Nobel prize 2024 - magic just got less wriggle room

OK, but then we also forget about relevance to the argument, because the argument relies on proximity.


That doesn't sound right, unless n - k data points (protein foldings) are identical with some other data point, which I don't think you'd suggest is the case here. [Edit: OK, but being more charitable, you probably don't intend that the representations are identical. I think I see what you mean.]


You seem to be putting up a straw man here. Maybe you can cite something relevant from the DI and explain how this AI result affects it.

Maybe it’s a straw man, but it could suggest that fewer semi-random walk iterations are needed to arrive at a biologically functional protein, given the constraints imposed by protein folding and natural selection.

There are others (who actually know more about biology than I do) who seem to make similar inquiries.

The recent development of artificial intelligence provides us with new and powerful tools for studying the mysterious relationship between organism evolution and protein evolution.

just as thermodynamic laws and collective order can emerge from the random motions of molecules, it may be possible to discover a “collective” trend consistent with the increase of organismal complexity

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9550990/

This part of our research I find particularly exciting. When we want to go back in time to look at evolution, the way we commonly do this is to compare the sequences between proteins in different species. By doing that we can try to guess what that sequence looked like in the evolutionary past.

https://deepmind.google/discover/blog/tr...n-of-life/

I thought I used the term “cluster” before you did but I may be wrong.

Anyway it’s now proven that “clusters” exists in protein space. Forget about proximity - If data has clusters it simply means that instead of being represented by n data points the same data can be represented by k data points with k< n =>protein space is (overwhelmingly) smaller than the narrative by DI.

Hold on - we are talking about the capability of an undirected random search to find certain rare viable and adaptive protein configurations that cumulatively produce an adaptive life promoting protein. Any random variation (say a cosmic ray particle) is vastly more likely to make a change that pushes the sequence farther from the needed target sequence or even out of the intrarelated "cluster" you are referring to, than it is to stay close within the cluster. Random is random. 

The protein space that random variation operates on is n in your notation. 

It takes outside intelligence (forbidden in Darwinism) to determine that the sequence is getting closer to producing a desired shape and to then make changes that are predicted to get closer to the target or onto the target. The statistical methods used by AI systems to predict 3D protein shape and which involve clustering (and also the AI systems themselves) are indeed invented by intelligence - ours.

Hold on - we are talking about the capability of an undirected random search to find certain rare viable and adaptive protein configurations that cumulatively produce an adaptive life promoting protein. Any random variation (say a cosmic ray particle) is vastly more likely to make a change that pushes the sequence farther from the needed target sequence or even out of the intrarelated "cluster" you are referring to, than it is to stay close within the cluster. Random is random. 

The protein space that random variation operates on is n in your notation. 

It takes outside intelligence (forbidden in Darwinism) to determine that the sequence is getting closer to producing a desired shape and to then make changes that are predicted to get closer to the target or onto the target. The statistical methods used by AI systems to predict 3D protein shape and which involve clustering (and also the AI systems themselves) are indeed invented by intelligence - ours.

So your position would be that the AI search was only successful because human intelligence used some a prior considerations to prune the search space?

And this would then translate to the necessity of some kind of mind to assist in evolution?

So your position would be that the AI search was only successful because human intelligence used some a prior considerations to prune the search space?

Not quite. I think the AI sysytems designed to do this are only partially using search algorithms - they are using a large learning data base of successfully solved by humans folding problems and are mainly extrapolating the data accumulated from these successful determinations of known amino acid sequences correlating with certain known protein 3 dimensional folding structures to try to find perhaps related amino acid sequences producing other perhaps related protein 3-dimensionsal folding structures (the structures desired for medical/biochemical reasons by the designers). It's a matter of utilizing such observed correlations to predict new correlations, without for the most part actually determining the details of the actual complicated deterministic atomic and chemical process that produces a certain folding shape from a given sequence. The latter brute force analytical approach is apparently not practical on a large scale. Of course the laborious analytic approach would directly be able to make the folding predictions by knowing exactly how each molecular bonding 3 dimensionally effects the folding through the atomic chemical/electrical forces involved.

And this would then translate to the necessity of some kind of mind to assist in evolution?

A
I think so. In order for intelligent humans in medical/pharmaceutical research to produce needed new proteins of particular shapes (which determine their biochemical actions in the body), they need to somehow generate the needed new amino acid sequences. They avoid an impractical brute force analytical approach which would directly calculate the resulting foldings. 

They also avoid an impractical random search algorithm approach (impractical because of the widely scattered nature of the search space). So they use an AI neural net/learning data set approach to predict foldings based on the training data configurations. This resembles the failure of the Darwinistic evolutionary theory to be able to find new required proteins by undirected semi-random walk RM + NS , and the resulting necessity of injecting outside intelligence into the process.