What is Organoid Intelligence?

What is Organoid Intelligence?

Organoid intelligence or OI is an emerging multidisciplinary field concerned with the development of biological computing models or biocomputers using three-dimensional cultures of brain cells. Some observers have noted that OI can either replace artificial intelligence or progress further the different goals and subfields of artificial intelligence.

Understanding Organoid Intelligence: History, Principles, Purpose, and Applications

History of Organoid Intelligence

The field of organoid intelligence draws principles and practices from other fields such as biology, computer science, engineering, and bioengineering. It is specifically an offshoot of biological computing which centers on using biologically-derived components such as cells or genetic molecules including DNA and proteins to perform computations.

Note that biological engineering or bioengineering was a term coined by German-born British scientist and media presenter Heinz Wolff in 1954 at the National Institute for Medical Research. It was during the same year that the field was recognized as its own branch at the institute for the first time. The United States also launched its first biological engineering program at the University of California in San Diego in 1966.

The field of biological computing or biocomputing emerged around the 1990s as a concept that took notes from bioengineering and other fields such as biology and computer science. American computer scientists Leonard Adleman developed DNA computing in 1994 as a field concerned with using molecular biology hardware as opposed to electronic computing.

Furthermore, in 1999, American biomedical engineer W. L. Ditto created one of the first biocomputers. It was composed of leech neurons at the Georgia Institute of Technology and it was capable of performing simple addition. Several biocomputers capable of performing simple computations were developed and demonstrated during the 1990s and early 2000s. More sophisticated biological computers emerged since.

Andre David Endy headed a team of bioengineers at Stanford University announced in 2013 that they created the first biological equivalent of a transistor. It was called a transcriptor and it was made from DNA and RNA that controlled the flow of enzymes. These enzymes corresponded to signals. Other experiments worked on storing data in DNA.

Researchers from the Biodesign Institute at Arizona State University and Wyss Institute for Biologically Inspired Engineering at Harvard University developed a biological computer inside the E. Coli bacteria. The team specifically called their biocomputer “ribocomputer” because it was composed of RNA inserted in E. Coli using CRISPR gene-editing technology. One of its demonstrated capabilities is that it was able to store video data.

Their experiments were published in Nature in 2017 and they demonstrated the potential of using living cells for computing tasks and storing information. It was also the first time that data was encoded in a living organism although previous experiments and even practical applications have demonstrated data storage using DNA as a medium.

Nevertheless, building on the developments from the aforementioned fields, organoid intelligence emerged as an offshoot of sorts. Researchers at the Australia-based company Cortical Labs showed in 2022 that brain cells grown on a chip can quickly learn to play the video game Pong. The process was called synthetic biological intelligence. Organoid intelligence also emerged as a novel field during the same year.

The participants of the First Organoid Intelligence Workshop held between 22 and 24 February 2022 made the Baltimore Declaration. It encouraged the international scientific community to explore the potential of human-brained based organoid cell cultures to advance our understanding of the brain and unleash new forms of biocomputing.

Working Principles

The Pong experiment involved growing a network of brain cells in a Petri dish. These cells were cultured cortical neurons dissected from the brains of embryonic mice and human stem cells reprogrammed into neurons. They were embedded in high-density micro-electrode array chips. These chips can record and simulate the electrical activity of these cells.

Researchers Kagan et al. called their creation the “DishBrain” and its capabilities are a demonstration of “synthetic biological intelligence” because it showcased the possibility of teaching or training cultured brain cells using predictable electrical stimuli and unpredictable electrical stimuli. These cells began learning while improving their performance in playing the Pong video game through constant exposure to electrical stimuli.

Smirnova et al. published a paper in 2023 that described how they grew balls of human brain cells from stem cells in a Petri dish to create a self-assembling aggregate of neurons. They called it brain organoid or “brainware” and further described it as living artificial intelligence hardware that harnesses the computational power of three-dimensional neural networks.

The same paper by Smirnova et al. described the core principle behind organoid intelligence. They specifically defined it as a multidisciplinary field that involves pursuing research and development approaches similar to the established goals and fields of artificial intelligence centered on the development and applications of computers and related technologies to perform tasks that normally require human intelligence.

Organoid intelligence specifically involved harnessing the computational capabilities of the human brain by developing biological hardware made of brain organoids from reprogrammed stem cells. An organoid will be attached with a custom-built device responsible for recording its neural activity and transmitting instructive information to it.

Purpose and Applications

The field of organoid intelligence promises to deliver unprecedented developments in computing speed, processing power, computing or processing efficiency, and storage capabilities while consuming power that is significantly lower than traditional computers. The field can also help in the further understanding of brain development and brain functions.

Canadian scientist Rémi Quirion explained the main technology behind silicon-based chips is nearing its physical size limit. Other scientists and even manufacturers are worried that this could halt technological developments and innovation in the future. He also noted that traditional computers are also power-hungry machines and this is evident from the power requirements of artificial intelligence applications and supercomputers.

Nevertheless, as mentioned by Quirion, biology could solve the limitations of traditional computers. He noted that computers are actually designed to model the human brain. Other studies have also demonstrated the unparallel capabilities of the human brain. It is true that computers are better at numbers but the human brain is better at learning.

The human brain is also more complex. The Fujitsu K supercomputer was able to simulate 1 second of 1 percent of human brain activity in 2013. It took 40 minutes to accomplish this task using more than 82000 processors. The estimated storage capacity of the human brain is also around 2500 terabytes. It would take 34 coal-fired power plants generating 500 megawatts each to power data centers that could hold the same amount of data.

Smirnova et al. also explained that computers are indeed better at dealing with numbers but a typical human brain is far superior in performing complex computations. The brain can learn to distinguish between two types of objects such as a cat or a dog using around 10 training samples while an AI algorithm would need millions or large datasets.

Remember that the AlphaGo computer program beat Go world champions Fan Hui and Lee Sedol. This demonstrated the superiority of computers over humans. It is still important to note that the computer program was trained on data from 160000 games. An individual would need to train for five hours each day for 175 years to attain this same level of performance. The human brain is arguably more efficient.

Smirnova et al. further added that the increasing computational requirements of modern societies have also resulted in greater utilization of energy inputs. Organoid intelligence can be considered a more sustainable approach to modern computing because brain organoids consume a fraction of the power requirement of silicon-based computers.

A biocomputer made from brain organoids is theoretically more powerful and power efficient than the most sophisticated and advanced computer available. These organoids can even hallmark the arrival of newer generations of sustainable supercomputers. Nevertheless, based on the aforementioned, the purpose of organoid intelligence is to utilize and maximize the advantages of the human brain over traditional, silicon-based computers.

Of course, more work needs to be done. Smirnova et al. noted that the practical applications of organoid intelligence would require scaling up current brain organoids into more complex and durable structures that are enriched with cells and genes associated with learning while also developing new learning models and tackling emerging ethical issues.

FURTHER READINGS AND REFERENCES

  • Hartung, T., Smirnova, L., Morales Pantoja, I. E., Akwaboah, A., Alam El Din, D.-M., Berlinicke, C. A., Boyd, J. L., Caffo, B. S., Cappiello, B., Cohen-Karni, T., Curley, J. L., Etienne-Cummings, R., Dastgheyb, R., Gracias, D. H., Gilbert, F., Habela, C. W., Han, F., Harris, T. D., Herrmann, K., … Zack, D. J. 2023. “The Baltimore Declaration toward the Exploration of Organoid Intelligence.” Frontiers in Science. 1. DOI: 3389/fsci.2023.1068159
  • Kagan, B. J., Kitchen, A. C., Tran, N. T., Habibollahi, F., Khajehnejad, M., Parker, B. J., Bhat, A., Rollo, B., Razi, A., and Friston, K. J. 2022. “In Vitro Neurons Learn and Exhibit Sentience When Embodied in a Simulated Game-World. Neuron. 110(23): 3952-3969.e8. DOI: 1016/j.neuron.2022.09.001
  • Shipman, S. L., Nivala, J., Macklis, J. D., and Church, G. M. 2017. CRISPR–Cas Encoding of a Digital Movie into the Fenomes of a Population of Living Bacteria. Nature. 547(7663): 345-349. DOI: 1038/nature23017
  • Smirnova, L., Caffo, B. S., Gracias, D. H., Huang, Q., Morales Pantoja, I. E., Tang, B., Zack, D. J., Berlinicke, C. A., Boyd, J. L., Harris, T. D., Johnson, E. C., Kagan, B. J., Kahn, J., Muotri, A. R., Paulhamus, B. L., Schwamborn, J. C., Plotkin, J., Szalay, A. S., Vogelstein, J. T., … Hartung, T. 2023. “Organoid Intelligence (OI): The New Frontier in Biocomputing and Intelligence-In-A-Dish.” Frontiers in Science. 1. DOI: 3389/fsci.2023.1017235