While accomplished both academically and professionally, Matrix Chief AI Scientist Professor Steve Deng has long been bothered by a perceived hole in his resume. Despite being a life-long fan of science fiction, he has always been too busy pushing forward the field of Artificial Intelligence to contribute in a meaningful way to the esteemed craft of longform science fiction.
That is, until now! The Matrix team is pleased to announce the release of 2019: The Bayesian Heart.
As a special thanks to the continued support of the Matrix community, Chapter 3 of 2019: The Bayesian Heart is included below. If you enjoy what you read, please head out and purchase 2019: The Bayesian Heart today! Available exclusively at Barnes and Nobles.
Chapter 3: Activation Potential
As his senses came seeping back through black unconsciousness, professor Deng noticed a dull heat building in the toe of his shoes as they dragged heavily on the carpeted aisle behind him. Two sets of hands fumbled to readjust their grip on his shoulders as the men unevenly hoisted and lugged his tall frame up the incline of the auditorium with jerky progress. He heard a nearly continuous chirr of digital shutters from camera phones over a rumble of gasps and muttering.
“Professor Deng!” a young woman cried out sharply. Yi-ling’s generally diffident voice rasped with fraught relief at the sight of his head lifting slightly. “You…you were, mid-sentence. At the podium. You hit the stage in such a… terrible, fashion.” Steve Deng was now fully alert, with the full inquest of his gaze studying the distress twisting Yi-ling’s delicate features.
“It was terrible!” Yi-ling repeated with a rising emphasis that precipitated tears. Professor Deng consciously softened his expression to a gentle receptiveness. “You…just collapsed. With this unearthly heaviness. From a full stand! The sound…” Her voice shattered and gave out. “This…inhuman thud. You were gone…instantly.”
Professor Deng now took to his feet, nodding slightly to acknowledge the crowd, then extended an assuring hand to Yi-ling’s shoulder before resuming an easy stride toward the exit. Quickly regaining her poise and indifferent to the gawking crowd still assembled for professor Deng’s talk, Yi-ling regrouped her steps and kept pace behind her mentor. He turned to Yi-ling with a gentle smile saying, “Call my family and let them know I have some important work to attend to and will be late.” Seeing the questions wavering in Yi-ling’s fretful gaze, he added “I’ll stay safe. Trust me.”
Back in his office, professor Deng strained to call back glimpses from the vision that set in during his talk. He made no effort to comprehend the physical episode Yi-ling described, and was engrossed in reclaiming some insight from that state, undisturbed by the vibration of his phone bleating continuously against the wooden desk. There was some vast expanse, with multiple planes of high-dimensional structured information. Not verbal, nor numerical — but highly correlated in some way that he was interpreting in that state. No, not interpreting. Generating.
Professor Deng decided to take the subway home to avoid potentially lapsing behind the wheel. As he made his way across the Tsinghua campus, his visual and auditory faculties had become gradually heightened. Not given to people watching, he’d nonetheless begun scrutinizing the physical bearing and movements of people around him. Even from considerable distance, he started decoding fine-grained expressions and strains of vocal affect with intense fidelity.
Supposing that thought could originate in some rational, pure platonic form, he wondered; through the sheer physicality of face-to-face conversation — from posture, to vocal waver, and sequence of micro-expressions, these ideas become unavoidably imbued with qualifiers. Over years, each person evolved into their own tightly synchronized mental-physical system. Each developed a unique psychosomatic disposition, literally woven into the fabric of their bones, tendons and muscles.
If abstract thought, like something executable by computer, started as a musical composition, then we were like instruments giving imprecise time and shape to each melody. All our minor structural imperfections coloring each note. A hypothetically identical idea adopted an altogether different character based on who expressed it. But if words aren’t received entirely objectively, what is our measure for evaluation? A honeyed voice? Shapely features? Assuming a demonstrative power stance from a TED talk? Or social constructs imparting status?
To what degree might physical conditions not only create feedback for thought, but exert a continuous possibly determinant influence? Professor Deng paused, suppressing a flaring hunger in order for a flurry of concepts to snap together in his mind.
He recalled in the 1800s Hermann von Helmholtz had formulated a curious materialist theory applying thermodynamics to humans as “energy systems.” Freud later picked up on this notion, extending it to the tangible transmission of psychic energy. Freud focused on how thought not only encoded itself in the brain and muscle memory, but unconsciously shaped all interactions whether social or with objects in our surroundings. This was the logic behind his obsessive focus on misplaced keys and apparently random slips of tongue.
This digression through the periphery of science flashed through professor Deng’s mind as a stepping stone to Rosalind Picard’s work in affective computing. “Emotion pulls the levers of our lives…” he voiced to himself softly, reciting from her seminal 1995 paper. This was not lyrical metaphor, but referred to the fact that the so-called advanced mental functions of prefrontal cortex originate in, or are substantially modulated by the limbic system — our emotional brain.
Now moving through the neon haze of Beijing streets under moonless nightfall, professor Deng’s vision phased out briefly as a violent pang of hunger twisted in his abdomen. He struggled to retrace his thoughts on affective computing, seeking to connect it to his research in Bayesian computing. Filtering out the clamor and haphazard movements around him, narrative thought eclipsed from his mind’s eye. Layered matrices of neural networks were transposed over an immersive model of 3-D brain — not in visual imagination, but in algorithmic space.
Professor Deng seemed to be directing a fly-through of the thalamocortical circuits in the brain, using Bayesian inference to parameterize how emotions modulate neuron discharge patterns across various cortical regions. Signals coursing through different cytoarchitectures were not visualized but rather understood as eigenspace. The immersive tour through cathedrals of fibers and synapses unfolded out elegantly, effortlessly assuming greater detail under his observation. It seems to be progressing toward some purposeful apex, like a pleasant dream — when once again the heavy curtain of unconsciousness fell swiftly.
Concluding that professor Deng might be diabetic, a well-meaning Aussie exchange student, had worked a straw into his lips and squeezed a mouthful of boba milk tea from a plastic cup. The pungent syrup fouled by the sour musk of milk roused professor Deng, with a volt of glucose bringing him to an upright seated position where he resumed staring vacantly toward the now vanished portal of sensory activity.
He had been building the bridge between Bayesian methods, affective computing and the limbic foundations of higher-order consciousness — but now, that beautiful hyperspace and the Escher-like structure spanning it were gone. Sitting on the sidewalk half a block from a busy intersection, Professor Deng remained unaware of the persistent well-meaning shakes from his Aussie benefactor. Now he finally turned to some self-reflection about his present condition.
“Dr. Bachmann,” he said to himself with deliberate, yet directionless clarity. He then remembered an uncharacteristic commotion Dr. Bachmann has made over the gift of a fitted Pittsburg Pirates hat back in grad school. Accounting for Steve’s abundance of hair, Bachmann had bought the largest size available, reasoning that some make-shift padding could neatly account for any excess room. To Dr. Bachmann’s unreserved surprise, the cap was too small even to be forced on. After delving into an appraisal of Steve’s hair, he exclaimed “why this wouldn’t account for more than 8 centimeters of diameter!” still clutching a flowing lock.
As much of a departure as this rumpus over the hat already was from Dr. Bachmann’s generally rational and stern nature, there was indeed more unabashed handling to come. Without a single word uttered from a stupefied Steve, Bachmann had started a veritable phrenological examination, which he referred to several times as “craniometry” in his narration of the inspection, and for which he had prepared a measuring tape.
Professor Deng now broke away from the lulling shoulder nudges of the Boba guy, but not before sharing a sincere smile and brief expression of gratitude. He made his way to a nearby noodle shop where he ordered and finished 3 bowls of lamb “knife-shaved noodles” and 4 cans of Jiaduobao.
When Steve had later regaled the hat incident over beers, a fellow Ph.D student said, “you do know Bachmann was one of the authors of the 1999 Lancet study of Einstein’s brain right?” That night Steve had read the Lancet paper with considerable interest. His subsequent attempts to downplay the boisterous comparisons to Einstein from classmates, they persisted for the remainder of his time at CMU. Despite the peculiarity of the entire episode, over all these years professor Deng had never considered any connection to something that happened shortly after.
Steve was in Dr. Bachmann’s office discussing the use of MCMC parameterization to model synaptic plasticity, when he was rendered unconscious. It was later described that a camera obscura replica, small but with considerable heft, had fallen from a high shelf striking him about an inch above his right ear. He awoke 37 hours later in Bachmann’s home, where he was purportedly brought to recover in more comfortable environs after a small-scale cranial incision was made at the University hospital to prevent brain swelling.
In a follow up with doctors they had mentioned that although it was clear there would be no cognitive impairment, he might expect some dramatic changes to his metabolism. His caloric intake had indeed surged to between 4,300–5,000 calories a day, though he maintained his trim physique with only sporadic light exercise. Professor Deng order two more lamb skewers, then politely averted the owner’s pleasantries about his healthy appetite.
The mental states from today, were certainly distinct from any form of human thought, yet not altogether foreign to professor Deng. They progressed like code. Only after finishing the second kebob, did professor Deng feel truly free from the tyranny of his metabolism. “But that was 13 years ago, why now?” he voiced to himself inaudibly. This was the first time he questioned why the surgery should have changed his metabolism so dramatically. The impact and incision should have had no impact on the hypothalamus. What was drawing the additional energy?
Now analyzing a potential mechanism behind today’s experiences, professor Deng felt each incident had unfolded like game-states. It was as if he were running simulations, like looking forward 70, 80 moves into a chess game, with each step factored the range of possible responses of all previous states. Once he began to consider the next possible outcome, it was as if the perfect model began pouring through streams of data picking out salient features. His mind seemed to be navigating data cubes with the same speed, precision and effortlessness that we easily make out a loved one in a crowded restaurant.
One thing that had evidently amazed Dr. Bachmann during his phrenological evaluation, was the width of Steve’s parietal and occipital portions of the skull — which he remarked on many times. “Your hairstyle does quiet a job obscuring it my friend! Sure, I had my suspicions, but this must be nearly 20% wider! Isn’t that something! Everyone’s always focused on the frontal cortex, but in reality, there’s more substance in that old platitude about what’s between your ears. Also, EQ, is grossly misunderstood and undervalued. All these Ph.Ds here lost in their own heads, bumping into people without so much as an apology. They’re missing at least half the picture — but not you Steve.”
The location of his surgery scar, description of the falling object, and arrangement of Bachmann’s office had never really come together convincingly for professor Deng — but he’d never had reason to question a friend and mentor? The study on Einstein’s brain had been notable due to the discovery of multiple structural differences that were highly uncommon, or possibly one-of-a-kind. Much research has been dedicated to how Einstein’s enhanced visuospatial and mathematical skills may have resulted from cortical reorganization, having a brain with different folding patterns. One notable difference was an additional fold in his right inferior frontal gyrus, presumable affording space for the region to expand.
A theory that Bachmann and co-authors of the Lancet piece advanced about Einstein actually missing his right parietal operculum has since been attributed to unclear original slides back in 1955. That interpretation did however present an interesting lead, suggesting that Bachmann might have been particularly fascinated with the insular cortex deep within the Sylvian Fissure. This region would be a compelling point of investigation for two reasons. First, it’s directly involved with multimodal sensory processing, and interpersonal dynamics. Second, it’s an important gateway for cognitive-emotional processes via the basal ganglia that orchestrates information to multiple cortical regions, the thalamus and brainstem.
Professor Deng now thought back to his fly-through look at the brain from earlier, “I was revealing the answer to myself.”
Could it be that Bachmann was actually hoping to use years of in vivo biological sampling to train an affective computing model? If so, perhaps he was able to retrofit the afferent pathways from the insular cortex into the striatum? That could potentially be a means for quantifying dopaergic modulation within the hyperdirect pathway. The striatum controls activation and inhibition feedback between the thalamus and many cortical regions, which would be critical for understanding how emotions calibrating higher-order reasoning.
When Dr. Bachmann was selecting a pioneering subject, he must have suspected me of sharing some key similarities in brain structure with Einstein, based on the width of my parietal-occipital width, and my work with 3-D model integration and advanced math. At the time, he would have still been working on the false assumption that Einstein had no parietal operculum, which would afford relatively clear access to the insular cortex. One of the most important selection criterion was that when the model came to maturity, it would require someone who understood Bayesian methods, who had designed them, to be able to integrate these new processing models into thought.
What was the nature of the device? How was it powered? Maybe piezoelectric? Perhaps there was a connection to his increased appetite? Professor Deng recalled approximately half of the brain’s energy is dedicated to maintaining the resting membrane potential. Could Dr. Bachmann have created a glucose biosensor with an electrolyte battery?
Bachmann, would already be well into his 70s, professor Deng needed to book a flight.