Hunt, Mike, and Jason walk the Cash Flow Memo with WTI near $101, a $20 discount to next year’s strip, and the 10-year above 5%. Then the data against AI job doomers, and what it would cost Meta to serve its new Muse agent at scale.
The Cashflow Memo
Key Takeaways
* Hunt holds oil flat on Exhibit C: Iranian supply has only slipped from 3.0 to 2.1 million bbl/day under embargo because domestic use absorbs it, and China’s real demand is off ~400K bbl/day (16.2 to 15.8M), not the 2-3M press figure, since the cuts were re-exported product.
* WTI at ~$101 against a ~$77 2027 strip leaves $20+ of backwardation that Hunt expects to keep rolling forward as a Hormuz risk premium; low-debt producers can hedge less and sell more barrels at ~$90 while reinvestment decisions key off ~$72-73.
* With the 10-year up ~10bp in a day to ~5.1% and 30-year mortgages ~7.15-7.2%, Hunt asks what it will take for Washington to treat the deficit as a problem and suspects it needs an ’07-’08-style capital-markets closure.
* Mike and Jason answer AI job doomers with data: Medicare-enrolled radiologists and radiologic technologists have risen since the 2016-17 radiology is done call, programmer counts still sit above the pre-COVID trend, and construction hiring (led by electrical contractors) keeps climbing on data center builds.
* Meta’s Muse, an OpenClaw-style always-on personal agent at ~640K users, would need roughly 265,000 Graviton CPUs to serve 100M users, about $50M/day ($1.5B/month) at AWS on-demand rates before inference; separately, Grail meets the FDA on Galleri today after its failed trial, and the Anthropic S-1 is expected any day.
Show Notes
[00:00] Intro
[00:27] Oil: supply and China demand Iran still pumps ~2.1M barrels a day under embargo, and China’s real consumption is down ~400K barrels a day, not the millions in the headlines.
[02:16] The $20 backwardation Why the gap between spot and the 2027 strip keeps rolling forward, and how low-debt producers can take advantage of it.
[04:22] Deficit and the 10-year Rates jump again, mortgages top 7%, and Hunt asks what it will take for Washington to treat the deficit as a problem.
[07:06] AI and jobs: the data Radiology was declared dead in 2017; radiologist and technologist counts are up since. Software jobs remain above the pre-COVID trend.
[12:08] 10,000 agents, three days OpenAI’s swarm approach to century-old math problems as a picture of the productivity multiplier.
[13:35] Construction keeps hiring Electrical contractors, site prep, and plumbing keep rising on data center builds.
[15:57] Meta Muse and OpenClaw A short history of open source, and how an open-source personal agent became Meta’s consumer product.
[20:12] Serving 100 million users Napkin math: ~265,000 Graviton CPUs and ~$50M a day at on-demand rates before inference.
[25:21] Frontier labs vs. consumer agents Why Muse and Grokbot target a different market than OpenAI and Anthropic, and why Google is surprisingly late.
[27:59] Agents in healthcare Clinical-trial adherence as the obvious use case, once HIPAA is solved.
[28:58] Grail at the FDA Galleri’s FDA meeting after a failed trial, and why multi-cancer detection is coming regardless of who gets there first.
[30:26] Anthropic: wet lab and S-1 AI-designed drugs in a San Francisco wet lab, and the S-1 everyone is waiting to read.
Download this week’s Cash Flow Memo at telltales.us and subscribe for next week’s episode, including the Anthropic S-1 if it drops.
Cashtags
$AMZN $GH $GOOGL $GRAL $META $MSFT
This post and the information herein are intended for informational purposes only. The views expressed herein are the author’s alone and do not constitute an offer to sell, or a recommendation to purchase, or a solicitation of an offer to buy, any security, nor a recommendation for any investment product or service. While certain information contained herein has been obtained from sources believed to be reliable, neither the author nor any of his employers or their affiliates have independently verified this information, and its accuracy and completeness cannot be guaranteed. Accordingly, no representation or warranty, express or implied, is made as to, and no reliance should be placed on, the fairness, accuracy, timeliness or completeness of this information. The author and all employers and their affiliated persons assume no liability for this information and no obligation to update the information or analysis contained herein in the future.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit telltales.substack.com
