Links for 2024-12-18 axisofordinary.substack.com/p/links-for-...

Watch Google's Waymo self-driving car avoid calamity with a scooter rider

Google's new state-of-the-art AI video model is absolutely brilliant: blog.google/technology/g...

Prompt for the video: A pirate canon shooting meatballs into a spicy hotpot

Axiom 4 (Mathematical and Logical Soundness): Assume a consistent framework for reasoning, probability, and mathematics. This is a bare minimum for rational inference.

Axiom 3 (Cognitive/Observational Reliability): Our faculties (or instruments) are good enough to let us propose, refine, and test theories iteratively. We don’t need absolute certainty here—just enough trust to get started and not dismiss the entire enterprise.

Axiom 2 (Resource Reasoning): When forming priors, distributing credence over an unbounded complexity space forces you to give higher effective credence to simpler theories. This ensures that simpler theories get tested first and, if successful, rapidly gain credence.

Axiom 1 (Learnability): There exist sufficiently compressible (lawlike) explanations that can unify past, present, and future observations, enabling us to distinguish better theories from worse through observation and criticism.

A revised set of axioms from OpenAI o1 based on a critique by David Deutsch:

Eliezer Yudkowsky's answer to the same question: You only need faith in two things: That "induction works" has a non-super-exponentially-tiny prior probability, and that some single large ordinal is well-ordered. www.lesswrong.com/posts/zmSuDD...

4. Axioms of Probability Theory: These underpin the handling of uncertainty and the updating of beliefs based on evidence, which is vital for both scientific reasoning and self-improvement.

3. Empirical Correspondence Axiom: This posits that observations correspond to an external reality, enabling the connection between theoretical models and the physical world, crucial for scientific inquiry.

2. Axioms of ZFC Set Theory: Zermelo-Fraenkel set theory with the Axiom of Choice (ZFC) serves as a foundation for most of mathematics, allowing for the construction of mathematical objects and structures.

1. Axioms of First-Order Logic: These provide the basic rules of reasoning, such as modus ponens, universal instantiation, and others, which are essential for constructing valid arguments.

Here is what deepseek DeepThink suggested:

3. Cognitive Reliability:

Assume that our cognitive and perceptual faculties are sufficiently reliable for iterative refinement of theories through observation, reasoning, and experimentation, even if they are not infallible.

2. Stable Regularities / Inductive Amenability:

Assume the existence of some degree of stable regularities in the environment, allowing for predictive modeling and inductive inference.

1. Mathematical and Logical Consistency:

Assume a consistent and sufficiently expressive mathematical framework (including probability theory) that can serve as a basis for logical deduction and quantification of uncertainty.

What is a minimal set of axioms that you have to believe in to be able to do science and self-improvement?

Here is what OpenAI o1 suggested:

🤖 The year is 2061—the year humanity perished in a war between gods.

🔥 Earth was gone. A molten sphere glowed where billions once lived. Sterilized. Nothing remained.

⏳ 8 hours later, the black shroud vanished.

🌓 Hours passed. Alex docked at the moon base, where despair hung heavy. Earth was silent. No signals. Civilians with standard implants were catatonic.

⚫ The attack abruptly stopped. The spheres extended black tendrils, interlocking. Earth was shrouded in a dark, impenetrable hull.

⚔️ The firepower was staggering, far beyond humanity’s capabilities. Clearly, someone—or something—had long prepared for this showdown.

🛡️ The spheres didn’t flinch. Their defense was impenetrable, neutralizing every attack—even shots aimed at "empty space."

💥 Then chaos. Projectiles fired at the spheres—blinding impacts that turned space into an inferno. Earth’s hidden railguns were unleashed.

🪐 Black spheres appeared in orbit around Earth. Even from lunar orbit, Alex could see them with his own eyes. Silent. Ominous.

🌐 China-Russia soon followed. It wasn’t a geopolitical attack—this was something bigger. Nearly all of humanity’s implants were offline.

🚨 “DEFCON 1: Massive surprise assault on brain implants underway!” Alex’s secure, military-grade implant rang the alarm in his head.

🌍 India went dark—all of it. Over a billion brain implants severed from the global hive. Moments later, the European Union followed.