When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
In the latest AI in biotechnology news, Techopedia recently interviewed Dr Fred Jordan, co-founder of FinalSpark, about their innovative work in neural research and biotechnology.
Jordan discussed how FinalSpark uses living neurons from human stem cells for advanced computing.
We also talked about the ethical considerations of this technology, the possibility of organoids (self-organized three-dimensional tissue cultures) gaining consciousness, and how these advances might transform the fields of artificial intelligence (AI) and biotechnology.
AI in Biotechnology: FinalSpark’s Neural Research Approach
Q: Can you explain how FinalSpark uses living neurons derived from human stem cells for computing?
A: Well, the very first answer is that it doesn’t work yet – but the “yet” is an important word.
Right now, we’re using induced pluripotent stem cells derived from human skin cells [induced pluripotent stem cells, or iPSCs, are specialized cells created in a lab from adult cells, which can then turn into any cell type in the body].
Then, we can differentiate IPSCs into neurons [cells in the nervous system that transmit information through electrical and chemical signals] and glial cells [supportive cells in the nervous system that provide structure, nutrition, and insulation to neurons].
We put these cells together, and they form a small brain organoid, very small – 500 microns, about half a millimeter.
So far, we’ve stored one bit of information. Think of it like the performance of the first quantum computers, which could store 1 “Qubit” (quantum bit) of data 20 years ago. Now, we have our first “Bibit”, or biological bit.
Q: How do you program these neurons to work with electronics?
A: Candidly, we don’t fully know yet; it’s our field of research. What I can tell you is about our approach, which gives some preliminary results.
We use stimulation, stimulating at the right moment with the right pattern to enforce the behavior we want. We have electrodes connected to a computer and decide precisely when to stimulate each electrode.
We use AI, and interestingly, sometimes we use artificial neural networks (ANNs). AI is used to decide the appropriate time to send a specific amount of current to an electrode.
Q: How do you stimulate the organoid exactly?
A: We stimulate the organoid by sending specific amounts of current to an electrode.
Another approach we’re using is chemical stimulation – we are currently working a lot with dopamine [a neurotransmitter in the brain that influences mood, motivation, and reward]. We release dopamine precisely at the right time directly to the brain organoid by using a process called uncaging.
We encapsulate dopamine in a molecular cage, invisible to the organoid initially. When we want to ‘reward’ the organoid, we expose it to specific light frequencies. This light opens the cage, releasing the dopamine and providing the intended stimulus to the organoid.
Organoid Intelligence and Ethical Implications
Q: How does the intelligence of these organoids differ from traditional artificial intelligence in terms of complexity?
A: It’s a challenge to quantify, but AI’s neural networks are far simpler than biological neurons, which are used in organoids.
Each neuron in our brain has about 10,000 connections, far more complex than anything in conventional AI. Furthermore, neurons work with spikes, sending signals in every direction, unlike the fixed input-output models in classical AI.
Q: Do you think organoid intelligence could eventually surpass human intelligence? AI has already exceeded human capabilities in certain areas, such as chess.
A: It’s a speculative path. History shows that each time we think machines have surpassed human intelligence, we realize they haven’t truly achieved it. We use real biological components with organoids, bringing a different perspective. However, it’s hard to predict whether they’ll ever surpass human intelligence.
Q: And the ethical considerations of using living neuron-based systems? How do we address that?
A: These are living systems similar to other biotech applications – for example, we use live organisms to make beer.
Of course, the ethical consideration is increased because we are using human cells. From an ethical perspective, what is interesting is that all this wouldn’t be possible without the ISPCs.
Ethically, we don’t need to take the brain of a real human being to conduct experiments. With a few stem cells, we can produce millions of them; they reproduce, and we can use them over and over.
People may question if we are crossing a line with technological advancements, but does it make sense? Where does the technology lead us? Is it better? Do we live better? The answer is yes, we live better, way better – I’m much happier here today than I would have been 1,000 years ago. This progress benefits me and the entire human race. So, there’s no doubt that moving in the direction of progress has been beneficial for us.
Speculations on Consciousness in Organoids: AI in Biotechnology Predictions
Q: As these technologies evolve, do you think there’s a risk of organoids developing sentience or consciousness?
A: It’s difficult to say. The concept of consciousness is not well-defined. In our experiments, we work on a small scale, and it’s beyond our current scope to determine if organoids can develop consciousness. Understanding and defining consciousness is the first step in such discussions.
Q: Considering the complexity of the human brain versus the current simplicity of the neural networks in organoids, do you think organoids need to reach a similar level of complexity to develop consciousness?
A: It’s a reasonable assumption, but it could also be incorrect.
We might have better chances of achieving this level of complexity by culturing biological neurons than by using digital computers to simulate artificial neurons. However, identifying consciousness in organoids would require a method of communication, much like how babies communicate discomfort through crying.
Q: So, in theory, could an organoid show some form of reaction or communication?
A: Biomarkers could potentially indicate reactions, but real-time analysis of these is challenging. We could observe the activity and spiking rate of neurons, but drawing parallels to human expressions of feelings, like crying, is speculative at this stage.
The Future of AI in Biotechnology: FinalSpark’s Goals and Projects
Q: What is the ultimate goal for FinalSpark in this field, and how do you envision this technology being used in the future?
A: The ultimate goal is to develop machines with a form of intelligence. We want to create a real function, something useful. Imagine inputting a picture to the organoid, and it responds, recognizing objects like cats or dogs.
Right now, we are focusing on one specific function – the significant reduction in energy consumption, potentially millions to billions of times less than digital computers. As a result, one practical application could be cloud computing, where these neuron-based systems consume significantly less energy. This offers an eco-friendly alternative to traditional computing processing.
Ultimately, the future of AI in biotechnology holds huge potential for various applications because it’s a completely new way of looking at neurons. It’s like the inventors of the transistor not knowing about the internet.
Q: Are there other projects happening at FinalSpark?
A: Yes, there’s something interesting to share. The COVID-19 pandemic significantly impacted us. We had limited access to our lab, which led us to automate our operations and make everything remotely operable. After three years, our lab became fully remote-operable. This development made us realize that if we could operate from home, research groups worldwide could also utilize our neurons.
We opened this opportunity to the research community and received responses from 32 research teams. Due to the high interest, we selected five of these projects. Currently, three applications are actively using our lab facilities remotely, one each in France, England, and the U.S.
These research teams are conducting experiments with our neurons, obtaining results, and furthering scientific understanding through their work. The idea is to foster a collaborative and accessible research environment.
About Fred Jordan
Dr Fred Jordan is a French physicist with a Ph.D. in signal processing. His two-decade career showcases his versatile expertise in digital technology, signal processing, and innovative research.
He and his business partner, Martin Kutter, founded their first company, AlpVision, in 2000. This company is known for its unique technology that distinguishes between real and counterfeit products using a smartphone app. Jordan serves as CEO of that company.
Jordan and Kutter’s passion for research led them to start another project, FinalSpark. FinalSpark is an advanced R&D project, which is risky and requires a lot of dedication, which is why they use their own money from AlpVision to support it.
