Biocomputing is one of the strangest frontiers in emerging technology. This is made possible by the fact that our neurons perceive the world and act using electrical signals, the same language as computers. Human brain cells grown in large numbers on silicon chips can receive electrical signals from a computer, try to understand them, and respond.
More importantly, they can learn. Our first encounter with this concept was in the DishBrain project at Monash University in Australia. What do you think it felt like? Dr. Frankenstein At that moment, researchers grew about 800,000 brain cells on a chip and placed them in a simulated environment. I watched this horrible cyborg monster learn how to play ping pong in about 5 minutes.. The project quickly received funding from the Australian Army and was spun off into a company called Cortical Labs.
cortex lab
When I interviewed Brett Kagan, Chief Science Officer at Cortical Labs, he said that even at this early stage, human neuron-enhanced biocomputers could learn much faster, using much less power than today’s AI machine learning chips, while also providing “more intuition.” “It shows,” he said. , insight and creativity.” After all, our brains consume just 20 watts to run nature’s most powerful necktop computer.
“We’ve done tests on reinforcement learning and found that it’s chalk and cheese in terms of how quickly you have to see before the system starts to show meaningful learning, like a biological system,” Kagan said. , even though it is now basic and janky, it still outperforms the best deep learning algorithms people have created.”
One downside, aside from the obviously tricky ethics, is that the “wet software” component must be kept alive. This means feeding them, watering them, regulating their temperature and protecting them from bacteria and viruses. Cortical’s record in 2023 was about 12 months.
Final Spark
We subsequently covered a similar project at Indiana University. Before plugging electrodes into brain cells, the researchers caused them to self-organize into three-dimensional ball-shaped “Brainoware” organisms. And Swiss startup FinalSpark has begun using dopamine as a treatment. Compensation mechanisms for Neuroplatform biocomputing chips.
If this is your first time hearing about Brain-on-a-Chip, pick your chin up from the floor and read a few of those links. This is truly amazing work. And now Chinese researchers say they are taking it to the next level.
The MetaBOC (BOC for Brain-on-Chip) project brought together researchers from Tianjin University’s Haihe Institute of Brain-Computer Interaction and Human-Computer Integration with another team from Southern University of Science and Technology.
Tianjin University
This is open-source software designed to act as an interface between a Brain-on-a-Chip biocomputer and other electronic devices, giving the brain organism the ability to perceive the world through electronic signals and operate with any given control. . Learn how to access and master specific tasks.
The Tenjin team says they are using ball-shaped organoids, as is the Brainoware team in Indiana. That’s because their three-dimensional physical structure allows them to form more complex neural connections, just like in our brains. These organoids are grown using low-intensity focused ultrasound stimulation, which the researchers say appears to provide a better foundation for intelligence.
The MetaBOC system also attempts to meet intelligence by using AI algorithms within the software to communicate with the biological intelligence of brain cells.
The Tenjin team specifically mentions robotics as an unifying target, providing the rather silly image above as if they were intentionally trying to undermine the credibility of their work. The team says the Brain-on-a-Chip biocomputer can now learn how to drive the robot, figure out how to control it, and attempt tasks such as obstacle avoidance, goal tracking, and how to use its arms and hands to grasp various objects.
Since the brain organism can only ‘see’ the world through the electrical signals provided, you can theoretically train yourself how to pilot your Mini Gundam in a fully simulated environment to make the most of its drops and drops. It does not endanger or interfere with the body’s intelligence engine.
Tianjin University
Now, to be clear, the pink lollipop-style brain organism fully exposed in the robot image above is a mockup (a “demo diagram of a future application scenario”), not a brain-controlled prototype. Perhaps the image below from Cortical Labs is a better representation of what this kind of brain on a chip might look like in the real world.
cortex lab
But either way, if we were to build tiny robots with appropriate sensing and locomotion capabilities, there’s no reason why human brain cells couldn’t soon try to learn how to drive them.
Science and technology, including projects like Neuralink, which aims to connect high-bandwidth computer interfaces directly to the brain, MetaBOC, which grows human brain cells into computers, and the burgeoning AI industry attempting to overtake the best of biological technology. It’s a wonderful time. Intelligence using strange facsimiles made entirely of silicon.
As science and technology confront the limits of our understanding, we have no choice but to become philosophical. Is the dish brain conscious? Is AI conscious? Both may become indistinguishable from sentient beings at some point in the near future. What are the ethics when that happens? Are biological intelligence and silicon-based intelligence different?
Kagan said in a wide-ranging interview: “Suppose these systems develop consciousness. I think it’s very unlikely, but let’s say it happens. Then we have to decide whether it’s actually ethically right to test it. We should test it in conscious animals. “Because it is tested, it is legal to eat animals without any worries.”
To be honest, I can hardly believe what I am writing. Humanity has begun to build cyborg minds that can intelligently control machines using the physical components of their own minds.
But this is life in 2024. We are accelerating at full speed toward a mysterious technological singularity, the point at which AI intelligence surpasses our own and begins to develop much faster than humans can. The point at which technological progress, already proceeding at an unprecedented pace, accelerates towards a vertical line and we become completely unable to control it.
What a great time to be alive and not just a bunch of cells connected to a chip on a plate. Well, as far as we know.
Source: Tianjin University
