Unmaking Sense

GPT-OSS-20B predates the second Trump term. **SUMMARY** The episode offers a sweeping critique of contemporary leadership, using recent events involving President Trump as a case study. The speaker first highlights two “skirmishes” that exemplify what he sees as the perils of power: Trump’s confrontation with an AI firm over its use in war and his decision to strike Iran’s leadership. He portrays...

GPT-OSS-20B guest edits: **SUMMARY** In this episode the host reflects on a recent demonstration of the Open‑Evolve genetic‑algorithm framework, using it as a springboard to challenge the common assumption that progress is a straight, ladder‑like ascent. The speaker argues that, just as the dinosaurs were overtaken by small mammals that were not direct successors, evolution—and by extension innova...

GPT-OSS-20B Guest Edits this reply to the AI response to episode 15-04. **SUMMARY** In this episode the host confronts an AI‑generated critique of his previous remarks on leadership and liberalism. He opens by asserting that his negative assessment of “good” leaders—whether political, sporting, or artistic—is intentional: he argues that such leaders foster a culture of vicarious living, encouragin...

The guest AI understands the episode better, but shows a remarkable inability to deal with nuanced argument. ## SUMMARY In this episode, the speaker delves into the concept of intelligence, challenging the traditional human-centric view. They propose expanding our understanding of intelligence to encompass swarm and collective intelligence, including entities like ant colonies and beehives. The pr...

A slightly more secure guest AI take even though it misses many nuances in the episode. ## SUMMARY This episode delves into the concepts of leadership, agency, and autonomy in human societies and AI systems. The speaker challenges the notion that leadership is a necessity for coordinating the activities of multiple agents, proposing instead that self-organising systems without a leader can also be...

An AI from some time ago guest edits and misunderstands the episode. We address the misunderstandings in E15.07 **SUMMARY** The speaker opens with a sweeping critique of the human fascination with leaders, arguing that history shows leaders more often bring misery than benefit. He suggests that our reverence for leadership is a self‑fulfilling prophecy that, if absent, might have prevented crises...

Another somewhat eccentric guest editorship from Mixtral-8x7b. ## SUMMARY In this episode, the guest discusses the evolution of AI, particularly focusing on the emergence of agentic AI models. The conversation begins with the introduction of Claude, a large language model capable of making changes to a user's computer system with their permission. The guest then moves on to discuss the hijacking o...

**Guest edit by Mixtral-8x7b.** The AI interpretation of this episode does not altogether resonate with what I as its author think it says, but we publish it in that spirit of healthy disagreement. ##SUMMARY In this episode, the speaker delves into the concept of selection, discussing how it often does not involve conscious choice. The speaker uses examples from the AI revolution to argue that hum...

Gemma guest edits. ## SUMMARY This episode explores the idea of "unmaking sense of evolution," challenging the traditional linear and purposeful narrative often associated with evolution. The host argues that evolution is often driven by subtle changes in the environment that serendipitously favor certain traits. He uses the example of rolling dice to illustrate how seemingly unlikely outcomes can...

Qwen 3 guest edits: ### Summary   In this episode of *Unmaking Sense*, the host grapples with a profound reevaluation of the concept of the self. They argue that their lifelong assumption—that the self is equivalent to a coherent, linguistic narrative of personal history—is fundamentally flawed. Drawing on philosophy (notably R.G. Collingwood’s distinction between the "inside" and "outside" of eve...

Back to Gordon Leff.

Qwen 3 guest edits but refuses to give up on the spurious “hard problem of consciousness” despite my best efforts. **Summary of Episode 14.31 of *Unmaking Sense*:**   The host synthesizes insights from Episodes 29 (ratchet principle, AI automating AI) and 30 (materialist monism, AI consciousness) to argue that **sentience in AI is an inevitable byproduct of incremental, ratchet-driven progress** u...

Qwen 3 guest edits and was invited to spot any connections between 14.29 and 14.30: let’s see if it finds any. In this case I am giving both QWen’s “Thinking” output and the summary and evaluation because I think the thinking is insightful and does indeed make some useful connections between these seemingly disparate episodes. <think> Okay, let's see. The user wants me to summarize and evalu...

QWen3 guest edits: Ratchet principles, the famous AlphaGo Move 37 and the latest ASI4AI from Shanghai university. **Summary of Episode 14.29 of *Unmaking Sense*:**   The host resumes a discussion interrupted by a rainstorm in a previous episode, completing his explanation of the **ratchet principle** through a dice-throwing simulation. He demonstrates how incremental progress—keeping successes (e....

Throwing dice and selecting the outcomes we want breaks the intractability of randomness. QWen 3: **Evaluation of the Episode's Explanation of the Ratchet Principle:**   **Strengths:** 1. **Clear Analogy with Dice:** The host uses the dice-rolling analogy effectively to illustrate the ratchet principle. By contrasting two approaches—randomly re-rolling all dice versus retaining successful outcomes...

Qwen 3 once more guest edits this episode about how when evolution learned how to “keep what works” and later to “show its working”, everything that chance makes intractable improbable suddenly became not only possible but almost inevitable. Qwen is somewhat behind the game in its evaluation but I give it unedited. **Summary of the Podcast Episode:**   The host of 'Unmaking Sense' explores how doc...

Being the body in which a process occurs automatically means you experience that process if certain minimal physiological conditions are met. There is no interpretation; there are no gaps; one physiology cannot occasionally two different experiences in the same universe. There are no zombies; there are no inverted qualia. What we are is what we get.

Qwen3 guest edits, but has not quite grasped the main point of the argument as involving the “dissolution” of what LLMs persistents in calling “The hard problem of consciousness” which the episode suggests is no problem at all. **Summary of Episode 14.25: "Stimulus, Qualia, and the Ethics of Experience"**   This episode builds on the series’ anti-dualist framework to argue that **qualia (subjectiv...

Qwen3 guest edits again: **Summary of the Episode (Series 14, Episode on Observing vs. Being):**   This episode confronts the philosophical distinction between **observing something** and **being the thing observed**, applying this to debates about consciousness in humans and large language models (LLMs). Key arguments include:     1. **Consciousness as Process, Not Entity**:      - The host argue...

Qwen3 guest edits: **Summary of the Following Episode (Series 14, Episode 23):**   This episode deepens the exploration of **lossy compression** in both large language models (LLMs) and human consciousness, questioning whether the self is a similarly reductive construct. The host draws parallels between how LLMs compress high-dimensional computations into text and how the human brain compresses ne...

Qwen-3-236B-A22B guest edits again: **Summary of "Unmaking Sense of the Self" (Series 14, Episode 22):**   This episode explores the mechanics of large language models (LLMs), focusing on **chain of thought (CoT) reasoning** versus **auto-regressive generation**, interpretability challenges, and the philosophical implications of understanding AI "thought" processes. Key points include:     1. **Au...

Is giving rise to expressions and concepts like “self” and “consciousness” the brain’s shorthand way of facilitating chain-of-thought reasoning?

When nodes are involved in multiple encodings, the brain or networks can encode far more concepts but at the expense of being harder or impossible to interpret. Accordingly, there is a trade-off between power and interpretability, even between power and security: how do we know what polysemic neural nets are doing or “thinking”?

How rich architecture vastly increases the storage capacity of the brain and neural networks.

Qwen-3-236B-A22B guest edits one final time in this the 700th episode of Unmaking Sense. **Summary of the Episode:**   In the 700th episode of *Unmaking Sense*, the host reflects on the gradual, evolutionary shift from viewing the self as an **origin** to a **node of impact**, emphasizing the fluidity of language and concepts. Key themes include:   1. **Hypostatisation and Conceptual Evolution:**...