Intro

Over the last 50 thousand years the human brain has shrunk by approximately 150 grams. This change is largely attributed to the reduced need for multitasking since the Cro-Magnon era and the transition to communal living with specialized social roles.

Nevertheless, the human brain remains the most efficient learning system. First, humans are about 10^6-fold more power-efficient than modern supercomputers for basic operations. Teaching a machine a complex task can require up to 10^10 times more energy than it takes for humans to learn the same thing. Second, biological learning relies on significantly less training data and iterations. Therefore human neurons achieve at least a one-million-fold advantage in both data and energy efficiency compared to current machines.

[MEA (Human Brain Organoids Multi-Electrode Array) © NeuroPlatform by FinalSpark]

A biocomputer is a system built from brain organoids - tiny clusters of neurons grown from human stem cells. These organoids, living in Petri dishes, are connected to digital interfaces via Multi-Electrode Arrays. Though still experimental, they can be trained using electrical stimulation and dopamine rewards.

The 2024 MindWare series are based on the FinalSpark biocomputing platform similarly to it's prequel PinkWare made in 2020-2021.

While AI aims to make computers function more like the human brain, OI (Organoid Intelligence) research focuses on adapting 3D brain cell cultures to operate more like computers.

OI experiments and the use of human brain surrogates raise numerous ethical concerns often summarized as the Greely Dilemma: “In a vat, no one can hear you scream.”


[MEA streaming data © NeuroPlatform by FinalSpark]

Although the brain reading (organoids Output) is already quite advanced and has enabled the control of prosthetic robots, the cellular understanding of writing (organoids Input) remains difficult, except in certain cases involving the visual cortex.
This limitation is one of the reasons why telepathy was chosen as the MindWare Input method, while Output is obtained by data scraping from the FinalSpark NeuroPlatform live stream and later via API access.

---> more about organoids and biocomputing research here in the Annex

  Maya Mirrors ✳

A soul selfie.

A series of works made with 4 Neural Networks: an organoid NN, a machine learning NN for signal detection and processing (Intan software for MEA interfacing), my NN, your NN.
The generative part is printed with an inkjet printer on craft paper.
The mantras in Slater ternary encoding language are a handmade ink calligraphy.
---> more about Slater in PostPost

In Hinduism and Buddhism Maya represents illusion or the way in which it manifests.

Maya Mirrors is a phygital ceremony that stages a real-time calligraphy exercise in a closed-loop cycle experiment. In this process, a human painter acts as a visualization medium, telepathically relaying the resulting image back to the biocomputer. By observing their mental reflections in this “mirror”, the organoids are challenged to distinguish between illusion and reality.

In turn, the painter is challenged to do the same in his own reality.

  Prayer for Player ╳

Praying for the game. And for the player.

- Middle part: a visualisation of an organoid elicited spikes signal.
- Upper part: his/her name retrieved from this signal, a kind of a signal hash or checksum.
- Lower part: begins with plaintext Om and continues with ciphertext of a Tibetan (Wylie transliteration) or Sanskrit world or a beginning of a mantra. Encrypted with post-quantum cryptography (the same method and keys as for the PostPost project) and written in Slater ternary language.



[ ---> The first ten inscriptions on Ordinals ]

  Code Contemplation ∞

Teaching meditation to an Ex Vivo computer.

Samatha-Vipassana meditation is a core practice across all Buddhist traditions and a primary method for cultivating freedom of consciousness, attaining enlightenment and ending suffering.
Samatha is the 1st stage - it aims to calm the mind. Vipassana is the 2nd stage - it aims to observe the mind. They are practiced either simultaneously or sequentially.

This Python code is written for the FinalSpark NeuroPlatform using Intan Technologies integrated neuroscience electrophysiology circuits and interface.

Here the Samatha function is defined as a dopamine reward for human brain organoids if the given organoid decreases elicited spikes activity through the training session.
Calm mind.

The Vipassana is a similar function that provides dopamine reward in case the organoid stays in a constant state - the state of observation - it doesn't sink neither in torpor nor in agitation.
Watch mind.

Inscribed on Ordinals: inscriptions 80,974,050 and 80,974,244.

  NeuroDoom 🎮

Fighting inner synthetic demons.

NeuroDoom - Biocomputing Telepathic game powered by human brain organoids.
Episode 1 - Knee-Deep in the Alive.

An attempt to port the classical Doom shooter video game to a biocomputer. Or rather to make a biocomputer play Doom, where the role of a human operator is to act as a generative visualization medium.

The NeuroDoom bridges Horo Mo X in Paris with a biocomputer in Vevey, Switzerland, performing a telepathic experiment over a distance of 400 km.

Watch the video and more ---> NeuroDoom project

  Annex

Brief history of human brain organoids biocomputing development

There are many terms for brain-directed computing as an alternative to silicon-based computing: wetware, biological computing, organoids computing, Biological Neuronal Network (BNN) by Eugene M. Izhikevich, Synthetic Biological Intelligence (SBI) by Brett J. Kagan et al. (Cortical Labs) and most recently Organoid Intelligence (OI) by Lena Smirnova et al. (Johns Hopkins University).

2004: Thomas DeMarse from University of Florida made a brain-in-a-dish out of 25,000 rat’s neurons and a 60 multi-electrode array that could control pitch and roll of an F-22 fighter jet simulator.

2008: Yoshiki Sasai and his team at the RIKEN Center for Developmental Biology in Japan were developing self-organizing cerebral tissue - miniature brain models - from embryonic mouse and human stem cells. Sasai’s work led to the creation of complex brain-like structures. His techniques were groundbreaking, allowing researchers to study early brain development and neurological diseases in 3D.

2013: first cerebral organoids developed by Madeline Lancaster and Jürgen Knoblich at the Institute of Molecular Biotechnology (IMBA) in Vienna. They pioneered a method to grow three-dimensional brain organoids using human pluripotent stem cells. Their method involved growing stem cells in a 3D matrix called Matrigel, then placing them in a spinning bioreactor to improve nutrient and oxygen distribution. This breakthrough allowed them to create more complex brain-like structures that mimic the human brain’s development and early architecture with the formation of distinct brain regions, a big leap forward compared to prior models.

2015: Paola Arlotta's group at Harvard developed protocols for generating region-specific brain organoids mimicking: forebrain/cortex, midbrain, hypothalamus and cerebellum.

2019: Generation of Functional Neural Networks and Brain Waves. Alysson Muotri at UCSD had cultured brain organoids capable of producing oscillating brain waves similar to those in preterm human babies.

Autumn 2020: Swiss company FinalSpark began web stream of its organoid computer data via the BioServer.

2021: the Launch of EU-funded international NEU-CHiP project at Aston University that aims to develop bio-inspired computing technology by utilizing human neuronal networks.

2022: Creation of Multi-Regional Brain Organoids. Multiple collaborative teams, including scientists from the University of Cambridge and Stanford University succeeded in creating more complex, multi-regional brain "assembloids" that simulate multiple interconnected regions of the brain, including cortical and subcortical structures.

Oct 2022: Australian company Cortical Labs made a brain organoid system DishBrain of 800k cells that have learned to play the 70s tennis-like video game, Pong.

Feb 2023: the team of Johns Hopkins University coined the term Organoid intelligence in their paper “Organoid intelligence (OI): the new frontier in biocomputing and intelligence-in-a-dish” - a seminal work on biocomputing architecture.

May 2024: FinalSpark launched its remotely accessible NeuroPlatform for research institutions.

Reading

Organoid intelligence (OI): the new frontier in biocomputing and intelligence-in-a-dish. -->

Playing Brains: The Ethical Challenges Posed by Silicon Sentience and Hybrid Intelligence in DishBrain. -->

In vitro neurons learn and exhibit sentience when embodied in a simulated game-world. -->

Open and remotely accessible Neuroplatform for research in wetware computing. -->

< all works