World's First Bioprocessor Uses 16 Human Brain Organoids, Consumes Less Power (tomshardware.com) 48
"A Swiss biocomputing startup has launched an online platform that provides remote access to 16 human brain organoids," reports Tom's Hardware:
FinalSpark claims its Neuroplatform is the world's first online platform delivering access to biological neurons in vitro. Moreover, bioprocessors like this "consume a million times less power than traditional digital processors," the company says. FinalSpark says its Neuroplatform is capable of learning and processing information, and due to its low power consumption, it could reduce the environmental impacts of computing. In a recent research paper about its developments, FinalSpakr claims that training a single LLM like GPT-3 required approximately 10GWh — about 6,000 times greater energy consumption than the average European citizen uses in a whole year. Such energy expenditure could be massively cut following the successful deployment of bioprocessors.
The operation of the Neuroplatform currently relies on an architecture that can be classified as wetware: the mixing of hardware, software, and biology. The main innovation delivered by the Neuroplatform is through the use of four Multi-Electrode Arrays (MEAs) housing the living tissue — organoids, which are 3D cell masses of brain tissue...interfaced by eight electrodes used for both stimulation and recording... FinalSpark has given access to its remote computing platform to nine institutions to help spur bioprocessing research and development. With such institutions' collaboration, it hopes to create the world's first living processor.
FinalSpark was founded in 2014, according to Wikipedia's page on wetware computing. "While a wetware computer is still largely conceptual, there has been limited success with construction and prototyping, which has acted as a proof of the concept's realistic application to computing in the future."
Thanks to long-time Slashdot reader Artem S. Tashkinov for sharing the article.
The operation of the Neuroplatform currently relies on an architecture that can be classified as wetware: the mixing of hardware, software, and biology. The main innovation delivered by the Neuroplatform is through the use of four Multi-Electrode Arrays (MEAs) housing the living tissue — organoids, which are 3D cell masses of brain tissue...interfaced by eight electrodes used for both stimulation and recording... FinalSpark has given access to its remote computing platform to nine institutions to help spur bioprocessing research and development. With such institutions' collaboration, it hopes to create the world's first living processor.
FinalSpark was founded in 2014, according to Wikipedia's page on wetware computing. "While a wetware computer is still largely conceptual, there has been limited success with construction and prototyping, which has acted as a proof of the concept's realistic application to computing in the future."
Thanks to long-time Slashdot reader Artem S. Tashkinov for sharing the article.
They already made a movie about this... (Score:2)
Re: (Score:2)
Re: (Score:2)
Given the training source (the internet) I think your point is valid
Re: (Score:2)
obligatory Star Trek (Score:2)
I RTFA and I call BS (Score:4, Informative)
I RTFA and I call BS.
TFA says, "Bioprocessors like this consume a million times less power than traditional digital processors."
But if you drill down to their citation for this factoid, they are comparing the energy consumption of a cluster of neurons, not to a computer doing the same computation, but to a computer doing a detailed simulation of a cluster of neurons.
That is intentionally misleading and is a meaningless comparison. It makes no more sense than comparing a calculator adding 2 plus 2 to a human doing the same by modeling the transistors in the calculator by manually solving the differential equations determining their saturation threshold.
If these neural organoids are really so efficient, why lie about it?
sellers suck (Score:2)
There's really not much else to be said.
Re: (Score:3)
I RTFA and I call BS.
TFA says, "Bioprocessors like this consume a million times less power than traditional digital processors."
But if you drill down to their citation for this factoid, they are comparing the energy consumption of a cluster of neurons, not to a computer doing the same computation, but to a computer doing a detailed simulation of a cluster of neurons.
That is intentionally misleading and is a meaningless comparison. It makes no more sense than comparing a calculator adding 2 plus 2 to a human doing the same by modeling the transistors in the calculator by manually solving the differential equations determining their saturation threshold.
If these neural organoids are really so efficient, why lie about it?
Computer folks think about inputs, output, and logical functions. It's not clear what these biological things are processing.
Furthermore, assuming that some controllable logical processing is possible, this would only be viewed as the ALU. As we are finding out with AI and GPUs, the logical processing is the easiest part. Data movement is the true challenge, both in terms of power and performance.
And these biological things need electronics to provide stimulus and convert back to electrical signals, and
Re: (Score:1)
It's a comparison a lot of people make. It would take more computing power (and energy) than exist in the universe to simulate a few neurons at the fundamental particle level, a large supercomputer to do it with reasonable biochemical fidelity, a small neural network by 2015 standards for spiking neurons, or a fairly ambitious high school CS project in the 90s for a reasonable equivalent without spiking.
What level is necessary? Evidence is pointing towards the light end, OMG brains are the most amazing thin
Re: (Score:2)
What level is necessary?
The apt comparison is to a computer performing the same task.
Can a human brain add two integers more efficiently than a computer? Of course not.
Can a human brain distinguish a photo of a dog from a photo of a cat more efficiently than a computer? No, it's closer, but a computer still wins. A brain uses 25 watts and takes a second, while my laptop uses 30 watts and takes a millisecond.
Re: (Score:3)
There are still tasks that a fairly ordinary human brain can do pretty well and electronic computers can only do very slowly, if at all. I expect that will change in the next ten or twenty years so much that any reasonable person will come around to your opinion, but a LOT of people still draw some bright line that they assume mere "machines" can't cross.
Biological brains are much more energy efficient for reasonable estimates of their computational capacity. They are worth studying. Neurons are picky basta
Re: I RTFA and I call BS (Score:1)
It doesnâ(TM)t take a second for a human to do it. It is so fast as to be nearly automatic, and is integrated with a hundred thousand other considerations as well. What might take a full second is internalizing the instruction, cuing the photo, formulating the answer, and then issuing a spoken response. But if you just glance around the room you are in, you will automatically identify lots and lots of things, without even having to formulate the query.
Re: (Score:2)
Kind of reminds me of how video game emulation is managed. Don't emulate the target architecture instruction per instruction, that's far too expensive for all but the simplest architectures. Instead, emulate things at the API level using native graphics calls.
Re: (Score:2)
It's even more than that. If you want to make a NAND gate you can do it with neurons, transistors, tubes, water valves, wood, metal, lego, etc. We've demonstrated over the last few centuries that you can make computing devices out of pretty much anything. Over the last 80 years we've demonstrated that putting enough NAND gates together gives you a machine that can compute anything computable, and "computable" is a very far reaching classification.
The argument for neurons is like arguing that Space Invaders
Re: (Score:3)
Agreed. (Score:2)
I came here to bring up the same idea. It's fun to compare things in a way that makes you feel good about your work, but 'end to end' or it didn't happen.
Just like when comparing the environmental effects of gas cars versus electrics. How long do they last, what kind of maintenance do they need, how expensive was it to make them in the first place, etc.
How much work is put into maintaining the environment for the organoids. Or growing them in the first place? How much space do they take up? How accurat
Re: (Score:2)
It’s a combination of cool, but early stage, research, combined with a perceived need to hype potential future capabilities to attract funding. Fairly common academic strategy where “thing” exists as a potentially enabling technology, but is not anywhere near commercialization.
In this particular case, what they have developed is actually pretty cool from multiple perspectives: long-lasting neuronal organoids in microfluidic devices, a remote orchestration system to do wetware expts, and an
Creepy (Score:3)
This may only be 16 random cells but at what point are there enough cells working together to have some level of consciousness and self awareness?
The problem is we really don't know; our understanding of these things is very rudimentary.
If they stopped at 16 I'd just shrug it off but obviously they're going to keep adding more and more cells. Shouldn't a self aware entity built from human brain cells have rights?
Re: (Score:1)
We probably should prohibit L5 pyramidal neurons in these systems. The microtubles in these neurons are responsible for consciousness.
When they are free to vibrate at 25GHz (IIRC) we have consciousness. When you introduce anesthetics those resonances are dampened and we get modern surgery.
Everyone agrees about that.
If you then wonder what that sort of oscillator at that frequency is tuning people become /remarkably/ unhinged. Some extremely smart scientists are studying it with little fanfare.
We probably s
Re: (Score:3)
I hope you get help
Re: Creepy (Score:2)
Re: Creepy (Score:2)
Far from everyone agrees about that. Itâ(TM)s a hypothesis that a couple of famous biologists and physicists have got behind, but itâ(TM)s not at all proven.
Re: (Score:2)
Let me guess we should classify them as fetuses.
Re: (Score:2)
everything human is sacred. human organoids have feelings, you insensitive clod!
Re: (Score:3)
This may only be 16 random cells
No. It is sixteen Organoids [wikipedia.org].
An organoid is a cluster of cells large enough to have some functions of an organ.
TFA doesn't say how many cells are in each organoid and is extremely light on other technical details as well.
Re: (Score:2)
Oh good point.
So there are some number of cells here > 16. Anyway, at some point there may be enough to be self aware. We still don't know what that number is. I expect they will hit that number eventually as more cells = faster compute, so "hey why not? What could possibly go wrong?"
Re: (Score:2)
If they're being honest, I'm sure there are more than 256 cells. And I extremely strongly suspect that organoids based on chicken cells would work just as well.
Re: Creepy (Score:3)
Re: (Score:2)
Shouldn't a self aware entity built from human brain cells have rights?
cows are completely self aware and we definitely should give them basic rights (eg to not be mass tortured and slaughtered), but that would be incompatible with eating them, so we don't.
weird how you consider awareness to be the boundary, yet it's only relevant if it's made out of human stem cells. it's all just spiced up carbon.
anyway, those blobs are property, and whatever they want to get they will have to either peddle, lobby or fight for, and good luck.
Re: (Score:1)
Surprisingly, humans and cows are not the same.
Anyway, I don't think you've seen an American cow. I doubt they're really self aware.
So you're cool with th4 idea of taking this all the way forward and having fully grown humans with real brains in 50 years we can use as slaves and sex toys with no rights because hey they're just carbon and we grew them, so fuck em if they want rights; they didn't come from a vagina so who cares?
Re: (Score:2)
so a handful of organoids on a petri dish is more aware than a cow. you're just doubling down on your ignorance and delusion. why am i not surprised ...
these are cows, get a clue:
https://www.youtube.com/watch?... [youtube.com]
nobody here is talking about hypothetical artificial sex slaves in 50 years, nothing even hints at that, that's just your feverish brain making up a straw man to justify your delusion of human exceptionalism.
there are indeed ethical considerations around this kind of research going forward but i'm a
Re: Creepy (Score:2)
I'm not sure you've ever met an American. Are you sure they're self-aware? I mean, look at Newark, and that's just for openers.
Re: (Score:1)
I'm pretty sure I know a few Americans but I've never been to Newark so I can't address that subset from personal experience.
Re: (Score:2)
We shouldn't torture them, but that's not to say we shouldn't slaughter them. Just that we should do it by surprise and as much as possible without pain. Nominally, that's our protocol. Unfortunately, actually following the protocol is a bit expensive, so lots of corners get cut.
All that live must die. (Just not all at the same time.) That's one of the facts of life. But pain can be minimized, and contentment maximized.
Remember, if people didn't want to eat cows and chickens, they would be extremely ra
Re: Creepy (Score:2)
This may only be 16 random cells but at what point are there enough cells working together to have some level of consciousness and self awareness?
You only need a handful.
Probably worth some precision there... (Score:2)
Personally being 'self-aware' isn't a magical boundary for me (I can't prevent the suffering of all self-aware things, and even some plants might be self-aware). Nor is 'coming from a human'.
But yeah, it brings up questions like you posed...
If it can't feel pain or want anything... does it need protection? I'm guessing these are parts of the neo-cortex, and I don't believe emotions/feelings exist there.
If an organism can be engineered to want to do what you want it to do... is that slavery? Even if it
Better Off Ted (Score:2)
I hope they give it time off.
https://www.youtube.com/watch?... [youtube.com]
I AM ALIVE, Proteus (Score:2)
Call to a tech support company (Score:4, Funny)
Do not we already have this? (Score:1)
"a million times less" (Score:1)
Enlarge your GENIUS! (Score:2)
Not only does it have 1/6000th THE POWER REQUIREMENTS of traditional processors, it also comes in THESE FUN COLORS for the modders out there. Resistance is foolish, the future *is* wetware. You will become acclimatized to it, trust us.
- A proto-Borg marketing director, "Bio/Tel Corporation"
Re: (Score:2)
Trust us, all the way to your receptacle, er... cube, er... office.
:-)
But seriously, the concept of "bioboxen" is pretty cool and I'm happy that someone is developing it.
But I'll pass on becoming one, myself.
Click Freight (Score:1)
Yeah, bullshit.