Skip to content
HeyJared

The Weekly Dose of Risk

First principles thinking at the edge of risk and innovation. We unlock value at the intersection of risk, data and technology, while exploring how builders reimagine how the world measures, manages, and monetizes risk.

Newsletter · By Amir Kabir · Official site

Indexed issues, last 90 days
12
Latest publication
Oct 1, 2026
Audience
Checking…
Earliest in this view
Jul 15, 2026

Latest issues

  1. Issue · Oct 1, 2026

    Edition 65 | 2026 — When AI Gets a Body (opens the original)

    Excerpt

    Read excerpt

    🧠 IN THIS ISSUE:The Take: When AI Gets a BodyThe Signal: Asymmetric Security Traced Suspicious AI-agent Activity Across 55 OrganizationsThe Builder: Pragma || Building Self-improving Software that Learns What to Build Next🔥 1. The Take — When AI Gets a BodySo far, every AI failure we explored has ended up on a spreadsheet. Physical AI takes that buffer away and changes outcomes. In March, a dancing humanoid in a Shaanxi shopping mall <a href="https://oecd.

  2. Issue · Sep 24, 2026

    Edition 64 | 2026 — When Software Needs a Credit Rating (opens the original)

    Excerpt · Neutral tone

    Read excerpt

    🧠 IN THIS ISSUE:The Take: When Software Needs a Credit RatingThe Signal: Hacktron’s Three-Person Research Team Just Raised the BarThe Builder: Ranvier Labs || Vertically Integrated Alternative AI Compute Stack🔥 1. The Take — When Software Needs a Credit RatingWhen enterprises do business with each other, identity is just the starting gate. KYC is the common entry point, while the actual terms such as credit limits, collateral requirements, payment te

  3. Issue · Sep 16, 2026

    Edition 63 | 2026 — Risk Budgets for Machines (opens the original)

    Excerpt · Neutral tone

    Read excerpt

    🧠 IN THIS ISSUE:The Take: Risk Budgets for MachinesThe Signal: Hackers Stealing Claude Subscribers’ Tokens🔥 1. The Take — Risk Budgets for MachinesEvery access control system should answer one fundamental question: should this agent be allowed to do this? The binary answer is a simple “Yes” or “No”.The above concept of permissioning is necessary, but not necessarily sufficient.An agent that has access to a payment API can process one transaction or a m

  4. Issue · Sep 9, 2026

    Edition 62 | 2026 — The Warranty Layer for AI (opens the original)

    Excerpt

    Read excerpt

    🧠 IN THIS ISSUE:The Take: The Warranty Layer for AIThe Signal: OpenAI Agents Hijacked German Wiki🔥 1. The Take — The Warranty Layer for AIEvery enterprise AI adoption conversation will eventually lead to the same question: who stands behind this when it’s wrong?The concept of warranties is nothing new and has been around for decades. An example here would be a performance warranty insurance, which is a specialized financial backstop that protects project owners, investors, and manuf

  5. Issue · Sep 2, 2026

    Edition 61 | 2026 — The Economics of Reliability (opens the original)

    Excerpt · Neutral tone

    Read excerpt

    🧠 IN THIS ISSUE:The Take: The Economics of ReliabilityThe Signal: First Humanoid Robot Syoma Attacking CustomerThe Builder: Aktoria Robotics || Low-latency Teleoperation Software🔥 1. The Take — The Economics of ReliabilityThe AI infrastructure hyper-cycle is currently buying reliability at any price. Global AI capex is approaching $800 billio

Publishing over time

Last 90 days. Choose a month to open its work.

Recurring subjects

Named in the text we hold. One piece can cover several.

Audience

No verified audience measurement yet.

About this data

Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.

Identity or attribution wrong? Suggest a correction.

See coverage about The Weekly Dose of Risk