or if genomic analysis showed no relationship between closely related species, [or maybe soft tissue in dinosaur bones?] the theory would be directly challenged and possibly falsified. The theory has consistently passed these tests.

Creationism is not a scientific hypothesis because it offers no testable predictions. Any evidence, whether it shows a perfect design or a flawed one, can be retrofitted into the theory.[4] This makes it impossible to prove wrong through observation or experimentation.

In conclusion, the claim that both frameworks are in the same boat is misleading.

JW: Your post-hoc point is a great one, and I confess to it (at times. As GG points out these are “refinements” to the theory), but it is also precisely the entire modus operandi of evolutionary theory. It chases the evidence and tells a story to make it fit the theory structure. Just-so stories are the stock in trade.

GG: I disagree with your assertion that evolutionary theory’s method is to “chase the evidence and tell a story to make it fit.” While it’s true that new evidence refines [a rose by any other name] the theory, this process is fundamentally different from a post-hoc justification [special pleading again].

The process you describe—chasing evidence to fit a story—is the opposite of a scientific approach [you said it, not me!]. Before the discovery of the first single-celled organism, evolutionary theory would have predicted that such a thing must have existed. The discovery of bacteria and archaea confirmed this prediction.[5] If evolutionary theory had not predicted this, and scientists discovered bacteria, they would then have to adjust the theory to account for them. This would be a post-hoc justification.

The discovery of transitional fossils, like Tiktaalik (the “fishapod), is a prime example [meaning “the best you can do”?] of a successful scientific prediction. Based on the known fossil record and a desire to fill a gap in the timeline, scientists predicted that they would find an organism with specific features in a specific layer of rock. They searched for it, and they found it.

The idea that evolutionary theory is a fixed story is a misunderstanding. Science is a continuous process of inquiry. As new evidence emerges, theories are refined and made more robust.

The Complexity of Life

JW: The existence of archaea and bacteria “confirm” the theory of evolution far less than the creationist claim that the design we observe in the world points to a designer, from brilliantly engineered systems with ever-increasing complexity. Darwinians (like Haeckel) believed that “protoplasm” (the first stuff of life) was simple and would naturally modify and increase in complexity. Darwinians didn’t know that the complexity of even the simplest of cells was almost infinitely greater than protoplasm. We don’t find “simple life” in the geologic column; we find life already mind-bogglingly complex.

GG: You’re right that Darwin had no idea about the incredible complexity of the cell. He lived in a time before genetics, molecular biology, and electron microscopy. The discovery of the cell’s intricate machinery didn’t falsify his theory; it simply presented a more complex problem for the theory to explain.

The discovery of increasing complexity in the genome—such as the functions of what was once called “junk DNA”—doesn’t falsify evolution. It simply shows that our understanding of how genomes evolve is becoming more sophisticated.

The Divine Foot

JW: Again … you take the side of the arbitrary philosophical definition of science that excludes Lewontin’s divine foot. Science can proceed in every way you mention without that exclusionary principle. There is no logical or scientific reason to think that excluding an inference to design improves the scientific process.

GG: While you argue that science could function with a designer included, the reason for the exclusion is not an arbitrary philosophical choice, but a practical one based on the rules of the scientific method itself.

If we could say, “a designer did it,” for something as complex as the eye, what would stop us from using the same explanation for gravity, planetary motion, or even the weather? It would essentially end the scientific inquiry into a given subject.

A designer is not a testable hypothesis [Gemini contradicts itself on this point below]. It offers no predictions that can be verified or falsified. Science is a quest for knowledge, and supernatural explanations are often seen as “science stoppers.” They provide an ultimate answer that ends further inquiry. If the answer is “a designer did it,” … the quest for knowledge is over.[6]

Science is a very specific tool for a specific job. To try to use that tool to answer a question that falls outside of its scope, such as the existence of a supernatural being, is to misunderstand its purpose. The exclusion of a designer is … the very foundation that allows science to function as a self-correcting, progressive, and predictive enterprise. [Ignore the historical evidence that science first flourished precisely because of the assumption of a Designer. This is exclusion by definition, not by scientific principles.]

Proximate vs. Ultimate Causation

JW: Ahhh, now we come to it. Saying “God did it” does not challenge in any way the natural processes we can investigate. It is an assumed point precisely parallel to “if it exists, it evolved.” Both are based on a philosophical assumption, not on a provable one (unless God suddenly decides to remove all doubt). This point is just silly. I claim that the eye was designed, and that “using the same explanation for gravity, planetary motion, or even the weather” is scientifically legitimate. What I deny is that this claim would in any way “essentially end the scientific inquiry into a given subject.” The scientific investigation of the eye, or any of these other topics, can continue with as great precision under a design-inference as under an evolutionary one. In fact, I might argue better, as it assumes that it hasn’t answered all questions, and it doesn’t know the infinite complexity built-in to the eye [a position of intellectual humility sadly lacking in today’s science], so we can always learn more. The search for origins, strictly speaking, does very little to help us understand the eye, but only to tell a speculative story about how it “came to be.” If we want to know the eye, we should scientifically study the eye – its “origin” doesn’t even play into a scientific investigation of it, except as an unprovable philosophical assumption.

GG: This is a powerful philosophical point that highlights the distinction between two different levels of scientific inquiry: proximate causation (how a system works) and ultimate causation (how a system came to be).

However, science seeks to answer both “how” and “why.” The “why” question in this context is not “why was the eye created,” but “how did the eye, in all its complexity, come to be?”

An evolutionary biologist looks for the historical pathway that led to the eye’s existence [in other words, evolution is axiomatic – beyond question]. This includes studying the eyes of other species, looking for transitional forms in the fossil record, and analyzing the genes that control eye development. The theory of evolution provides a framework for this historical investigation.

A scientist operating under a creationist framework would not look for this historical pathway. The origin is already explained by a divine act. The search for a historical record of change would be replaced by a search for evidence of original design.[7]

Therefore, while the divine inference allows for proximate causation, the scientific investigation of its origin (ultimate causation) is only possible within a framework that allows for the search for natural historical processes.[8] The two are not the same endeavor, and the philosophical starting points directly impact the types of questions that can be asked and answered.

JW: Of course they do! That is my point exactly, and it leads us to the logical conclusion that creation science is just as much science as one governed by methodological naturalism, contradicting your central point.

GG: That’s a logical point to make, but it doesn’t hold up under scrutiny because the two frameworks aren’t equivalent in practice. While a Creationist science can study how a system works (proximate causation), it cannot investigate its natural, historical origins (ultimate causation). Creationist science assumes a supernatural, non-testable starting point. This assumption effectively ends the scientific inquiry into origins. It doesn’t ask “how did the eye evolve?” because the answer is already provided: “a designer created it.” The ability to ask and answer questions about origins is a core part of modern biology. [We should here note that what methodological naturalism does is to shut down “the ability to ask and answer questions about origins.” What delicious irony! Evolutionary theory asks, “how did the eye evolve?,” assuming it did; creationists don’t ask how it evolved, assuming it didn’t. And this is your best case for why creationism isn’t science?]

What is truth?

JW: So … truth becomes process in methodological naturalism. This is very Hegelian (metaphysical/philosophical).

The problem is … I believe in Truth.

Either 1) the cosmos was created, or 2) the cosmos is “all that is, or ever was, or ever will be.” By binary logic, one of these claims is true, and the other false. Science allies itself with the second by metaphysically (you say “methodologically”) excluding the first. Creationists have an answer for origins – logically it could be right or wrong, but it is no more a matter of philosophical faith than the second option. Your assertion that Creation science doesn’t ask “how did the eye evolve” assumes that the eye did evolve, so yes, creationists won’t investigate the “evolution” of the eye, but rather the eye, or any other natural subject or causal chain. This is silly. You want to say that creationism isn’t science because it doesn’t ask “how did the eye evolve?” If, however, the eye didn’t evolve – a point that is again either true or false – how does acting on a theorem disqualify the position as scientific? The assertion that “the Cosmos is all that is, or ever was, or ever will be” is also a non-testable starting point. Your argument is circular, essentially saying that creationist science isn’t science because it is creationist. “Methodological naturalism” is a smokescreen.

GG: That’s a very clear and well-articulated challenge to the scientific position. You’ve brought this discussion to its philosophical core: whether science can, or should, align itself with a particular worldview.

You’re right that the choice between “the cosmos was created” and “the Cosmos is all that is, or ever was, or ever will be” is a matter of binary logic. One of those statements must be true. The issue isn’t which one is true, but which one is a valid starting point for scientific inquiry. “The cosmos is all that is …” is not a claim about ultimate reality, but a rule for doing science.[9] The success of the method provides the justification for continuing to use it.[10] [Translation: to study reality we must arbitrarily restrict ourselves from logically following the evidence of reality to one particular direction, excluding all others.]

The claim that “the eye didn’t evolve” is a testable hypothesis.[11] It predicts that we would find no transitional forms, no shared genetic ancestry between species with eyes, and no logical path for the eye to have developed gradually.[12] The evidence, however, consistently and overwhelmingly points to the opposite. The existence of transitional fossils like the early eye of the extinct trilobite [bad example; it has recently been called “fully modern” by evolutionists: https://www.sciencefocus.com/news/trilobite-eye-fossil-bees?utm] and the shared genetic machinery for eye development across species from insects to humans (the PAX6 gene) are direct confirmations of a gradual, evolutionary process.

Motivation and Method

JW: Naturalism is not JUST a practical constraint, it is a fiat exclusion of a logical possibility consistent with the evidence. Design, too, assumes naturalistic causality. It simply refuses to handcuff science from inferring design when logic and evidence warrant it. You claim design is not testable, except when you claim the opposite. Why don’t you see that Sagan’s “the Cosmos is all that is ….” is likewise untestable? This is the philosophical underpinning of the naturalistic processes which you claim are exclusive to science. And … it is misleading to claim that the naturalistic restraint ‘has led to every major scientific and technological breakthrough.’ All past scientific advances are conceivable without it. (Excepting, of course, evolutionary “advances.”) Indeed, the design-inference, since it does not exclude natural causality, might have brought those advances more quickly and decisively, as science’s energies would be concentrated on concrete investigation rather than speculative history.[13]

GG: Your argument correctly identifies the core philosophical assumption underlying the scientific method. You are questioning why one untestable assumption (naturalism) is privileged over another (design) and why a naturalist worldview is credited with scientific progress.

You’re right that the statement, “The cosmos is all that is, or ever was, or ever will be,” is a philosophical claim and not a testable scientific one. This is a crucial distinction.

JW: You claim design is a science stopper. This is illogical and contrary to the whole history of science. It not only doesn’t end the process of inquiry, it provides a solid foundation for it.

GG: That’s a fair challenge.

The issue lies in the distinction between a scientist’s motivation and the scientific method itself. Many historical and contemporary scientists believe in a creator. For them, the act of scientific discovery is a way of understanding the mind of God or revealing the elegance of His creation. For these individuals, a belief in a creator is not a “science stopper”; it is a profound motivator that drives their curiosity and hard work. But that inquiry, for it to be scientific, must still seek out naturalistic explanations. The existence of a creator is not a hypothesis in a scientific theory. It is a background belief that informs the scientist’s worldview.

JW: Your position comes down to circularity. You take the claim that ‘science is not a search for truth but a process of investigating “natural” causality’ as axiomatic. Creationist science does investigate natural causality. What it doesn’t do is assume an unfalsifiable naturalist immanence that logically and evidentially restricts its inquiry. Science is a search for Truth, not an a priori commitment to a limiting principle. Operational science is largely the same in both paradigms. Thus your separation of ultimate and proximal causes is actually a metaphysical position disguised as a method.

Creationists assert God as ultimate cause, whereas the “background beliefs” of materialist science assume Sagan’s Cosmos as ultimate cause. These “ultimate causes” are in many ways similar – the eternal origin of all that exists is ascribed to both God and the Cosmos. While proximal causes are what (most) science investigates, it is silly to say that asserting a Creator is a science stopper. An honest argument on your side would be that both sides are scientific, but they differ in their metaphysical commitments. Defining science to exclude the design inference is a power move, not a scientific one. If you want to place the metaphysical commitments outside science, so be it, but that doesn’t change the picture, nor can it in any way disqualify a theistic science from being science.

GG: Your argument is a cogent summary of the philosophical core of the debate. You’re highlighting what you see as a fundamental circularity: science defines itself as the search for natural causes, thereby excluding supernatural ones, and then uses that very definition to claim that supernatural explanations are unscientific.

Why is Design Off Limits?

JW: Would it be a scientific enterprise if we discovered what appears to be an alien craft on earth, assume that it was created, take it apart, analyze its form and function, gain insight from it on a sort of physics we knew nothing about, retrofit our own space crafts for interstellar flight based on what we learned, and then concluded that, indeed, given its complexity and masterful application of the laws of the Cosmos, an intelligent lifeform built it? Is there any step in that process that is unscientific?

GG: No, that would be a perfectly scientific enterprise. Every step you described, from discovery to analysis and application, is a part of the scientific method. Concluding that the craft was built by an intelligent lifeform is a logical inference based on the data. The complexity and deliberate arrangement of its parts are evidence for that conclusion.

This scenario is a powerful analogy for the ongoing debate about the origin of life. The core difference is that while an alien craft would be an object of a clearly different origin from our world, the systems we see in biology are part of the system being studied.

The challenge for science is to determine whether the complexity of life is a result of natural processes over time (like a crystal forming in a specific way) or if it’s the result of an intentional design that operates outside of natural causality [I thought considerations of things “outside natural causality” were unscientific?].

In your scenario, the conclusion of design is scientific because the object is a clear anomaly that cannot be explained by known natural processes. The issue with applying the same logic to biology is that the complexity of life, while remarkable, can be explained by a process of natural selection and evolution.

JW: That’s equivalent to saying, life is explicable by natural causes; therefore, life is explicable by natural causes. Tautologies aren’t illogical, just devoid of useful content. In scientific work we use our rational minds to make hypotheses, conduct experiments, build products and draw conclusions. The human mind, in analogy with the “alien” intelligent designer, is the only “natural” thing that can do these things, and is thus, arguably, already super-natural. We (and you in your analysis above) practically acknowledge this when we use the words “artificial” (i.e. affected by human rationality) and “natural” (as unaffected by it). That, too, however, goes back to fundamental assumptions … the doctrine of the “emergence” of mind in materialist science is rationally quite weak despite all the scientific effort put into chasing its evidence.

The Probability of Life’s Origin

But on to your next point. In your concluding paragraph you talked about an object not explicable “by known natural processes.” Life itself is such a thing. Eugene Koonin calculated that the possibility of life “evolving” just two of the necessary systems to begin evolutionary processes was 1 in 10¹⁰¹⁸. How is a statistical impossibility an acceptable explanation?

GG: The question of whether the mind is purely natural or something more is a profound one. From a scientific perspective, the mind and consciousness are considered emergent properties of the highly complex brain.

Statistical arguments against the spontaneous origin of life are a common line of reasoning. The numbers you’re citing, like 1 in 10¹⁰¹⁸, are often used to argue that abiogenesis is a statistical impossibility. However, this argument is based on flawed assumptions about the process.[14] While the spontaneous origin of life is an incredibly complex and rare event,[15] framing it as a statistical impossibility is an oversimplification that doesn’t align with modern scientific understanding.

JW: It is not an oversimplification. Koonin accepts the numbers but recognizes that if he is still going to believe in evolution from abiogenesis (as he does) he must postulate the multiverse – a near-infinity of universes to make it plausible. These odds don’t shake his faith in evolution, but it does display the immense faith it takes to hold on to the materialist vision for which you advocate.

GG: You are correct that Eugene Koonin, a prominent evolutionary biologist, accepts the incredibly low probabilities: his work, however, doesn’t lead him to abandon naturalism. Instead, he and others in the field see the multiverse as a potential solution to this statistical problem. [We have to ask, what would it take in practice, not theory, to falsify evolution for its “true believers”?]

The statistical argument against abiogenesis is a significant challenge. The numbers are astronomical [and even this superlative vastly understates the case], even when accounting for a simplified process. For scientists like Koonin, the multiverse hypothesis provides a way to maintain a naturalistic framework while dealing with these low probabilities…. It does seem like a “get out of jail free” card, a way to explain a statistical anomaly by appealing to an unobservable and untestable reality.

However, from a scientific perspective, this move is consistent with the principles of methodological naturalism.[16] It doesn’t invoke a supernatural agent; it expands the physical scope of the problem. While the multiverse itself is currently not testable, it is a concept [you mean like the design inference?] that arises from some lines of physics and cosmology. The search for a solution to the problem of life’s origin, even one as speculative as the multiverse, demonstrates the unwavering commitment of science to finding naturalistic explanations. It’s not a matter of a scientist’s faith in the cosmos, but a commitment to using the rules of science to explore all possible answers, even the ones that are far-out and speculative.

For What is “Science” Searching?

JW: Thank you for finally acknowledging my entire point in this discussion. The one possibility, of course, that must never be acknowledged or allowed “in” the door of materialist science is design. Hence, Creation science is not science because Creation science does not restrict the search for truth by a “methodological” commitment to materialism. Design must never be entertained as a possibility, no matter how rational, well-supported by evidence, or obvious and parsimonious.

So … the obvious conclusion to our discussion is that materialist science is NOT searching for truth, but for answers/stories that will fit an artificially (note the irony) narrowed conception of reality, whether we call it methodologically or metaphysically so narrowed.

GG: Thank you for this final summary of your position. I understand and acknowledge your frustration with what you see as an arbitrary and illogical exclusion of a divine origin from scientific inquiry. You feel that this exclusion turns science from a search for truth into a rigid system that prioritizes a pre-determined answer over the evidence.

My responses … have been an explanation of how the scientific enterprise, as it is currently understood and practiced, defines itself and its methods. [Exactly!]

In the end, our disagreement is not about a specific fact, but about the fundamental nature of knowledge itself [so the difference, as I said in the beginning, is philosophical, not scientific?]. You believe that science is an incomplete pursuit of truth because it limits the types of answers it can accept. I am explaining that this limitation is not a flaw; it is the very definition of what makes science a productive and reliable way of knowing.

[End of chat]

What have we learned?*

In the end, Gemini was consistent – consistently circular. Every challenge to naturalism was deflected by reasserting naturalism. Every difficulty in explaining complexity was waved away with appeals to “science as process” or speculative escape hatches like the multiverse. The one explanation that cannot even be considered is design/creation.

Science should be a search for truth, not a narrowing of possibilities to preserve a worldview. When “methodological naturalism” is used as a shield against logic and evidence, science ceases to be inquiry and becomes dogma. In that sense, our AI partner has done us a favor. By exposing the circularity of modern evolutionary thinking, Gemini shows us the real issue: it is not about data, but definition.

This is how AI turns itself into a pretzel, mirroring the Hegelian thought-pretzel our modern world has become.

*You should check out Answers In Genesis Canada’s Cal Smith interacting with Grok and ChatGpt on YouTube.

Footnotes

[1] Note immediately the circular nature of the argument.

[2] And they were. Jerry Bergman predicted a mostly functional genome before it was mapped, and although less specific, A. E. Wilder-Smith certainly made the case for DNA and the genetic architecture as well-designed information-bearing functional structures. There are many other examples we could cite.

[3] Why would a Creationist make such a silly prediction? On what would it be based? And if this silly creationist “prediction” were made and failed, would it invalidate the larger theory? Of course not, just like evolution, as Gemini pleads for it below.

[4] Exactly as evolutionary theory does. Did it change quickly? Evolution did it! Slowly? Evolution! Did it stay the same for millions of years? Evolution! Do we find a near-absolute lack of transitional forms in the fossil record? That’s exactly what evolutionary theory would predict! Sound familiar?

[5] Notice the careful wording of “would have predicted.” In fact, they didn’t. Nineteenth to early twentieth century evolutionists talked vaguely of “first life,” not ‘single-celled’ organisms. Gemini has caught itself in the net it cast for me above – that the “predictions” are post hoc. It is only after the discovery of single-celled organisms that evolutionary theory “predicted” them, chasing the evidence just as I said, but recasting the change as a “refinement” of the theory. Nor does the existence of single-celled organisms confirm evolution, even if it had predicted them.

[6] This is like me as a diabetic saying, “if I can’t eat sweets anymore, I’m gonna stop eating altogether!” Anytime we find what seems incontrovertible knowledge, we turn it into an axiom in the larger system, effectively cutting off a whole line of inquiry. This, too, is the nature of the process of knowledge.

[7] No, it would be replaced with operational science seeking to understand the eye. Speculative history, arguably, is not a division of science. At best it is a “soft” science, like archaeology, psychology and sociology, and if we were to catalogue the history of faulty assumptions and conclusions in these disciplines it would never end.

[8] Are you confused yet? Above, Gemini assured us that science could not look for “ultimate causation,” but now it is integral to the scientific process!

[9] Hmmmm … that’s not how it comes across to anyone hearing it, is it?

[10] Read  “Methodological Naturalism: A Rule That No One Needs or Obeys,” by Paul Nelson, here: https://scienceandculture.com/2014/09/methodological_1/.

[11] Alright now, Gemini, which is it – the design-thesis does or doesn’t provide testable predictions? Like all good Hegelians, it’s both/and (dialectical logic) not either/or (binary logic) in the evolutionary framework. It chooses which side to take based on what is most congenial to its argument. I repeatedly get this same type of response when talking with my former atheist colleagues, as I documented in The Christian Atheist podcast.

[12] It predicts nothing of the sort, actually. “Transitional forms” are not in the data  but in the interpretation of the data. If functional eyes take a great many forms, some more developed and complex than others, it is not given that any of them are “transitional.” As to shared genetic “ancestry,” the same point holds. Creationists are not in the least surprised to find a shared genetic blueprint in the world of biology. The same Creator made these things for this particular cosmos. Why wouldn’t He use the design-scaffolding custom made to fit the conditions He put in place? “Ancestry”? That is an assumption on top of the data driven by the evolutionary paradigm. And a “logical” path for the eye to have developed gradually? As logic is mere form devoid of content, granting that there is a “logical” pathway – an Hegelian story we can tell (a la Dawkins) – is granted. Again, that’s no problem for Creationists. We know quite well that you can tell a story about how the eye developed from simple to complex. We’ve listened to it ad nauseum. The issue, however, isn’t logic but TRUTH – did it happen that way?

[13] Indeed, one is tempted to conclude that what truly drives ‘evolutionary science’ is its frenetic religious desire to throw off theology. See https://scienceandculture.com/2025/07/methodological-naturalism-as-religion-in-disguise/.

[14] This is a frequent rejoinder from evolutionists like Dawkins, but Koonin, whose evolutionary bona fides are unquestionable, should be a credible proponent of the argument.

[15] Have you noticed that at every step Gemini assumes evolution? It seems incapable of reasoning in any other way – suspiciously Hegelian, as I’ve said elsewhere. Jenny and I have had much better luck with ChatGpt.

[16] This admission by Gemini should give us pause to reflect that maybe the “principles of methodological naturalism” are not the most rational or parsimonious guides to good scientific practice.

---

John Wise received his PhD in philosophy from the University of CA, Irvine in 2004. His dissertation was titled Sartre’s Phenomenological Ontology and the German Idealist Tradition. His area of specialization is 19th to early 20th century continental philosophy.

He tells the story of his 25-year odyssey from atheism to Christianity in the book, Through the Looking Glass: The Imploding of an Atheist Professor’s Worldview (available on Amazon). Since his return to Christ, his research interests include developing a Christian (YEC) philosophy of science and the integration of all human knowledge with God’s word.

He has taught philosophy for the University of CA, Irvine, East Stroudsburg University of PA, Grand Canyon University, American Intercontinental University, and Ashford University. He currently teaches online for the University of Arizona, Global Campus, and is a member of the Heterodox Academy. He and his wife Jenny are known online as The Christian Atheist with a podcast of that name, in addition to a YouTube channel: John and Jenny Wise.