HeyJaredCash & Cache
Where strategy meets execution — exploring how AI is transforming businesses, markets, and value.
- Indexed issues, last 90 days
- 4
- Latest publication
- Aug 6, 2026
- Audience
- Checking…
- Earliest in this view
- Jul 16, 2026
Latest issues
Your Claude Skills Are Passing Every Check and Still Wrong. (opens the original)
Read excerpt
A blank permissions field almost gave a scheduled AI task full access to my Gmail, Slack, and calendar.TL;DR: Claude Skills go quiet-broken, not loud-broken — they keep running, keep producing plausible output, while silently ignoring an instruction or reading from the wrong place. This piece covers four real instances of that failure found in one afternoon, at four different layers of a real build, and the scheduler I built to catch them automatically: what it checks, how often, and what it rou
Your AI Sounds Confident About Your Portfolio. Make It Prove It. (opens the original)
Read excerpt
Four years on a desk taught me there's no correct portfolio. Only weights, and what each one quietly costs you.🧠TL;DR: Asking AI “what is a bond” gets you a fluent, generic, useless answer. Instead: ground it in the real document (Gemini Notebook, which only answers from what you give it and cites the passage), then make a second AI (Claude) argue with that answer. Four reusable prompts, one risk-reward matrix, and you being able to conduct a stress-test on what your portfolio could look like ba
The 3-Layer AI Accuracy Framework (Most People Only Use 1) (opens the original)
Read excerpt
The output didn’t get more accurate. Instead, it got more convincing.🧠 TL;DR: AI accuracy fails for a boring reason. Polished output makes us stop checking it. This piece gives you three checkpoints to run instead: constrain what the AI sees, verify with real independent cross-checks instead of “are you sure,” and force yourself to engage with the unfinished parts before you act on anything. Paid subscribers get a Claude Skill that builds and calibrates all three automatically for whatever recur
Your Org Remembers Everything. It Understands Less. (opens the original)
Read excerpt
🧠TL;DR: AI is solving institutional memory. Meetings recorded, decisions documented, context retrievable in seconds. The paradox: organizations may remember more while understanding less. Neha Kabra, McKinsey partner and returning Cash & Cache collaborator, introduces the Context Stitching Layer and explains why. I close with what this means for anyone building AI workflows right now.On Cash & Cache Live last month, drew a distinc
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.