Dish Brain - Synthetic Biological Intelligence

Dec 2022 - Harnessing the computational power of living neurons to create synthetic biological intelligence (SBI), previously confined to the realm of science fiction, may now be within reach of human innovation.

https://www.sciencedirect.com/science/ar...7322008066

Integrating neurons into digital systems may enable performance infeasible with silicon alone. Here, we develop DishBrain, a system that harnesses the inherent adaptive computation of neurons in a structured environment. In vitro neural networks from human or rodent origins are integrated with in silico computing via a high-density multielectrode array. Through electrophysiological stimulation and recording, cultures are embedded in a simulated game-world, mimicking the arcade game “Pong.” Applying implications from the theory of active inference via the free energy principle, we find apparent learning within five minutes of real-time gameplay not observed in control conditions. Further experiments demonstrate the importance of closed-loop structured feedback in eliciting learning over time. Cultures display the ability to self-organize activity in a goal-directed manner in response to sparse sensory information about the consequences of their actions, which we term synthetic biological intelligence. Future applications may provide further insights into the cellular correlates of intelligence.

The DishBrain protocol was refined over three pilot studies, each increasing the density of sensory information...[ ]...Pilot study 3 adopted the layout in Figure 4D and changed to the combined rate (4–40Hz) and place-coding method of data input. This combined rate and place coding has compelling biological similarities conceptually to the rodent barrel cortex, suggesting this encoding is physiologically coherent

   

Dec 2022 - Harnessing the computational power of living neurons to create synthetic biological intelligence (SBI), previously confined to the realm of science fiction, may now be within reach of human innovation.

https://www.sciencedirect.com/science/ar...7322008066

Given the speed of modern computers, it is hard to see what using real neurons would give you over just simulating some neurons - unless perhaps they would connect to the other realm that many suspect contains all consciousness.

If they do make that connection, would we be able to recognise it?

David

Given the speed of modern computers, it is hard to see what using real neurons would give you over just simulating some neurons - unless perhaps they would connect to the other realm that many suspect contains all consciousness.

If they do make that connection, would we be able to recognise it?

David

Something along these lines (the new development of a "biological computer gel-pack" approach) was used as a plot element for a very advanced technological replacement for the previously used silicon-based large scale integration comprising the starship Voyager's distributed computational modules, in the '90s Star Trek Voyager TV series. The advantages were supposedly mainly very high logic device packing density, great efficiency along with self-repair capabilities, and a high degree of self-assembly of replacement modules. Of course there were tradeoffs - the "gel-packs" could become infected by viruses and bacteria, their extreme complexity could allow them to somehow become sentient, conscious, and the interface between silicon-based computation and biological neuron-based structures is very problematical with progressive deterioration of the junctions. Not good things for essential computational elements in a starship's digital electronics' design. 

As you might tell, I would consider this new research development mainly science-fiction, not a practical ground-breaking new digital computational element.

As you might tell, I would consider this new research development mainly science-fiction, not a practical ground-breaking new digital computational element.

Yeah my guess is we are hitting the limit of machine "learning" - the disasters surrounding driverless cars and the lowered investment interest in them is likely going to extend to other "AI" industries.

So this is more about getting some hype for a new way to take people's money with the promise of amazing technology.

Given the speed of modern computers, it is hard to see what using real neurons would give you over just simulating some neurons - unless perhaps they would connect to the other realm that many suspect contains all consciousness.

If they do make that connection, would we be able to recognise it?

David

Well neuron structures are clearly doing a lot more than is currently accepted, as an example, Anirban Bandyopadhyay and his colleagues are doing very detailed and careful work to show just how complex the electromagnetic processes within the neuron could be. A technically challenging paper from 2021, but worth struggling with, I find it very exciting stuff... it's worthwhile reading just to grasp the idea of time-like geometrical shapes... that you could have geometry in time...

https://doi.org/10.31083/j.jin2004082

The most important finding reported here is that
all prime contributors have threshold resonance frequencies
that burst energy. So, we found that the dc potential burst
of the membrane is the last or final event in a sequence of ac
electromagnetic energy bursts. Membrane firing is not alone.

Chemically, Hodgkin-Huxley has shown earlier in their Nobel-prize winning work [74] that neuron fires even if HeBHiZ does not exist. HeBHiZ is not essential for firing. However, our theory and experiment both argue that it edits the time gap between spikes, essential for brain cognition. If the microtubule-neurofilament structure is dissolved, then the membrane attached actin-spectrin periodic lattice-like architecture disappears. Their physical relationship was never measured before.

In summary, we triggered a millimeter-wave to fire a neu- ron even using a sub-threshold pulse where it should not fire by conventional wisdom [76] (Fig. 5a). We could even stop the inevitable firing under an above threshold pulse when mixed with a suitable millimeter-wave [84, 85] (Fig. 5b). A membrane fires even without the filaments inside. The fila- ments only modulate the spike frequency [75]. That is why even when the neuron does not fire, a natural wave flows through the filaments (Fig. 5c). The whole neuron turns to an integrated vibrating system where distant ion channels, ir- respective of their separation, are coupled to signal each other 103 times within a span of a single nerve spike via filaments. Thus, a deep neural substructure [53, 64] is the third non- somatic contributor to a nerve spike, in addition to the cur- rent axonal computing [82] and gliotransmitters editing the synapse activity. We envision that the improved version of SDIM would reveal many more different signals at various time scales.

Max,

I don't think you really take on board the philosophical power of David Chalmers' Hard Problem.

All the complexity in the world doesn't explain how the sensation of green, or love, or interest, or disgust, or ANYTHING is created.

Compare this with, say classical physics. Once you knew that molecules whizzed around - faster when hot, and slower when cool - you could explain a mountain of facts even if the details were complicated. I mean it was obvious that molecules hitting a piston would push on them.

The trouble is there is no conceptual link between the chemicals inside neurons doing their thing, and any sensation of any kind.

David

Max,

I don't think you really take on board the philosophical power of David Chalmers' Hard Problem.

All the complexity in the world doesn't explain how the sensation of green, or love, or interest, or disgust, or ANYTHING is created.

Compare this with, say classical physics. Once you knew that molecules whizzed around - faster when hot, and slower when cool - you could explain a mountain of facts even if the details were complicated. I mean it was obvious that molecules hitting a piston would push on them.

The trouble is there is no conceptual link between the chemicals inside neurons doing their thing, and any sensation of any kind.

David

Where I am now, none of that is really a problem for me anymore... I tried to deal with some of this in this thread, suggesting that people struggle with Nima's presentations, one of which I linked to.

Where I am now, none of that is really a problem for me anymore...

So if it isn't a problem for you anymore, how do you understand the Hard Problem?

I tried to deal with some of this in this thread, suggesting that people struggle with Nima's presentations, one of which I linked to.

That is some really (and I think speculative) complex physics you want to introduce. Are you sure it really relates to the operation of a human brain?

David

That is some really (and I think speculative) complex physics you want to introduce. Are you sure it really relates to the operation of a human brain?

David

Yes, Nima's work seems likely to become a better story to understand my experiences (whether they are of a brain, or a piston, or love).