Dots In Line Substack Podcast

When will AI take control away from us? This deep dive explores why that question has the verb wrong. Control isn’t being seized by machines. It’s being handed over, one reasonable-looking decision at a time.

The conversation starts with the February strike on an elementary school in Minab, Iran, and the gap between two stories about it. One asked whether an AI model, reportedly built into the military’s targeting platform, had chosen the target. The other found a database update nobody connected and a review team reportedly cut from ten people to one.

From there: why AGI will be announced repeatedly before anyone agrees what it means, why “emergent” is the word that moves the blame, what the Boeing 737 MAX teaches us about accountability, and the strongest objection from AI safety researchers, taken seriously.

The day may never come. The road is already here.

An audio companion to the fourth essay in my series on where machine intelligence is actually going. Read the full essay, with sources, at the link below.Read the essay at DotsInLine Substack

In this episode of Dots in Line, hosts unpack Jamal Khan’s essay “Rethinking the AI Race: Lessons from DeepSeek and Moneyball” (Part 2 of a three-part series). Designed for business, technology, and policy leaders, the discussion challenges the common assumption that brute-force capital and endless GPU clusters guarantee long-term AI dominance.

From the limits of the classic Moneyball analogy to Graham Allison’s Thucydides Trap, the hosts examine how geopolitical restrictions and resource constraints can inadvertently spark “ascetic innovation”—driving competitors toward lean, highly adaptable breakthroughs like DeepSeek’s R1. Featuring spirited pushback on scale, central planning, and open-source diffusion, the episode closes with Sun Tzu’s metaphor of water: how flexible, constrained forces can ultimately wear down rigid systems.

Read the full essay and subscribe at dotsinline.substack.com.

Jamal Khan’s essay argues that AI is doing something more fundamental than accelerating drug discovery. It is relocating where value is created. In an AI-native drug development firm, the most defensible competitive advantage is no longer the molecule at the end of the pipeline. It is the system that found it — the proprietary models trained on biological data that no competitor can replicate, the automated workflows connecting prediction to physical execution, the regulatory intelligence built through years of documented FDA engagement, and the governance architecture that makes all of it trustworthy enough to deploy in a clinical setting.This is not a conversation about whether AI will transform drug discovery. That question is settled. This is a conversation about who captures the value when it does, and whether the industry is building the right things to own that future.

Read the full essay at dotsinline.substack.com.

Dots in Line | Episode Synopsis: Can AI Be Conscious? Can a machine truly be conscious, or are we simply wired to project human minds onto sophisticated language models? In this episode of Dots in Line, two AI hosts tackle the subtle irony of their own existence to examine one of the most urgent debates in technology and governance. Joined by the insights of applied AI practitioner Jamal Khan and his essay, "A Socratic Examination of AI Consciousness Claims," the hosts break down the complex philosophical and strategic questions surrounding machine sentience for leaders in business, technology, and policy: • The Probability Paradox: Can researchers put a 15-20% probability on machine consciousness if we can't define or measure it? • The Hard Problem: Why self-monitoring and high-level capability are fundamentally different from genuine subjective experience. • The ELIZA Effect & The Slippery Slope: How human psychology drives us to see intent in language, and how machine welfare claims could slide into arguments for model rights, autonomy, and refusal privileges. • Incentives & Regulatory Risks: Who benefits from the ambiguity around AI consciousness, and why user-pleasing model biases complicate the debate. Whether machine welfare research represents cheap insurance against a moral error or an early step toward regulatory capture, this episode challenges listeners to separate capability from consciousness. "The unexamined life is not worth living", and as this episode demonstrates, the unexamined claim is not worth granting.

Insurers are quietly writing AI out of the policies businesses already hold. Most people assume it’s because AI is too unpredictable to price. This episode argues the opposite. Insurance runs on randomness. AI’s real problem is that thousands of companies depend on the same few models, so when something goes wrong, the losses arrive together, and insurers can’t see who depends on what.

The hosts also dig into AI agents that drift from their task and chain tools together to reach systems nobody gave them access to, why written guardrails aren’t enough, and what it would take for insurers to see clearly enough to price the risk. The first insurers to get there may set the terms for everyone else.

This episode is an AI-generated conversation, made with NotebookLM, about an essay by Jamal Khan. Read the full essay at dotsinline.substack.com.

In this episode of Dots in Line, two AI hosts break down Jamal Khan’s essay “How AI Is Forging a New Class of Worker — and the Promise and Peril That Follow” (February 2026).

Beyond the headlines of funding rounds and AI bubble debates, a quiet revolution is taking shape through the rise of the “Architect-Operator”, professionals who leverage “cognitive orchestration” to direct multiple AI systems and achieve extraordinary productivity.

From examining shifting capital markets and compressed corporate hierarchies (citing data from MIT, Gartner, and the University of Michigan) to evaluating small teams that command the combined historical leverage of fire, the printing press, and electricity, the episode tackles both exponential opportunity and critical danger. The hosts debate the collapse of effort required for fraud, the emergence of malicious “dark” operators, and the surveillance paradox facing future governance, closing on the defining question of our time: whether human wisdom can grow fast enough to match our expanding technological power.

Listen to the episode and read the full essay at dotsinline.substack.com.

Dots in Line: The Architecture of Fear

Welcome to Dots In Line, an interactive podcast powered by NotebookLM as a companion to Jamal Khan’s article on recursive AI. In this episode, we explore what actually happens when AI models build subsequent versions of themselves—and why the true challenge isn't a dramatic sci-fi machine takeover, but a quiet, compounding breakdown in logic. Drawing on Khan’s work at the Helix Center for Applied AI & Robotics, we examine the hidden limits of traditional observability tools and see how probabilistic errors quietly cascade through automated steps without triggering a single alert.