🤖A.I News. Cortical Labs: Blurring the Line between Biology and Artificial Intelligence - From AI to Bio-AI 🧠🧟🧫😬⚡IA-ismo

In the name of accessibility and convenience, we're introducing this new audio format.

Biological Artificial Intelligence (B.A.I.) is the pursuit of programmable biological computing, using cultured human brain cells (brain organoids) as opposed to silicon circuits. The cutting-edge research in this field is spearheaded by Cortical Labs and Monash University, and this week they secured nearly A$600,000 from the Intelligence and National Security Discovery Research Grants Program.

This grant will enable them to cultivate and "train" 800,000 brain cells for specific tasks. This burgeoning field brings significant ethical and legal questions to the table, and aspects of this technology are undeniably controversial. In this article, we delve into these issues with the intention of sparking informed debate on the field's future.

So, what is organoid intelligence?

Organoid intelligence (IoI) is an attempt to engineer biological computers by growing miniature brains (brain organoids) from human stem cells. Unlike traditional Artificial Intelligence, it leverages the inherent information-processing capabilities of cells, rather than depending on programmed algorithms. It could potentially be more energy efficient and surmount the current limitations of AI in tasks involving complex sensory information processing and "lifelong learning". Despite being in its infancy, this nascent field could dramatically shift our understanding of intelligence.

The Experiment: Neurons Playing Pong

Cortical Labs, a pioneering enterprise in the domain of neuroscience and artificial intelligence, carried out an enthralling experiment that bridges the gap between biology and technology. In this experiment, they employed cultured brain cells from both mice and humans, positioning them on a high-density multiple electrode array (HD-MEA). The goal was to observe and scrutinize how these brain cells could adjust their activity via stimulation.

The task given to these brain cells was no less challenging than playing the legendary 1980s game, Pong. Although this game has a straightforward design, it demands swift responses and incessant adaptation to the fluctuating conditions of the game. Astonishingly, these brain cells demonstrated a rapid adaptation to this task in a virtual setting, displaying goal-oriented behavior in a matter of minutes. This experiment served as a compelling showcase of brain plasticity and the capacity of brain cells to learn and adapt to new circumstances.

Key Concepts to Grasp in Order to Understand the Experiment:

• Brain Plasticity - This is the brain's capability to perpetually transform and adapt in response to new experiences, learning, or injury. This attribute will allow neurons to swiftly adapt to playing Pong.
• Friston's Free Energy Principle (FEP) - This principle suggests that all self-organizing systems that are separated from their environment strive to minimize their variational free energy (VFE). Simply put, it implies that brain cells will endeavor to reduce "surprise" or uncertainty in their environment.
• Active Inference - This is a process through which neurons tailor their behavior to minimize uncertainty. In the context of this experiment, it means that brain cells are consistently modifying their behavior in response to the information received during the game.
• Awareness - It is believed that the biological neural network (BNN) lacks sufficient mechanisms to be aware that it is engaged in a game of Pong, even though it exhibits a conditioned response to its surroundings.

El Experimento: Dishbrain

In the experiment conducted by Cortical Labs, glutamatergic neurons were employed, which were differentiated from stem cells using the NGN2 protein. These neurons, integral for learning and located in areas like the hippocampus, were chosen due to their aptitude to form strong connections and their efficiency for mass production through rapid maturation.

The DishBrain system was designed to interact with these neurons, tracking their electrical activity and supplying 'sensory' electrical stimulation. This system emulated the game of Pong, transmitting data to a predefined sensory area and recording the 'motor regions' activity to control a paddle in the game.

Through self-organization, the neurons adopted diverse firing patterns to engage with the game. The outcomes demonstrated a substantial correlation between the activity in the sensory region and the motor regions, along with dynamic adjustment of activity in the motor regions during gameplay.

This experiment showcased the functional plasticity of neurons and their ability to adapt and change, which was significantly enhanced during gameplay. Moreover, information entropy, a measure of average 'surprise', decreased during predictable interactions with the environment during the game, suggesting that neurons can learn to anticipate and adapt to stimuli.

Sensory and predictable stimuli, administered at a voltage of 75 mV, fall within the normal physiological ranges and are not expected to harm neuronal cells. While the unpredictable stimulus, even though delivered at a higher voltage (150 mV), remains within the parameters used in prior studies without damaging neuronal cells. Nevertheless, the unpredictability of this stimulus could introduce some stress to the neurons, which are known to be sensitive entities.

One theory posits that the BNN (Biological Neural Network) might harbor internal 'beliefs' (like predictions) about the state of the world, and learning would involve modeling its behavior to update these beliefs to minimize their impact on the environment.

Biological artificial intelligence emerges as an opportunity and a competitive edge for countries investing in research, promising treatments for neurological, neurodegenerative, and psychiatric disorders, as well as the potential to experiment and test drugs. However, further research is necessary along with a legal and ethical framework to ensure the safety of neurons and prevent them from being subjected to any form of stress.

And don't forget, if you like this content and want to support us so we can keep bringing you more interesting information, you can buy us a symbolic coffee on Ko-fi! ☕️💪

#AI, #ArtificialIntelligence, #Neuroscience, #Dishbrain, #CorticalLabs, #BrainResearch, #BioComputing, #Biology, #Technology, #Innovation, #StemCells, #Neurons, #BrainCells, #NeuralNetworks, #ScienceNews, #Research, #BrainPlasticity, #Pong, #Gaming, #Learning, #Adaptation