Intelligent design and "design detection"

Well, similarity to human genes is an odd measure of sequence conservation, since human genes are not primitive. All it’s really measuring is genetic distance, which has something to do with conservation but more to do with time since divergence. And why choose humans as the reference sequence? Pinnacle of creation, or what?

Indeed. But that’s the definition that gpuccio provided that Szostak did not.


Prolly. Why ignore human polymorphism if you’re claiming to measure the prevalence of function in sequence space? The whole thing makes no sense.

I only vaguely remember gpuccio’s arguments, but didn’t he produce a graph purporting to show increase in functional information over time, which was basically a line showing time since divergence from humans? Do you have a link to his very sciency publication?

Lucky you!

Again, these claims are “not directly detecting the presence of design, but merely inferring design”, so again:

Calling any of this “design detection” would be a bit like calling the presence of fire engines and hoses a “smoke detector”.

Likewise, a police officer might detect the presence of cyanide by the smell of bitter almonds – they would not detect its presence, but rather infer or deduce it, from the the fact that an autopsy showed that the murder victim died of cyanide poisoning.

Indirect inference or deduction tends to involve more assumptions, and therefore be less certain than direct detection – so mislabeling “Design deduction” as "design detection* is a form of exaggeration.


You’re thinking of this impressively scientific-looking figure concocted by Gpuccio:

The textbook cargo-cult science post that originated the whole thing:

One among numerous threads we had on it here:

I like how Gpuccio begins his post with the pseudoscientific rhetorical device that someone here taking the time to look over his post is an “acknowledgement.” Better pump up that “we’re being taken seriously/getting acknowledged” number as much as possible.

1 Like

Yes, exactly. The “big jump” is between the similarity of cephalochordate proteins to human and the similarity of shark proteins to human. Well, duh. The two divergences are very far apart in time, and it only looks like a sudden jump because the x axis is arbitrary and doesn’t represent time or anything else. And the measure of “information” is just silly, as I had vaguely recalled. I do wonder what that graph would look like if the x axis did represent time.

More importantly, if this is @Giltil’s idea of ID science, that says a lot.


The focus on human genes is probably heavily influenced by various religious beliefs, although non-religious people are just as susceptible to ego as anyone else.

From my understanding, the argument boils down to the probability of producing a highly conserved protein sequence. Their claim is that only that sequence could have carried out that function which is evidenced by how highly conserved it is. What they don’t seem to consider is contingency through time. Protein interactions can cement the importance of a specific sequence after it has appeared, especially if multiple proteins interact with the same target. It’s a bit like being amazed that the hole in the ground exactly fits the shape of the water in the hole.


But Gpuccio isn’t even measuring conservation. He’s measuring similarity to the human sequence. Not in any way the same thing.

1 Like

If an orthologous sequence is 99% similar between sharks, frogs, birds, and humans then that would be a highly conserved sequence.

1 Like

For sure. But that isn’t what’s happening there. The comparison is only between some species and humans. Chimps, by that measure, would have almost every sequence highly conserved, while sharks would have very few.


I think I am confusing things in my mind. I am remembering arguments based on a single or handful of strongly conserved genes, but that doesn’t appear to be the argument being cited here. Apologies.

IIRC, Gpuccio was plugging the bit values from BLAST alignments into Szostak’s equation(s) and claiming “information” or something like that.

Clearly he doesn’t understand Szostak’s definition. But the graph that’s been shown doesn’t even use an equation. It just reports the bit values.

The graph is meant for people like @Giltil, not for anyone who has the first frigging clue. So, as we can see, it is a very good graph in terms of achieving its intended goal.

What you call a “flaw” maybe simply that the analogy is not exactly the same as the original claim. If it was exactly like the original claim would it be an analogy?

What would be your argument that it is “flawed” or a poor analogy?

In one case we know the existence of the designer in the other we need to infer existence. While direct evidence of the designer would be preferred does lack of direct evidence really nullify the inference?

Exactly that.

While direct evidence of the designer would be preferred does lack of direct evidence really nullify the inference?


Some people may disagree with this as it would disqualify the use of abductive and inductive reasoning :wink:

I was one of many who participated in the long and very tedious discussion of gpuccio’s functional information criterion. In fact I wrote two posts at TSZ (here and here) that led to discussion on that. In the end we found that unlike Szostak he was assuming that all sequences that had less functionality than the one which we saw, were so much less functional that there was no evolutionary path to the observed sequence. This is way different from Szostak and Hazen’s FI. They were not even attempting to use FI to discuss whether evolution could reach that sequence.

BTW, I am failing to access TSZ, there is some 404 problem.


The links work fine for me right now.

1 Like