Brain-in-a-Jar Biocomputers can Now Learn to Operate Robots

Living brain cells integrated into organoid-on-a-chip biocomputers are now capable of learning to control robots through MetaBOC, an open-source intelligent interaction system. This extraordinary initiative aims to relocate human brain cells into artificial bodies.

Biocomputing stands at one of the most surreal frontiers of emerging technology, enabled by the shared language of electrical signals between neurons and computers. Human brain cells, cultured in large quantities on silicon chips, can interpret electrical signals from a computer, process them, and respond accordingly.

Even more significantly, these biocomputers are capable of learning. This concept first emerged in the DishBrain project at Monash University in Australia.

Cultivating Brain Cells on Chips

In what could be likened to a Dr. Frankenstein scenario, researchers cultured approximately 800,000 brain cells on a chip, placed it into a simulated setting, and observed as this startling cyborg entity mastered the game of Pong in roughly five minutes. The project swiftly received funding from the Australian military and evolved into a company named Cortical Labs.

Biocomputers with Human Neurons

In an interview with Brett Kagan, Chief Scientific Officer at Cortical Labs, he revealed that even in their early stages, biocomputers enhanced with human neurons seem to learn significantly faster and with far less power compared to modern AI machine learning chips. They also exhibit greater “intuition, insight, and creativity.” Remarkably, our brains, using just 20 watts, operate as nature’s most powerful processors.

Kagan explained, “We’ve conducted tests against reinforcement learning and found a stark contrast in performance. The biological systems, despite being rudimentary and imperfect, still surpass the best deep learning algorithms in terms of how quickly they require fewer samples to begin meaningful learning. It’s quite astonishing.”

However, there are drawbacks, including ethical concerns and the need to maintain the “wetware” components. These biological elements must be nourished, hydrated, temperature-regulated, and shielded from germs and viruses. As of 2023, Cortical Labs achieved a maintenance record of about 12 months.

Advances in Brain Cell Organization and Neurochemical Integration

We’ve also reported on similar projects, such as at Indiana University, where researchers allowed brain cells to self-organize into a three-dimensional spherical “Brainoware” organoid before inserting electrodes, and at Swiss startup FinalSpark, which has begun using dopamine as a reward mechanism for its Neuroplatform biocomputing chips.

If this is your first encounter with brain-on-a-chip technology, it’s worth taking a moment to absorb the astonishing nature of this work. Chinese researchers are now advancing this field even further.

The MetaBOC (Brain-On-Chip) project combines efforts from Tianjin University’s Haihe Laboratory of Brain-Computer Interaction and Human-Computer Integration with teams from the Southern University of Science and Technology.

Enabling Brain-Organoid Interaction with Electronic Devices

The open-source software connects brain-on-a-chip biocomputers to electronic devices, allowing brain organoids to interpret signals, interact with their environment, and learn tasks.

The Tianjin team uses spherical organoids, like Indiana University’s Brainoware, which form complex neural connections similar to human brains. These are grown with low-intensity ultrasound stimulation, enhancing their intelligent capabilities.

The MetaBOC system uses AI algorithms to communicate with the brain cells’ biological intelligence, integrating artificial and biological intelligence.

The Tianjin team, while presenting humorous images, identifies robotics as a key application. They claim a brain-on-a-chip biocomputer can now control a robot, learning tasks like obstacle avoidance and object manipulation.

Because the brain organoid perceives the world through electrical signals, it can train in a simulated environment to operate a robot, reducing risks to its biological component.

Conceptual Mockups vs. Practical Brain-on-Chip Systems

To clarify, the brain organoids depicted in the robot images above—resembling exposed, pink lollipops—are conceptual mockups, illustrating future application scenarios rather than actual brain-controlled prototypes. A more realistic depiction of what these brain-on-chip systems might resemble in practical terms is shown in the image below, provided by Cortical Labs.

It’s increasingly likely that human brain cells could soon be integrated into small robots, learning to operate them effectively.

This era marks remarkable advancements in science and technology. Initiatives like Neuralink aim to directly interface high-bandwidth computer interfaces with the brain, while projects such as MetaBOC grow human brain cells within computers. Simultaneously, the AI industry is striving to replicate and potentially surpass biological intelligence with silicon-based models.

As these frontiers expand, profound philosophical questions emerge. Are dish-brains or AI systems conscious? Could they become indistinguishable from sentient beings? What ethical considerations arise between biological and silicon-based intelligences?

“In my extensive interview,” Kagan reflects, “even if these systems were to develop consciousness—which I consider unlikely—ethical dilemmas would emerge, akin to those surrounding testing on conscious animals or consumption habits.”

It’s astonishing that humanity is now using the physical components of the mind to develop cyborg intelligences capable of precise machine control.

In 2024, we’re accelerating toward the technological singularity, where AI could surpass human intelligence and drive unprecedented advancements faster than humans can manage.

What an exhilarating time to be alive—though our existence may not simply be as cells wired to a chip in a lab dish, at least not as far as current knowledge suggests.

Read the original article on: New Atlas

Read more: ZenRobotics 4.0 Enhances Intelligence in Waste Sorting Automation