\n", " \n", "\n", "\n", "\n", " \n", " \n", "\n", "\n", "
\n", " \n", "\n", "\n", "
\n", " \n", "- What makes for a suitable problem for AI (Demis Hassabis, Nobel Lecture)?\n", "
\n", "\n", "
\n", "- Space: Massive combinatorial search space
\n", "- Function: Clear objective function (metric) to optimize against
\n", "- Time: Either lots of data and/or an accurate and efficient simulator
\n", "- Guess what else fits the bill (Yours truly, amateur philosopher)?\n", "
\n", "\n", "
\n", "- Space\n", "
\n", "\n", "
\n", "- Intestines/villi
\n", "- Lungs/bronchioles
\n", "- Capillary trees
\n", "- Network of lymphatics
\n", "- Dendrites in neurons
\n", "- Tree branches
\n", "- Function\n", "
\n", "\n", "
\n", "- Energy
\n", "- Aerobic respiration
\n", "- Delivery to \"last mile\" (minimize distance)
\n", "- Response time (minimize)
\n", "- Information
\n", "- Exposure to sunlight for photosynthesis
\n", "- Time\n", "
\n", "\n", "
\n", "- Nourishment
\n", "- Gaseous exchange
\n", "- Oxygen & Nutrients (Carbon dioxide & \"Waste\")
\n", "- Surveillance for antigens
\n", "- Coherence of functions
\n", "- Water and nutrients from soil
\n", "
CG-BEST: Karl Marx remains surprisingly relevant 200 years after his birth. He rightly predicted some of the pitfalls of capitalism, but his solution was far worse than the disease. Source: The Economist
\n", "\n", " Resources: Data 📉; Pretext, 95/5 (deconstructed)\n", "\n", "This looping is essential. Ukubona believes in feedback—not just statistical, but moral. If a model suggests disproportionate risk to a population, that is not just data—it’s indictment. It is signal that something is wrong, not with the patient, but with the structure around them. The sea is biased. Ukubona doesn’t pretend otherwise.\n", "\n", "The platform, in this regard, becomes more than code—it becomes protest. Against opaque risk scores, against unjust eligibility criteria, against the flattening of human variety into actuarial tables. Ukubona’s ship is navigated not by captains alone, but by all who ride it. Patients, clinicians, families—they all get to steer.\n", "\n", "This ethos makes Ukubona dangerous to systems that prefer silence. Where other tools prioritize [efficiency](https://en.wikipedia.org/wiki/Department_of_Government_Efficiency), Ukubona dares to [prioritize](https://www.nobelprize.org/uploads/2024/12/hassabis-lecture.pdf) understanding. It does not believe that care must be fast to be just. Sometimes slowness—reflection, customization, hesitation—is the more ethical stance.\n", "\n", "But that stance requires tools. Ukubona’s screwdriver, its most humble icon, is also its most radical. It affirms that models must be maintained, adjusted, localized. That no model is final. That even truth, once seen, must be tuned. This is what separates Ukubona from static dashboards and inert PDFs.\n", "\n", "The screwdriver, too, protects against the pirate. It is not enough to resist once. Maintenance is ongoing. Pirates, whether in finance, insurance, or ideology, reemerge. Ukubona builds systems that expect attack—not in paranoia, but in wisdom. It assumes the ship will be tested.\n", "\n", "And so it trains its users. Not with lectures, but with interface. Ukubona teaches through interaction. When a user tweaks a curve, it is not mere aesthetics—it is education. It says: you have agency. You can intervene. Even in statistics, your story matters.\n", "\n", "That story is not singular. Ukubona's visualizations allow for overlay—multiple lives, multiple paths. It is a comparative theology. Not in a religious sense, but in a metaphysical one: which future shall I serve? This or that? The curve becomes sacred, a scrying mirror for the soul.\n", "\n", "But the mirror reflects only what is fed into it. This is why Ukubona insists on open data pathways—why it allows CSV uploads, API inputs, voice entries. If the platform is to serve the many, it must listen to the many. Its ears are open.\n", "\n", "It listens not just to patients, but to analysts. Ukubona’s back-end is transparent. The Cox regression is not a black box; it is documented, modifiable, inspectable. The platform respects those who question it. This, too, is moral: no model should be above critique.\n", "\n", "\n", " \n", "\n", "```{raw} html\n", "\n", "\n", "
\n", " Bequest: Filter 🌴; Subtext, 80/20 (project 2025)
\n", " Strategic: Choice 🪵; Text, 50/50 (low-propensity voters)
\n", " Motive: Scaling 🪡 🐫 ; Context, 20/80 (once-in-a-lifetime)
\n", " Distributed: Canopy 🌿; Metatext, 5/95 (symphony)
\n", " — Yours Truly\n", "
![\"Eco-Green](\"https://upload.wikimedia.org/wikipedia/commons/c/c5/Peacock_Plumage.jpg\")
\n",
"\n",
"\n",
"\n",
" \n",
" \n",
"\n",
"\n",
"```{figure} https://www.ledr.com/colours/white.jpg\n",
"---\n",
"width: 1\n",
"height: 1\n",
"name: figures/cgbest\n",
"---\n",
"We just might have figured out how to number non-python images. This is exciting\n",
"```\n",
"\n",
"