TODAY'S SIGNALS · SUN NOV 22, 2026 · ISSUE 25 · COVERS NOV 21–22
Today's Signals
Weekend read: Visser says this year is about doubting public stocks because of AI, Bitcoin trades below its estimated mining cost, and an AI-assisted math proof passed computer checking.
Today's Signals
Signal 01
Jordi Visser (on The Pomp Podcast): Visser: "this year has been about doubting public stocks" because of AI
Who this is
Jordi Visser runs Visser Labs, writes on AI and macro, and spent decades on Wall Street (background not rechecked this run). Anthony Pompliano hosts The Pomp Podcast.
What happened
In this week's Pomp Podcast episode, Visser discusses why markets are falling: AI pressure on capitalism, cyber attacks shaking markets, the AI stock selloff, the Anthropic IPO, and why Bitcoin still matters (episode description). Pomp's 93-second clip sums up Visser as saying public stocks face serious trouble. The wording was checked as quoted in Pomp's post; the video itself was not checked.
“If everyone takes one thing away from this, this year has been about doubting public stocks...and its because of AI.”Jordi Visser as quoted in Anthony Pompliano's post on X (verbatim from the post; video not checked). · VERIFIED
Why it matters
His claim is that AI makes any company's long-term future hard to trust, so investors pay less for public stocks overall. If AI can disrupt almost any business quickly, 'safe, steady' stocks may not feel so safe. This is education, not advice.
Next dominoes
If the selloff keeps hitting companies across many sectors, not just AI names, that supports his 'doubt in public stocks' view. If AI leaders rebound while old-economy stocks lag, the story becomes winners versus losers. The Fed meeting on Dec 9 is the next big rates event.
The other side: doubt in public markets has shown up near bottoms before, stocks are now in a correction, and tech valuations are reported at multi-year lows, so much fear may already be priced in. AI could also raise profits for companies that adopt it.
Click here to copy and paste into your AI for further conversation.
Signal 02
Anthony Pompliano: Bitcoin below its estimated mining cost; Pomp: "Bitcoin's volatility works in both directions."
Who this is
Anthony Pompliano hosts The Pomp Podcast and runs investment firms; he is a long-time Bitcoin advocate. JPMorgan's note was led by strategist Nikolaos Panigirtzoglou (per The Block).
What happened
JPMorgan estimates the average cost for public miners to produce one Bitcoin at about $86,000 (Nov 18 note, reported). With Bitcoin reported around $59k to $61k this weekend, higher-cost miners may sell reserves or shift more machines to AI data-center work (reported). Pomp says the 48-hour drop is driving fear but Bitcoin can recover just as fast. Only Pomp's wording was checked; the JPMorgan figure and prices are reported.
“Bitcoin's volatility works in both directions.”Anthony Pompliano, post on X (verbatim). The JPMorgan estimate and prices are reported, not checked. · VERIFIED
Why it matters
When Bitcoin trades below the cost to mine it, the weakest miners get squeezed first. Miners that also host AI computers have a second way to earn; pure miners do not. This is education, not advice.
Next dominoes
If weaker miners shut machines off, Bitcoin's network automatically makes mining easier (difficulty adjusts about every two weeks), which lowers costs for the rest. If miners sell reserves into a weak market, that adds selling pressure. Watch for miner AI-hosting deals or shutdowns.
The other side: 'production cost' is an average estimate, not a floor. Bitcoin has traded below such estimates before and stayed there for months, and the network's self-adjustment can close the gap without any price rise.
Levent Alpoge is a mathematician at Anthropic (per Scientific American). The Lean and Mathlib communities maintain the proof-checking software used to verify math by computer. Elon Musk is CEO of Tesla and SpaceX and runs xAI.
What happened
On Nov 20, Alpoge released a 17-million-line computer-checked proof of the Riemann Hypothesis, a 167-year-old problem about prime numbers, built with Anthropic's Claude. The Lean and Mathlib communities say the code compiles and proves the exact standard statement. Musk called it 'yet another mind-blowing thing achieved by super intelligence' (his post, Nov 21, 10:49 AM PT). Only the Lean statement wording was checked.
“We have full confidence that Alpöge's proof of the Riemann Hypothesis is correct, and we congratulate him on this achievement.”Lean community blog statement (verbatim). · VERIFIED
Why it matters
A problem the best human mathematicians could not settle for 167 years was cracked with AI help and double-checked by a computer. The 'AI is overhyped' story and the 'AI is a real breakthrough' story are both running while stocks fall. This is education, not advice.
Next dominoes
If mathematicians accept a human-readable version, expect more AI-plus-proof-checker work on other famous problems. Anthropic just went public, so this could shape how investors value AI labs. It also fits Visser's point: if AI can do this, few businesses are safe from disruption.
The other side: a computer check confirms the logic, but mathematicians still need to understand the ideas, and that review takes time. A math breakthrough does not directly translate into revenue for anyone.
Click here to copy and paste into your AI for further conversation.
Signal 04
Financial Times (NVIDIA / Megaspeed report): FT investigation ties a Megaspeed-linked network to China chip flows
Who this is
The Financial Times is a global business newspaper. Megaspeed is a Singapore-based company that was a large NVIDIA customer in Southeast Asia (reported).
What happened
An FT investigation (reported Nov 22) says a network linked to former Megaspeed executive Alice Huang, reported as Jensen Huang's partner, used shell companies and gaming firms to route restricted NVIDIA chips to China. Reports say Megaspeed recorded about $4.6B of imports from Aivres, the U.S. unit of blacklisted Chinese server maker Inspur, in 2025, and that NVIDIA blocked sales to Megaspeed in 2025. The FT article is paywalled and was not read; details come from secondary reports.
Why it matters
NVIDIA's problem is shifting from leadership drama to whether export rules were broken through its customers. When the most important AI chip company is under export and accounting scrutiny, the whole AI trade feels riskier. This is education, not advice.
Next dominoes
If Commerce adds Megaspeed to the trade blacklist (the FT reported it was delayed), more chip sales into Southeast Asia could face checks. If NVIDIA shows it blocked these buyers early, the damage may stay with the intermediaries.
The other side: smuggling through third parties affects every chipmaker, and reports say NVIDIA cut off these buyers in 2025, so this may be old exposure, not new risk.
Click here to copy and paste into your AI for further conversation.
Worth a look (not checked)
Signal 05
Binance: Proof of reserves
Who this is
Large crypto exchange.
What happened
Binance says its latest proof that it holds customers' crypto one-to-one is ready. No one outside Binance has checked it here, and a proof like this is not a full audit.
Why it matters
Proof that an exchange holds what it owes is a trust signal in crypto, once it's checked independently.
Next dominoes
Watch for an outside review.
Nov 6, 2026 · post on X (carried) · UNVERIFIED Binance on X
Click here to copy and paste into your AI for further conversation.
Signal 06
Oklo: Q3 2026 results
Who this is
Advanced nuclear company. Builds small reactors aimed at power for AI and industry.
What happened
Oklo released its third-quarter results on Nov 6. We haven't checked revenue or earningsEarnings: A company's profit. An earnings report is its regular update on sales and profit. Glossary against the company's own tables, so we're not printing them.
Why it matters
Oklo is watched for nuclear power aimed at AI. Not advice.
Next dominoes
Next check: the company's financial tables or filed quarterly report.
Nov 6, 2026 · company newsroom (carried) · UNVERIFIED Oklo newsroom
Click here to copy and paste into your AI for further conversation.
Weekly closes · New week: prices as of Fri Oct 30 close
Ticker
Tied to
Fri Oct 23
Fri Oct 30
Last week
NVDA
Huang
210.96
204.64
-3.0%
TSLA
Musk
372.22
399.81
+7.4%
BLK
Fink
1,118.25
1,147.54
+2.6%
IBIT
Fink/crypto
45.96
43.36
-5.7%
META
Ives
642.62
614.99
-4.3%
AMZN
Camillo
248.00
240.35
-3.1%
ETN
Policy: transformers/grid
366.11
369.97
+1.1%
VRT
Policy: power/grid
194.32
232.50
+19.6%
MP
Policy: China mineral pause
38.30
39.57
+3.3%
AMAT
Policy: chip export rules
377.24
390.09
+3.4%
ASML
Policy: chip export rules
1,571.80
1,565.90
-0.4%
A new week starts today. These are Fri Oct 30 official closes vs Fri Oct 23 closes, until today's close. Source: Yahoo Finance daily history, pulled Oct 30, 2026, about 1:40 PM PT. Prices only; no advice. Yahoo Finance
Gaps
Left out: no posts found from Jensen Huang, Larry Fink or Chris Camillo in the window, and no posts of his own from Jordi Visser (covered through Pomp's episode). Dan Ives posted only TV promos, not used. Elon Musk posted 'Tesla FSD V15 is a banger release!' (x.com/elonmusk/status/2124294755316310346, Nov 21, 4:11 PM PT, verified wording); a fan claim that V15 is '25x safer' than V14 is unverified. A Kraken flash crash on thin Canadian-dollar pairs briefly priced Bitcoin far lower before recovering; Kraken says trades processed correctly (news summary, not checked). Markets are closed this weekend. Still open: Bitcoin fund flows (reported November total about -$3.47B; Nov 20 figure not checked), whether Anthropic holds above its $83 IPO price, NVIDIA's CEO question and SEC probe, any Commerce statement on Megaspeed, the Dec 9 Fed meeting, the MSCI and Strategy deletion after the Nov 30 close, outage root causes, and Oklo's figures (unverified). This is educational research, not financial advice.
VERIFIED quotes were checked word for word against the linked source. UNVERIFIED items and paraphrases are labeled. Neutral briefing; no opinions. Calendar · Archive.