I'm no expert in the biochemistry of the brain, but surely thoughts and memories depend as much upon the fluid medium of the brain as well as the more permanent structures of its neurons—the complex cycles and oscillations in dissolved ions, neurochemicals, &c. that cease to exist when the brain is killed and preserved with glutaraldehyde or whatever. and yes, they're proposing exactly that: https://nectome.com/the-case-for-glutaraldehyde-structural-encoding-and-preservation-of-long-term-memories/ glutaraldehyde, H(C=O)(CH₂)₃(C=O)H, is a somewhat less toxic alternative to formaldehyde in the preservation of soft tissue by chemical cross-linking.
so you preserve the physical structures of the brain—so what? what does that get you, when you've lost the complex chemical medium in which a living brain functions? and you can't get that back because the glutaraldehyde preservation has reacted with and thus destroyed much of the fine chemical structures that were present in the living brain. it's cross-linked all the proteins, denatured them. the idea of somehow reconstructing thoughts and memories from these dead and permanently altered tissues seems a bit like thinking that you can reconstruct Windows 10 by meticulous microscopic examination of an Intel microprocessor that happened to be running Windows 10 before it was removed for examination. that's not an analogy I want to push too far because a brain and its neurochemicals aren't the same as a computer's hardware and software, but still—organic life and organic thought are so dependent upon transient phenomena, chemical oscillations that are destroyed upon death, that I don't get why this wild notion is supposed to be plausible.
anyway, if Sam Altman wants to convince me that he ought to be able to preserve his brain, first he should convince me that there's anything worthwhile in it.
~Chara
context below the cut, for people who feel that their life is enhanced by context
https://openworm.org/ is an academic research project aimed at advancing our understanding of neuroscience by simulating a flatworm
this particular type of flatworm has something like 120 brain cells. something in that close neighborhood, we don't remember exactly, but every worm has the same number of brain cells and the same layout of them and the exact function of every single one has been documented extensively
there is a thing you can teach a flatworm to do: you can catch it in a loop of wire and wait for it to escape. there is a particular trick to escaping. you can tell whether a flatworm has learned the trick by how long it takes to get out. you can also look at its brain, because learning the trick changes the way the neurons connect in a certain precise, documented way
with all of this knowledge, some might think it should be easy to identify how memories are stored in the flatworm's brain, and even simulate it
that turns out not to be the case
our understanding is that neuroscientists don't seriously believe that the mere topology of neurons is the entirety of memory storage, there's more going on, we just have only the barest guesses at what
and that is why we don't worry about brain uploads with current technology
