3 ms·
> It's incredible that the information necessary to create a human is just about 750 MB uncompressed Hold up, isn't the point of this article that genes do not
by Terr_ 2y ago
> It's incredible that the information necessary to create a human is just about 750 MB uncompressed
Hold up, isn't the point of this article that genes do not have all the information?
The, er, bootup environment of a freshly-fertilized human egg normally provides a lot more than merely protection and raw materials for nine months. Likely a lot of required parameters, and definitely a lot of important tuning optimizations.
> For example the very specific shape of the scapula bone or fear of spiders...
There were some studies a decade back about mice inheriting fears of certain smells from the father, I wonder if anyone discovered the mechanism (or disproved the effect) by now.
- abecedarius 2y agoTo a programmer it's reminiscent of a self-hosting compiler that comes with working source and binary, but not the sequence of bootstraps. "Trusting trust" makes a nice example of the hastiness in taking the source to be all you really need to know. Similarly, metacircular interpreters that leave evaluation order ambiguous. Should be very interesting to quantify what's needed for a "clean bootstrap" of a minimal model organism that can reproduce.
- agumonkey 2y ago> Hold up, isn't the point of this article that genes do not have all the information? Yeah it is. The informational aspect (and the precision level) Levin and his friends are looking for is the main factor here. Cells are not just tissue building blocks, they're agents, limited but still, themselves with a lot more capabilities than we generally assume. I believe multiple fields are mature enough (biology, computational biology, optics) to care and analyse this and make a new step in our understanding of complex organisms.