I still disagree with the edited version. An improbable explanation can be the most plausible explanation - that’s a relative ranking. The argument doesn’t establish that a designing intelligence is a plausible explanation except in cases where the presence of a potential designing intelligence is unproblematic. That is emphatically not the case of the origins of earthly life. Indeed, there’s a real possibility that a designing intelligence requires “FTDI” - human beings do.
Yes, that’s also a good point. The slippage from the term “improbable” in P2 to “most plausible” in the Inference is doing a lot of heavy lifting here.
Yeah, I don’t think I can quite agree with this. P1 states only that a designing intelligence is a possible producer of FDI (the soundness of which is a separate matter), and P2 states that other causes are improbable (dito). Even if “plausible” were to be defined as “possible and not improbable” – which I may be willing to grant for the sake of argument, but has not been explicitly so declared – we would still be lacking a step stating that designing intelligence is not improbable before we can say that it is plausible by that definition. Maybe that was somehow meant implicitly, but I think if one is going to put our arguments into syllogistic form like that, then one should fully expect any reader to apply the full rigor of logic when they interpret it. Midhun had all the time in the world to write it up in a rigorous, valid way, and either failed to do so or chose to not do so.
After a quick read of that paper, here is what I learned. Please point out if missed any important thing:
The g3p minor coat protein present in the wild-type bacteriophage fd-tet has 3 domains: D1, D2, and D3. The D2 domain plays a major role in the adsorption of the phage to the host cell’s F-pilus, thus initiating the infection. Researchers modified the phage genome by replacing the D2 coding region with a sequence that codes for a random soluble polypeptide of 139 amino acid residues. They retained D1 and D3 as is. They called the modified phage fd-RP. The infectivity of fd-RP was found to be many orders of magnitude lower than the wild-type fd-tet. Then the researchers introduced random mutagenesis to the replaced sequence, and after multiple rounds of mutation and selection, they found mutants that had better infectivity than fd-RP.
I don’t understand how the results of this study dispute my statement that no known natural processes can direct/bias the arrangement of individual nucleotides towards ‘x’ (provided, ‘x’ is a function or outcome that can be encoded only by a minute fraction of digital strings in the total combinatorial space of strings).
I think you misunderstood my point. By the term “arrangement of individual nucleotides,” I meant the joining of nucleotides by chemical bonds and thus the initial forming of sequence.
We are discussing the origin of systems that occurred in the historical past, which were not witnessed by any humans since humans did not exist at that time. Nevertheless, the positive aspect is that we can infer the action of a non-human designing intelligence, even if we have no knowledge of the nature or identity of that non-human intelligence. I’ll illustrate this with an example. You might be familiar with the Arecibo message, which was transmitted into outer space. The message consisted of 1679 binary digits encoding basic information about humans and Earth. Now, imagine a hypothetical scenario in which humans receive a similar information signal from an extraterrestrial source. Even without knowing anything about the nature or identity of the source that produced such a signal, we would still infer the action of a [non-human] designing intelligence.
The reasoning we would be employing can be outlined as:
Premise 1: Information signals (FDI) can be generated by the action of a designing intelligence.
Premise 2: The probability that this kind of information signal (FDI) originated by non-intelligent causes is extremely low.
Inference: The activity of an extraterrestrial [non-human] designing intelligence is the most plausible explanation for the origin of this message.
This addresses @Gisteron too👇
Is it? I apologize. It could be due to English not being my native language.
“Digital data” is a well defined concept and genetic information encoded in DNA is an example of naturally occuring digital information.
Read this and this.
To explain that, the number of random trials that are plausible in a realistic scenario are to be considered. I avoided explaining that in my initial comment for the sake of brevity. Since you asked let me explain it using a hypothetical scenario:
Imagine a person purchases a box containing 3D models or blocks representing English alphabets. Each alphabet occurs 100 times within the box, resulting in a total of 2600 blocks (26×100). These blocks possess a distinctive feature—magnetic strips attached to their left and right sides. Consequently, when brought close, the blocks automatically join together. A block can connect to a maximum of two other blocks, one on the left and one on the right. Upon arriving home, the person unboxes the blocks onto the floor of their room and then leaves the house, unaware of what transpires with the blocks in his absence.
Within the house, there are other family members who are intelligent and proficient in English. They can arrange the blocks, and there are also pet animals present. The presence of running fans can cause the blocks to move, and due to their magnetic properties, they can join if in close proximity. Upon the person’s return home, he observe numerous strings of various lengths formed from the connected blocks. In this scenario, any meaningful english words that are 15 letters or longer, if formed, can be considered examples of FDI. Because those words satisfy all the three criteria.
(1) Digital information since the words are formed by discrete alphabets.
(2) Here the function ‘x’ = meaningful english word of atleast 15 letters long.
Number of strings that can encode ‘x’ is 10⁴ (since there are roughly ten thousand such words in english language).
Total combinatorial size of strings = 26¹⁵ + 26¹⁶ + 26¹⁷ + …
Fraction of functional strings in combinatorial space= 10⁴/ (26¹⁵ + 26¹⁶ + 26¹⁷+ …)
This is a very small fraction
Maximum number of random trials possible that can produce strings containing atleast 15 letters = 2600/15 =185
Comparing both values we know that the number of random trials is inadequate in order to achieve ‘x’.
(3) It is the relative arrangement of discrete alphabets within the string that imparts upon its function ‘x’ and no natural and non-intelligent causes known to us can direct/bias the arrangement of individual units towards ‘x’.
All strings that encode ‘x’ (here ‘x’ = meaningful english word of atleast 15 letters long) are examples of FDI. As I said in the beginning, FDI is an indicator of Design. The argument can be outlined as
Premise One: A designing intelligence that knows English language can produce meaningful english words.
Premise Two: It is improbable to get those FDI strings through non-intelligent causes. (Because FDI strings are improbable by definition. See criteria number 2)
Inference: The activity of a designing intelligence that knows English language is the most rational or the only plausible mechanism behind the formation of those FDI strings.
Yes, but then look at his original definition of FDI (my emphasis):
1 Like
The argument appears to be based on a relative evaluation of probability, but there is no probability stated for “FDI generated by the action of humans” (or any sort of designing intelligence). Technically there are no probabilities anywhere as premise 2 just says it’s “improbable”.
It we are to say A is more likely than B, we need the probabilities of both, with justification for those numbers, so that we can assess the rational basis (if any) for those given numbers.
That’s why that argument fails.
Hi🙂
That’s true.
1 Like
Even if we credit his definition, there’s a difference between “no natural process known to us” and “no natural process”. How do you go from there to “must be a supernatural process, which must be Jesus”?
Human voice is an example of analog data.
I find no problem with that. Instead of honing in on a single sentence within a book, I’ve decided to examine the entire book, which, in this context, represents the entirety of the biological information encoded in biopolymers necessary for the origin of a modern-type cell.
I’ve already described what I think is implausible to arise without the action of a designing intelligence. Please read my initial comment again.
Given that evolution through built-in mechanisms doesn’t conflict with the concept of Intelligent Design, I recommend providing a list of such evidence. These evidences will greatly enhance my future discussions on built-in mechanisms, and I would greatly appreciate your contribution😉
I will cite a few studies
For eg, a 2001 study estimated that “…a library of ≈10^24 members would have been needed to obtain AroQ mutases”.
The team admitted that there is “…a low probability of finding catalysts” or
“…extremely low incidence of enzymes in completely unbiased libraries…”
A 2003 article by Jack W. Szostak stated:
“…the extreme rarity of functional sequences in populations of random sequences (typically 10^−10 to 10^−15 for aptamers and ribozymes isolated from random RNA pools)…”
Also see this 2017 study that estimated the fraction of sequences that can fold to a target protein structure. This study estimated the fraction of around 10 different protein structures. (I acknowledge that in some cases, distinct protein structures can perform similar functions)
The fraction of sequences capable of encoding a function can fluctuate depending on the specific function. For eg, this fraction will be lower for functions involving slow catalysis or weak binding compared to those associated with high-affinity binding or rapid catalysis. In my definition of FDI, I stipulated that the functions being considered should have a low likelihood of occurring within the sequence space.
But what is the probability? We need a number, not just the “a designing intelligence can”.
If you’re going to say A is more likely than B to produce X, we need the probabilities of both. You can’t just say B is improbable therefore the unknown probability of A is preferred. Tell us how you get the probability of A.
You’re skipping a step: How did the “designing intelligence” come to known English language in the first place?
If it hasn’t been taught English by previous generations, it can’t just magically know how to design English sentences. Learning is a combination of trial and error (testing variants of behaviors (analogous to mutations), such as enunciating certain sounds and evaluating the response those sounds elicit, analogous to natural selection), and deliberate tutoring (having information basically copied into your brain, analogous to replication of the genome).
A designing intelligence is basically a learning algorithm like evolution. It can’t just magically solve problems evolution can’t solve if it did not benefit from something analogous to the process of evolution: Copying successful information gathered from previous experiences to the next generation, and by testing lots of different variants against the environment and getting some sort of feedback.
There’s a simple test to show that a designing intelligence isn’t this magical problem solver that can do stuff evolution can’t. If there is no smooth fitness surface that can give you informational feedback that tells you that you are getting closer to the real solution, then you can’t design your way towards the solution, such as a long digital string. The example is a large random password. Only one solution will do, and any deviation from the complete correct string will not work. Nothing tells you that you are getting closer to the “target” password. This is the ultimate problem that evolution can’t solve, that design is supposed to be able to solve. The case where the “solution” is incredibly rare and there is no feedback. Can design solve it then? Nope. All you can do is brute force random guessing for millions of years until you hit the correct one by chance.
Your mind, your “designing intelligence” is no better than evolution. It’s just another learning algorithm that benefits from previous generations of trial and error and copying of successful results. If you have not been given this benefit, you are as worthless as brute force guessing.
That is why your argument fails, and always will fail.
1 Like
Here’s another 2003 Nature article that conveniently repeats the same mistake.
Btw, do you understand the difference between digital and analog data?
That’s due to my not-so-perfect English (it’s not my native language). I didn’t mean ‘exactly one’ x; I meant ‘at least one’ x.
So, with that clarification, let me pose the question: Regardless of whether the sequences contain just a single ‘x’ or multiple 'x’s, how does it affect my argument?
I didn’t get it. Elaborate.
This part:
Demonstration of what? The capabilities of built-in mechanisms or the evolution from a [hypothetical] proto-cell to a modern-type cell? You misread my statements. I suggest revisiting my initial comment and my reply.
“would have had”? Great. You could have also stated “evolutionary process without the aid of built-in mechanisms would have had created the built-in mechanisms and later would have had created a modern-type cell”.
What an effective and powerful way of debating!
1 Like
I’ll illustrate the problem with your argument using an example. You might be familiar with the Arecibo message, which was transmitted into outer space. The message consisted of 1679 binary digits encoding basic information about humans and Earth. Now, imagine a hypothetical scenario in which humans receive a similar information signal from an extraterrestrial source. Even without knowing anything about the nature or identity of the source that produced such a signal, we would still infer the action of a [non-human] designing intelligence.
The reasoning we would be employing can be outlined as:
Premise 1: Information signals (FDI) can be generated by the action of a designing intelligence.
Premise 2: The probability that this kind of information signal (FDI) originated by non-intelligent causes is extremely low.
Inference: The activity of an extraterrestrial [non-human] designing intelligence is the most plausible explanation for the origin of this message.
If we replace the broader term “designing intelligence” with a narrow term “humans” like you insist, the argument will look like:
Premise 1: Information signals (FDI) can be generated by the action of humans.
Premise 2: The probability that this kind of information signal (FDI) originated by non-humans is extremely low.
Inference: The activity of humans is the most plausible explanation for the origin of this extraterrestrial message!!
Correct. Hence the random string that constitutes the D2 domain evolved towards acquiring the biological function of increasing infectivity of phage. That’s a biological function. The string of amino acids that replaced the D2 domain was a random string. And evolution found and improved it’s function.
I don’t see why the nature of the chemical bond that links nucleotides together makes any difference (digital information isn’t defined by phosphodiester bonds as opposed to amide bonds). The bonding (or what causes the bonds to form in the first place) is irrelevant. It’s the fact that the string is “random”, and that there is enough function in random strings that it just takes a few mutations to find one of relevance, and that further mutations subject to selection can improve it, that shows you’re wrong.
That’s a direct test of the postulate that evolution can’t create new functional digital information in a nonfunctional sequence. It was demonstrated that it can.
1 Like
I clearly wrote in my initial post that evolution mediated by built-in mechanisms can generate FDI. I only doubt the creative potential of evolution in producing ‘x’ when ‘x’ is the biological information encoded in biopolymers required for the origin of modern-type cells. I defined ‘x’ like that because in the [hypothetical] first self-replicators or proto-cells, built-in mechanisms would not be there to aid evolution.
You didn’t say what the problem with the above argument is.
1 Like
That’s also a good point.
And, as it happens, he is not even consistent on that: