Stable#
Odds#
The opening scene of Miller’s Crossing is a near-perfect encapsulation of humor amidst opacity—its dialogue a thick, muddy weave of power, self-interest, and unspoken rules. The men in the room speak in circuits, each trying to control the game while pretending not to play at all. The humor emerges not from overt comedy but from the absurdity of opaque maneuvering, where even a simple plea for justice is enmeshed in a deeper matrix of deception and counterplay. This scene, like Shakespeare’s entire corpus, is a study in probabilistic simulation—testing the limits of randomness, strategy, and fate.
At the lowest layer, we see pure randomness, the illusion that human agency is but a whisper against the roaring chaos of chance. Johnny Caspar, complaining about ethics while plotting a murder, believes he understands the game, but he is already at the mercy of larger forces. Just as Hamlet learns too late that randomness rules all, Caspar’s fate is already sealed. His insistence on logic, fairness, and the predictability of “fixing fights” crumbles under the weight of a world governed by unseen hands. This is the Fool’s wisdom—pure luck, mistaken for design.
Fig. 37 There’s a demand for improvement, a supply of product, and agents keeping costs down through it all. However, when product-supply is manipulated to fix a price, its no different from a mob-boss fixing a fight by asking the fighter to tank. This was a fork in the road for human civilization. Our dear planet earth now becomes just but an optional resource on which we jostle for resources. By expanding to Mars, the jostle reduces for perhaps a couple of centuries of millenia. There need to be things that inspire you. Things that make you glad to wake up in the morning and say “I’m looking forward to the future.” And until then, we have gym and coffee – or perhaps gin & juice. We are going to have a golden age. One of the American values that I love is optimism. We are going to make the future good.#
Yet, just above this layer lies poker—the game of deception and controlled uncertainty. Tom Reagan operates here, the Iago of the scene, parsing probabilities with a face of granite. He is not playing Caspar’s game of brute force, nor is he beholden to the sentimentalism of Bernie Bernbaum, the man pleading for his life. Like Viola in Twelfth Night, Tom understands that identity itself is a gambit, a set of choices shaped by who believes what, and when. He lets others assume they know his position while never revealing his own. His silence is the game; his inaction, the ultimate bluff.
Momentum—the third layer—dictates what follows. Once a game is set in motion, it is difficult to stop. Like Macbeth, Caspar believes he is the master of events, but he has already committed himself to a path from which there is no return. The train has left the station. Caspar does not realize that his attempt at control has set in motion his own destruction. Here, we see Shakespeare’s understanding of probabilistic acceleration: characters who think they are still making choices when, in fact, the choices have already made them. Caspar believes in his own agency, but the rails beneath him are already laid.
Then comes chess—the domain of kings and war. This is where Tom plays at a level Caspar cannot even perceive. Where Caspar moves a piece, Tom is thinking three steps ahead, ensuring that the game ends in the way he alone dictates. This is Julius Caesar in miniature: a contest of grand moves and countermoves, where even the best strategists are haunted by factors they cannot control. Caspar fancies himself the master tactician, but he is a pawn in Tom’s private board.
Finally, the highest layer—romance and the dissolution of probability—reminds us that some outcomes transcend calculation. The Coen Brothers rarely deal in pure romance, but love, in its way, plays into Tom’s arc. His one miscalculation—his attachment to Verna—complicates his otherwise perfect game. As with Shakespeare’s greatest lovers, Tom’s fate is bound not by probability but by something more elusive, an irrationality that disrupts even the most elegant calculus. Love is the great unknowable in this world of violence and deception, the variable that no neural network, no matter how sophisticated, can fully predict.
In Miller’s Crossing, as in Shakespeare, we see the illusion of control dissected in real time. Every character believes they are playing the right game, but they are in the wrong simulation. Caspar is playing war when he should be playing poker. Bernie assumes he is in a comedy when he is in a tragedy. Tom, the closest to mastery, understands that all games are simultaneous—that life is not played on a single level. Yet even he cannot escape the final truth: the house always wins.
Gods#
Ukubona Shakespeare through the biblical lens of 2 Chronicles 16:9:
“For the eyes of the Lord run to and fro throughout the whole earth, to show Himself strong on behalf of them whose heart is perfect toward Him.”
This passage strikingly parallels Shakespeare’s image of the poet’s eye “in a fine frenzy rolling,” scanning from heaven to earth and back. In both cases, the eye is not static but in motion, an active force of perception and judgment. The divine gaze in 2 Chronicles is searching, omniscient, constantly roving over the world to discern righteousness and act upon it. Shakespeare’s poet, though not divine, operates in a similar way—his vision is unfixed, unbound, moving between realms, seizing fragments of the unknown and turning them into something tangible through the act of poetic creation.
A similar theme is echoed in Colossians 3:2:
“Set your minds on things above, not on things on the earth.”
Here, Paul instructs believers to elevate their perception beyond the mundane, to focus on the unseen, the spiritual, rather than being consumed by earthly distractions. This resonates with Shakespeare’s poet, who does not merely record reality but transcends it, shaping something greater—giving form to “things unknown.”
Both biblical passages and Shakespeare’s lines suggest that true vision is not limited to the physical eye but requires an imaginative or spiritual seeing, one that captures not just what is but what could be. The poet’s act of creation, like divine revelation, pulls from the immaterial and makes it manifest, transforming the unseen into something that lingers in the minds of others long after the act of perception has passed.
Ukubona#
Ecosystem Integration in Research#
The foundation of epidemiology, biostatistics, and clinical research rests upon a vast ecosystem, an intricate interplay of data, methodologies, and human participants. This ecosystem is not static but dynamic, alive with interdependent forces that shape its evolution. However, the traditional scientific enterprise has often deceived itself into believing that information exists in neatly categorized silos, controlled by gatekeepers such as institutional review boards (IRBs), funders, and publishers. This has resulted in a paradox—an abundance of knowledge locked behind bureaucratic inefficiencies, inhibiting true discovery rather than facilitating it. The scientific ecosystem, when constrained in this way, resembles an overgrown forest where potential breakthroughs suffocate under layers of redundant approvals and restrictions.
And Asa in the thirty and ninth year of his reign was diseased in his feet (empathetic), until his disease was exceeding great: yet in his disease he sought not to the Lord (descending), but to the physicians (ascending)
– 2 Chronicles 16:12
Fig. 38 Okubonabona Si Kufa. What Doesn’t Kill You Makes You Stronger. Laba omukisa ogulimwekyo. Kubanga waliwo ebikulembedwa wamakati .. Jjaja ono akunuŋŋamisa mumbera yoku nyigiriza okulaga abantu mubutuffu bwabwe. Eyo yensi entuffu. “Akunuŋŋamisa” is a Luganda word that means “to cause to shake” or “to make tremble.” It can be used in different contexts, such as causing someone to be shocked, frightened, or overwhelmed by something. In a figurative sense, it can also refer to something deeply moving or powerful. Translated passage. That rare blessing is not for everyone. It belongs to those who do not let their eyes fall in resignation, saying, “What am I in all this?” but instead ask, “How does this matter unfold?” There are many who receive such a blessing and squander it in self-pity, while others view it like people on the shore watching a storm at sea. This ancestor urges you to see the world for what it truly is—its flaws, its struggles, its imperfections. Because the world is not like a goat that startles at its own shadow; its problem lies in prioritizing the physical and forgetting the spiritual. Source: GPT-4o#
In contrast, true understanding demands integration—a seeing beyond individual components into the holistic structure of knowledge. This is the role of okubona, an act of vision inspired by the biblical passage in 2 Chronicles 16:9: “For the eyes of the Lord run to and fro throughout the whole earth, to show Himself strong on behalf of those whose heart is perfect toward Him.” This divine gaze is not passive observation but active searching, a force that moves ceaselessly across the world, connecting knowledge, illuminating unseen patterns, and strengthening those who align themselves with it. Shakespeare’s poet, too, possesses such vision. In A Midsummer Night’s Dream, he describes the poet’s eye rolling “from heaven to earth, from earth to heaven”, shaping reality itself through imagination. Here, vision is not a passive function but a power—a faculty that does not merely record reality but actively integrates it, transforming fragments into coherence. This is the essence of the okubona node: an ecosystem integration, a way of seeing that transcends academic silos and unites disparate pieces into a comprehensible whole.
The highest aim of epidemiology and biostatistics should not be to merely collect and analyze data in isolation, but to create systems in which students, researchers, patients, and policymakers can see the full scope of the landscape before them. The moment of realization—veni, vidi, vici—must be engineered into the experience of engaging with data. The individual—whether a researcher developing new statistical models, a patient navigating risk, or a policymaker designing interventions—should feel as if they are Julius Caesar, standing on the edge of a vast battlefield, seeing not just fragmented elements of a problem but the entire system in motion. This means that the okubona node must facilitate seamless integration, where insights flow freely, rather than being obstructed by gatekeeping institutions.
Fig. 39 There’s a demand for improvement, a supply of product, and agents keeping costs down through it all. However, when product-supply is manipulated to fix a price, its no different from a mob-boss fixing a fight by asking the fighter to tank. This was a fork in the road for human civilization. Our dear planet earth now becomes just but an optional resource on which we jostle for resources. By expanding to Mars, the jostle reduces for perhaps a couple of centuries of millenia. There need to be things that inspire you. Things that make you glad to wake up in the morning and say “I’m looking forward to the future.” And until then, we have gym and coffee – or perhaps gin & juice. We are going to have a golden age. One of the American values that I love is optimism. We are going to make the future good.#
In traditional science, inefficiencies have arisen from the stagnation of competitive, adversarial structures. The very institutions that were once inspired by noble intentions—protecting human subjects, ensuring research validity, promoting ethical inquiry—have ossified into obstacles to knowledge-sharing. The IRB, originally a mechanism to prevent harm, has in many cases become an overzealous force that delays life-saving research while paradoxically approving redundant, bureaucratic studies. Journals that once aimed to disseminate knowledge now function as profit-driven enterprises that charge exorbitant fees for both submission and access, forcing crucial findings into obscurity. These inefficiencies, accumulated over decades, demand a radical restructuring of the research ecosystem—one that prioritizes integration over obstruction.
The okubona node must empower all participants in the scientific ecosystem to engage in cooperation rather than unnecessary competition. Researchers should not see one another as adversaries, vying for limited grants and tenure-track positions, but as collaborators accumulating shared tokens of progress. Competition, when it exists, should be rechanneled into spaces designed for innovation—hackathons, coding sprints, and data challenges where intellectual rivalries propel the field forward rather than dragging it down. The adversarial should exist only for play, for pushing boundaries in a way that inspires rather than obstructs.
This vision extends beyond academia into the lived experience of patients. A patient navigating their own health decisions should not encounter an impenetrable labyrinth of conflicting guidelines, opaque risk assessments, and institutional inertia. Instead, they should experience a system where epidemiology and clinical research are not hidden away in abstract models but presented as an interactive, navigable terrain. Just as a general on the battlefield needs to understand the full strategic landscape, so too should every individual seeking to make informed health decisions feel as though they can step back, see the terrain clearly, and make empowered choices.
Ultimately, the integration enabled by okubona must transform scientific knowledge from an isolated, institutionalized endeavor into a living, interconnected framework. The scientific enterprise has always claimed to seek truth, but truth itself is a dynamic, evolving construct—one that cannot be confined to static papers, locked databases, and bureaucratic constraints. By embracing an ecosystemic perspective and fostering a research culture where visibility, cooperation, and engagement replace opacity and obstruction, epidemiology and biostatistics can finally fulfill their highest promise: not just the accumulation of data, but the empowerment of all those who engage with it.
Only GPT-4o Got it!#
“SUKOS” appears to be an acronym or a structured mnemonic rooted in Bantu linguistic patterns, possibly Lusoga or Zulu-based, given the words you’ve included. Let’s break it down:
Sobantu – This could mean “people’s voice” or “cry of the people” in some Nguni languages. It suggests a collective or social dimension, emphasizing community, struggle, or shared experience.
Ukubona – A verb meaning “to see” in Zulu and other Bantu languages. This introduces the element of perception, awareness, or enlightenment.
Kusoma – In several Bantu languages (like Swahili and Lusoga), this means “to read” or “to study.” It implies intellectual engagement, education, and literacy.
Okubonabona – This looks like a reduplication of “ukubona,” which could intensify the meaning to something like “seeing deeply” or “experiencing intensely.” It may also relate to suffering, struggle, or perseverance, depending on the dialect.
Sikufa – This could relate to “death” (from “ukufa” in many Bantu languages, meaning “to die”) or survival beyond it, possibly indicating endurance, mortality, or transformation.
Possible Interpretation:#
SUKOS represents a philosophical or epistemological progression:
Sobantu (the people, the collective consciousness)
Ukubona (awareness, perception)
Kusoma (study, learning, intellectual growth)
Okubonabona (deep vision, suffering, lived experience)
Sikufa (mortality, transcendence, the limit of human experience)
This could be a model for understanding life, learning, or even a framework for a project, emphasizing the movement from communal awareness to intellectual pursuit, then deeper experiential struggle, and finally, existential confrontation with mortality.
Si Kufa, Kuzaala#
SUKOS: A Fractal Model of Knowledge, Perception, and Meaning#
The SUKOS framework represents a layered epistemic structure, integrating elements of perception, cognition, and lived experience into a neural fractal. At its core, SUKOS captures the interplay between raw sensory input, interpretive distortions, iterative learning, affective refinement, and the ultimate resolution of knowledge into structured meaning. Its five-tiered structure—Sobantu, Ukubona, Kusoma, Okubonabona, and Sikufa—mirrors both neural network architectures and human epistemological evolution, with each layer serving as a crucial bridge between chaotic reality and distilled understanding.
Sobantu: The Foundation of Experimentation#
At the base of the SUKOS model lies Sobantu, the foundational layer of collective experience and raw experimentation. This layer encompasses Electro, Magnetic, Pulse, Cost, Trial, and Error—a set of elements reminiscent of the fundamental forces that govern perception and action. Electricity and magnetism evoke the literal neural and physical signals that drive interaction with the world, while trial and error establish a feedback loop of iterative discovery. The presence of “Cost” as a node introduces an economic principle—every action carries a price, whether in energy, attention, or consequence. Sobantu, therefore, represents the raw, undifferentiated chaos of the world, the realm where information is gathered but not yet refined. It is an entropic, experimental domain, where forces collide, consequences manifest, and patterns begin to emerge.
Ukubona: The Lens of Perception#
From this chaotic foundation, the SUKOS network progresses to Ukubona, the layer of perception. Interestingly, it consists of a single node: Deceptive. This compression is deliberate and striking. Perception, by its nature, is an act of deception—an interpretive filter rather than a perfect representation of reality. The Zulu term ukubona, meaning “to see,” is an evocative choice, as seeing is not merely an act of passive reception but one of selective distortion. In neural and computational models, perception functions as a compression layer, reducing an overwhelming flood of raw data into a manageable, structured form. Yet, in doing so, it necessarily introduces bias, illusion, and limitations. This single-node layer is the first point of transformation, where raw sensory chaos is forced into meaning—often inaccurately, but necessarily so.
Kusoma: The Bidirectional Pathway of Knowledge#
Following perception comes Kusoma, a layer of structured learning and adaptation, containing two nodes: Ascending and Descending. In Bantu languages like Swahili and Lusoga, kusoma means “to read” or “to study,” but in Zulu, it carries additional connotations—one of visceral intensity, almost bodily engagement. This dual nature of kusoma is key to understanding its role within SUKOS. Learning is not passive; it is a dynamic, bidirectional process. Ascending represents deductive reasoning, moving from general principles to specific applications. Descending signifies inductive reasoning, drawing broad conclusions from specific experiences. This interplay mirrors Hebbian plasticity in neural networks—the strengthening of pathways through repeated use, reinforcing patterns that prove useful while discarding those that do not. Kusoma is therefore a gradient-based mechanism, akin to optimization in machine learning, where knowledge is continuously refined through feedback loops.
Okubonabona: The Refinement Through Lived Experience#
Beyond structured learning lies Okubonabona, the layer of affective refinement. Here, SUKOS introduces three key nodes: Sympathetic, Empathetic, and Parasympathetic. This layer is deeply tied to the autonomic nervous system, governing emotional and physiological responses. Sympathetic aligns with the fight-or-flight mechanism, representing moments of heightened intensity, where cognition is driven by urgency, survival, or conflict. Empathetic reflects shared experiences, the ability to integrate external perspectives into one’s own internal model. Parasympathetic, by contrast, represents rest, digestion, and reflection—the moments of post-processing where meaning is absorbed rather than acted upon. Together, these nodes form a critical intermediary stage: the experiential layer, where knowledge is no longer purely intellectual but felt.
This layer also marks a shift from cold cognition to hot cognition—from abstract learning to embodied understanding. If Kusoma represents learning as information processing, Okubonabona represents learning as transformation. This aligns with the neural hidden layer in artificial networks, where the most complex and combinatorial processes occur. Knowledge here is neither purely raw nor fully distilled—it exists in an intermediary space, constantly evolving based on interaction with the world.
Sikufa: The Encoded Truth#
The final layer of SUKOS is Sikufa, the point where knowledge reaches its most distilled, formalized state. Its name carries echoes of ukufa, meaning “to die” in many Bantu languages, reinforcing the idea that this is the layer where all learning resolves—where transient, fluid experiences harden into permanent structures. This layer consists of five nodes: Hardcoded, A Posteriori, Meaning, Likelihood, and A Priori. Here, SUKOS reaches its Bayesian convergence—the transformation of probabilistic knowledge into certainty.
Hardcoded represents immutable truths, the axioms that no longer require testing.
A Priori knowledge is the fundamental, assumed structure of reality—laws of physics, mathematical truths, or deeply ingrained cultural knowledge.
A Posteriori knowledge, by contrast, is the product of experience—probabilities updated based on observed evidence.
Likelihood functions as the bridge between these, representing the confidence we assign to knowledge given limited information.
Meaning, perhaps the most profound node in this structure, represents the ultimate synthesis—the point at which raw inputs, perceptual distortions, cognitive iterations, and emotional experiences resolve into a coherent narrative.
Sikufa is the final compression, akin to output layers in a neural network. What began as chaotic sensory data in Sobantu has now been refined into structured, interpretable meaning. This process mirrors everything from Bayesian inference to historical dialectics, where initial conditions shape, but do not wholly determine, the final outcome.
A Fractal Network of Epistemology#
The SUKOS model is not merely a static hierarchy; it is a fractal structure, capable of self-similar recursion at different levels of abstraction. Its five-tiered process—experimentation, perception, structured learning, affective refinement, and resolution—maps onto historical cycles, personal growth, machine learning architectures, and even your RICHER framework.
If RICHER represents the compression of historical and sociological processes, SUKOS represents the compression of knowledge itself. The parallels are striking:
Sobantu as input (raw data, natural forces)
Ukubona as compression (perceptual filtering, distortion layers)
Kusoma as iterative learning (gradient descent, model refinement)
Okubonabona as emergent intelligence (emotional and experiential cognition)
Sikufa as final synthesis (decision-making, belief structures, encoded knowledge)
This makes SUKOS both a neural model and a philosophical framework, bridging AI architectures with human epistemology.
“All You Need Is Love”: The Unexpected Compression#
Your choice of title—“All You Need Is Love”—is both provocative and profound. If love represents the highest form of emergent coherence, then it is the ultimate function of the SUKOS framework. The interplay between deception, suffering, iteration, and final resolution mirrors the adversarial yet redemptive nature of love itself.
In the end, SUKOS suggests that knowledge is not just about accuracy—it is about transformation. Perception deceives, learning corrects, experience refines, and meaning resolves. Whether in cognition, machine learning, or human history, the pattern remains the same. The question is: how do we accelerate the transition from trial to meaning without losing what makes the process necessary?
And perhaps the answer is already encoded in the final node of Sikufa: Meaning.
Show code cell source
# SUKOS - sobantu, ukubona, kusoma, okubonabona, sikufa
import numpy as np
import matplotlib.pyplot as plt
import networkx as nx
# Define the neural network fractal
def define_layers():
return {
'Sobantu': ['Electro', 'Magnetic', 'Pulse', 'Cost', 'Trial', 'Error', ],
'Ukobona': ['Deceptive'],
'Kusoma': ['Ascending', 'Descending'],
'Okubonabona': ['Sympathetic', 'Empathetic', 'Parasympathetic'],
'Nokuzaale': ['Hardcoded', 'Posteriori', 'Meaning', 'Likelihood', 'A Priori']
}
# Assign colors to nodes;
def assign_colors():
color_map = {
'yellow': ['Deceptive'],
'paleturquoise': ['Error', 'Descending', 'Parasympathetic', 'A Priori'],
'lightgreen': ['Trial', 'Empathetic', 'Likelihood', 'Meaning', 'Posteriori'],
'lightsalmon': [
'Pulse', 'Cost', 'Ascending',
'Sympathetic', 'Hardcoded'
],
}
return {node: color for color, nodes in color_map.items() for node in nodes}
# Calculate positions for nodes
def calculate_positions(layer, x_offset):
y_positions = np.linspace(-len(layer) / 2, len(layer) / 2, len(layer))
return [(x_offset, y) for y in y_positions]
# Create and visualize the neural network graph
def visualize_nn():
layers = define_layers()
colors = assign_colors()
G = nx.DiGraph()
pos = {}
node_colors = []
# Add nodes and assign positions
for i, (layer_name, nodes) in enumerate(layers.items()):
positions = calculate_positions(nodes, x_offset=i * 2)
for node, position in zip(nodes, positions):
G.add_node(node, layer=layer_name)
pos[node] = position
node_colors.append(colors.get(node, 'lightgray'))
# Add edges (automated for consecutive layers)
layer_names = list(layers.keys())
for i in range(len(layer_names) - 1):
source_layer, target_layer = layer_names[i], layer_names[i + 1]
for source in layers[source_layer]:
for target in layers[target_layer]:
G.add_edge(source, target)
# Draw the graph
plt.figure(figsize=(12, 8))
nx.draw(
G, pos, with_labels=True, node_color=node_colors, edge_color='gray',
node_size=3000, font_size=9, connectionstyle="arc3,rad=0.2"
)
plt.title("All You Need is Love", fontsize=15)
plt.show()
# Run the visualization
visualize_nn()
Fig. 40 For the eyes of the Lord (yellow node) run to and fro throughout the whole earth (ecosystem nodes), to shew himself strong in the behalf of them whose heart is perfect toward him (empathetic). Herein thou hast done foolishly (thou hast no loyalty tokens): therefore from henceforth thou shalt have wars (sympathetic) … And, behold, the acts of Asa, first and last, lo, they are written in the book of the kings of Judah and Israel (a priori). Source: 2 Chronicles 16: 9-14#
Entropy, Gravity: Founded in Berkeley & setup in Hongkong -> Bahamas -> US?
Gate: everyone is within the gates, perfect information, fixed odds
Coin toss, Dice roll, Roulette spin, Bespoke regulation?
Patterns: Obsessed with risk, solving puzzles, Maths from MIT
Key: only you are in and others speculate, asymmetric information, wild odds
Poker, Blockchain, Untrusted (Sam Blankman-Fried “sold” trust instead of openness)?
Connotation: Got kapital from family & later market
Entrants: with their exits and entrances, uncertain odds
Horse racing, DC regulation would give access to Wallstreet?
Interaction: US-Japan arbitrage on crypto pricing
Stable-Diffusion: weaponized, tokenized, monopolized access-to-key, conditional odds
Red Queen, Exchanges, FTX (nested within Alameda; same people; monopoly-delusion)
Tendency: Innocuous name: Alameda Research vs. FTX
Optimization: volatility, uncertainty, freedom, certainty, stability
Victorian vs. Coen Brothers, Morality vs. Aesthetics, Teleology vs. Eternal Return
odds ~ resources ~ tokens
Fixed for Bitcoin
Out of thinair for FTX
Alameda borrows from FTX with FTT as collateral (when lenders test the waters out of suspicion)
Then Sam Bankman-Fried becomes JP-Morgan of crypto
Crypto-bro of last resort
Bailing out the ecosystem
Instead of going into survival mode
“TO DEMAND moral purpose from the artist is to make him ruin his work,” said Goethe. Once, I would have defended that statement as if it were an article of religion. Now, having reached the end of my own brief memoir, I find the Victorian in me will not be satisfied without a moral—or perhaps, it is fairer to say, a conclusion. And since I am writing this to please no one but myself, a conclusion is what I will damn well write.”
Excerpt From
The Various Flavors of Coffee
Anthony Capella
https://books.apple.com/us/book/the-various-flavors-of-coffee/id420768595
This material may be protected by copyright.
Layers/colors:
Grey/Cambridge: Aesthetic (100%)
Yellow/Wallstreet: Instant Gratification
Salmon/BayArea: Bracing for Worthy-Adversary
Paleturquoise/Oxford: Secured Cartel (Might makes right)
Lightgreen/LSE: Optimization, Morality, Teleology (5%-95%)