For decades, the field of artificial intelligence (AI) has strived to create computers with the capabilities of a human brain.
Now, a new study proposes a “new frontier” for computing called “organoid intelligence” (OI) that could surpass the learning capabilities of any machine.
OI is using organoids—small lab-grown tissue that resemble mature organs—as a form of “biological hardware” and a potentially smarter alternative to the silicon chips in AI.
Researchers at Johns Hopkins University in Baltimore think a “biocomputer” powered by an organoid made up of millions of human brain cells could be developed in our lifetime.
While previous studies questioned whether a biocomputer would cross an “ethical boundary,” the team says organoids would be used in a safe and “ethical way.”
Scientists unveil a revolutionary path to advance computing: organoid intelligence, where lab-grown brain organoids act as biological hardware. Pictured, a magnified view of a lab-grown brain organoid with fluorescent labels for different cell types (neurons are shaded pink)
The new study was authored by an international team of researchers led by Professor Thomas Hartung at Johns Hopkins University.
“We present a collaborative program to implement the vision of a multidisciplinary field of OI,” they say in their paper, published today in the journal Frontiers in science.
“This aims to establish OI as a form of true biological computing that utilizes brain organoids using scientific and biotechnical advancements in an ethically sound manner.
“While AI aims to make computers more brain-like, OI research will explore how to make a 3D brain cell culture more computer-like.”
Although AI can process calculations involving numbers and data faster than humans, brains are much smarter at making complex logical decisions, such as the difference between a dog and a cat.
Developing systems that can make such logical decisions will be where OI will excel – although the researchers stress that AI and OI can co-exist in a “complementarity” fashion.
Like the nerve cells in our brain, OI systems would use chemical and electrical signals to follow instructions and perform tasks.
OI could also be faster, more efficient and more powerful than AI while requiring only a fraction of the energy.
Pictured are brain organoids in the lab at Johns Hopkins Bloomberg School of Public Health
Organoids – small, three-dimensional tissue cultures derived from stem cells – are not new and date back to the late 1990s.
They have already been used to study and treat cancer, heart problems, tissue transplants, neurological disorders and much more.
Different cells in the organoids correspond to different areas – for example, brain organoids have been used to investigate cellular aspects of learning and memory.
These tiny spheres—about the size of a pen tip—have neurons, synapses, and other features that enable basic functions like learning and remembering.
While most cell cultures are flat, organoids have a three-dimensional structure, increasing their cell density by 1000 times, meaning the neurons can form more connections.
Professor Hartung started growing and assembling brain cells into functional organoids in 2012 using cells from reprogrammed human skin samples.
Each organoid contains about 50,000 cells, about the size of a fruit fly’s nervous system – although this is currently too small.
“For OI, we should increase this number to 10 million,” Professor Hartung said.
The research team envisions building a futuristic computer with brain organoids connected “to real-world sensors and output devices.”
Each organoid contains about 50,000 cells, about the size of a fruit fly’s nervous system – although this would need to be scaled up to build an OI computer
Researchers describe their roadmap for organoid intelligence (OI) and its possible applications in the journal Frontiers in Science
Brain organoids would then be trained “using biofeedback, big-data warehousing and machine learning methods,” the team explains.
OI computers with this “biological hardware” could begin in the next decade to alleviate supercomputing power consumption demands, which are becoming increasingly unsustainable.
A supposed OI computer could be many times more powerful than Frontier, the world’s fastest supercomputer, while consuming less energy.
“We’re reaching the physical limits of silicon computers because we can’t put more transistors in a small chip,” said Professor Hartung.
“But the brain is wired completely differently — it has about 100 billion neurons connected through more than 1,015 connection points.
“It’s a huge power differential compared to our current technology.”
The ethics of using AI to do a human’s job is well documented and even touched upon in popular culture, such as the Black Mirror episode “White Christmas.”
But the use of human cells could prove even more controversial, raising questions such as whether millions of them grown in succession can evolve consciousness.
Scientists already warned that an ‘ethical boundary may have been crossed’ when miniature brains were made from human tissue.
A supposed OI computer could be many times more powerful than Frontier, the world’s fastest supercomputer (pictured)
Organoids can feel or even experience pain, which means controls must be in place to prevent them from suffering, the critics said.
Another potential problem is donors whose cells are used to make biocomputers without their consent, but the study’s authors claim they are aware of all ethical considerations.
“An important part of our vision is to develop OI in an ethical and socially responsible way,” said Professor Hartung.
For this reason, we’ve been working with ethicists from the very beginning to establish an “embedded ethics” approach.
“All ethical issues will continue to be reviewed by teams of scientists, ethicists and the public as research evolves.”
