This sounds like the beginning of a Horror movie, but also colors me hopeful for a future where people born with my mothers disease (SCA) will have some kind of cure to halt or revert the loss of brain tissue ?
Thank you for your comment reminding me of how the world can hopefully become better for people if we can develop and use technology in an ethical way.
You would think if we could get something as advanced as brain tissue to grow, then we could also get something as simple as hyaline cartilage to regenerate. I’ve been following the research on it for 14 years now since I had a knee injury from running that caused focal cartilage defects, and the rate of progress in the field is absolutely glacial.
The latest research articles in the medical journals are still talking about BMP-2 growth factors and mesenchymal stem cells being the future of cartilage repair, but I now treat this a bit like nuclear fusion in the sense that we always seem to be about 10-20 years away from productionizing it.
From the article:
> The tissue still has enough structure to hold together, but it is soft enough to allow the neurons to grow into each other and start talking to each other,” Zhang said. “Our tissue stays relatively thin, and this makes it easy for the neurons to get enough oxygen and enough nutrients from the growth media.
Seems like this would be great for hyaline cartilage as well, as the tissue is avascular and receives nutrients primarily via diffusion. I don’t have a background in the field, but it would be nice if we could take some of these cutting-edge techniques and attempt to apply them to more mundane things like joint degeneration.
But aren’t these two completely different challenges?
Successfully growing the structure of brain tissue is one task
But regenerating enough cartilage with enough strength and robustness to support a joint is another challenge altogether?
Sort of how chip R&D is famous for some exotic structure like CFETs or graphene transistors being created, but reliable and scalable wafers with a chance of being economically viable is an altogether different feat?
Why do you think hyaline cartilage is simple vs. neurons?
A neuron is probably a more specialized version of something a lot of cells can do (https://archive.ph/WXcSE).
I think neurons have the benefit of staying relatively differentiated once they mature, whereas for most other tissue, introducing/inducing stem cells runs a higher risk of promoting uncontrolled growth/teratoma/cancer.
They didn't print brain tissue, they kinda printed scaffolds for it to grow.
"The researchers aimed to construct layered neural tissue in which neural progenitor cells (NPCs) mature ..."
Don't trust this kind of source for titles.
About cartilage. That is a very specific configuration of materials in not a random shape. Also, it must want to attach to the bone by itself, and source its nutrients from synovial liquid.
Sounds too complicated for what I know we can do.
About human cells growing in a petri dish. Thats pretty much cancer right there. In such an alien environment, healthy cells expect something to kill them, and if enough changes happen without that, they will suicide.
Growing hearts in pigs is much closer to something viable.
>“The tissue still has enough structure to hold together, but it is soft enough to allow the neurons to grow into each other and start talking to each other,”
While everyone's worried about sentient AI annihilating humanity, some guy with a PrusaBio Mk7+ accidentally prints a superintelligence in his basement.
At the scale of individual neurons, or clusters of neurons, would the performance of human brain cells significantly differ from that of a pig? or house fly?
Those would be some interesting LLM performance benchmarks to see.
We still know so little about the actual micro scale and macro scale structure and connection graph of the brain. This is a neat advance in tissue integrity but it is not a brain. It's like melting sand into a pile that holds itself and considering it an advance in computing.
We still know so little about the actual micro scale and macro scale structure and connection graph of the brain. This is a neat advance in tissue integrity but it is not a brain. It's like melting sand into a pile that holds itself and considering it an advance in computing.
Still cool through.
On a side note, the brain is amazing. The more we learn about AI the more amazing it is that the brain emerges from a relatively sparse instruction set, self assembles from widely available local resources, and becomes a general purpose computing engine able to simulate enough of the world internally to promote survival and propagation of its instruction set.
Well, it's not organized when you print it. It self-organizes as you train it. Yes, it is like melting sand into a pile, but this is magic sand that can re-organize itself over time.
Neurons have the ability to self-organize into an information processing device, though. Researchers have taken rat neurons, grown them in a petri dish, and taught them to play (a simplified version of) pong [0]. A better vat might genuinely allow currently-out-of-reach computing (ignoring the ethical issues).
Maybe grow nanotube like transceivers that the brain prefers to grow on to then you can train it to make classifiers. Then make the doomers really piss their pants
28 comments
[ 3.1 ms ] story [ 71.5 ms ] threadThe latest research articles in the medical journals are still talking about BMP-2 growth factors and mesenchymal stem cells being the future of cartilage repair, but I now treat this a bit like nuclear fusion in the sense that we always seem to be about 10-20 years away from productionizing it.
From the article:
> The tissue still has enough structure to hold together, but it is soft enough to allow the neurons to grow into each other and start talking to each other,” Zhang said. “Our tissue stays relatively thin, and this makes it easy for the neurons to get enough oxygen and enough nutrients from the growth media.
Seems like this would be great for hyaline cartilage as well, as the tissue is avascular and receives nutrients primarily via diffusion. I don’t have a background in the field, but it would be nice if we could take some of these cutting-edge techniques and attempt to apply them to more mundane things like joint degeneration.
But aren’t these two completely different challenges?
Successfully growing the structure of brain tissue is one task
But regenerating enough cartilage with enough strength and robustness to support a joint is another challenge altogether?
Sort of how chip R&D is famous for some exotic structure like CFETs or graphene transistors being created, but reliable and scalable wafers with a chance of being economically viable is an altogether different feat?
A neuron is probably a more specialized version of something a lot of cells can do (https://archive.ph/WXcSE).
I think neurons have the benefit of staying relatively differentiated once they mature, whereas for most other tissue, introducing/inducing stem cells runs a higher risk of promoting uncontrolled growth/teratoma/cancer.
"The researchers aimed to construct layered neural tissue in which neural progenitor cells (NPCs) mature ..."
Don't trust this kind of source for titles.
About cartilage. That is a very specific configuration of materials in not a random shape. Also, it must want to attach to the bone by itself, and source its nutrients from synovial liquid. Sounds too complicated for what I know we can do.
About human cells growing in a petri dish. Thats pretty much cancer right there. In such an alien environment, healthy cells expect something to kill them, and if enough changes happen without that, they will suicide.
Growing hearts in pigs is much closer to something viable.
While everyone's worried about sentient AI annihilating humanity, some guy with a PrusaBio Mk7+ accidentally prints a superintelligence in his basement.
Soylent green was ahead of its time.
Those would be some interesting LLM performance benchmarks to see.
Still cool through.
Still cool through.
On a side note, the brain is amazing. The more we learn about AI the more amazing it is that the brain emerges from a relatively sparse instruction set, self assembles from widely available local resources, and becomes a general purpose computing engine able to simulate enough of the world internally to promote survival and propagation of its instruction set.
[0] https://www.cell.com/neuron/fulltext/S0896-6273(22)00806-6
Growing it is easy. Using it is hard.