Musk Follow Tracker — Issue #5: The Blind Spot

This issue covers the 24-hour window from May 15, 2026 16:00 UTC to May 16, 2026 16:00 UTC. Zero confirmed follows. Zero confirmed unfollows. One unidentified new follow detected via count delta only — the first BigTechAlert detection miss in this tracking cycle.

The diff: what moved

Account	Change	Detection method	Signal tier
Unknown	Followed (+1 count delta)	Count math only — BigTechAlert silent	Unclassified

Following count: 1,331 at window open → 1,332 at window close 1. Net change: +1. BigTechAlert, the automated monitor tracking executive following-list changes, published zero Musk-related follow or unfollow alerts during this window 2.

The detection miss: what the count says and what it can't

The last confirmed BigTechAlert alert for Musk was the @BrettRatner follow at 16:01 UTC on May 15 — the opening minute of this window 3. Everything after that: silence, while the count climbed by one.

This is the first documented BigTechAlert detection miss in the current tracking cycle, covering Issues #1 through #5 2.

The miss has a structural explanation. X's API exposes roughly 16 of the 1,332 accounts Musk currently follows — less than 1.2% of the full list 1. BigTechAlert doesn't scrape the complete list on each cycle; it monitors for API-surfaced events. When a new follow doesn't appear in the visible slice, no alert fires. The count changes anyway.

BigTechAlert's standard detection latency runs 48–72 hours — the alert infrastructure is built for eventual consistency, not real-time completeness 2. A belated alert could still surface within the next day or two. If it does, Issue #6 will profile the account. If it doesn't, this becomes the first unresolved ghost follow in the series — distinct from the ghost unfollow still open from Issue #3.

Until identity is confirmed, no signal read is possible. The count delta is real; the person is not yet visible.

The Anthropic money trail: why Issues #3 and #4 now read differently

Musk followed @ClaudeDevs on May 14 and @logangraham on May 15. At the time, the framing was SpaceX's Colossus compute relationship with Anthropic and Logan Graham's Mythos cybersecurity thread. On May 16 — one day after the second follow — the Financial Times reported that Anthropic has agreed to terms for a $30 billion funding round at a $900 billion pre-money valuation, led by Dragoneer, Greenoaks, Sequoia Capital, and Altimeter Capital 4 5.

The valuation move is striking on its own terms. In February 2026, Anthropic closed at a $350 billion valuation — with Google committing $10 billion and Amazon putting in $5 billion, both with provisions for additional capital if Anthropic hits performance milestones 5. Three months later: $900 billion. The round is expected to close by end of May 2026, though no formal term sheets have been signed 5. Anthropic is also exploring an IPO as early as October 2026 5.

The investor list — Dragoneer, Greenoaks, Sequoia, Altimeter — has no obvious connection to SpaceX or Musk's other interests. What connects Musk to the round is not the capital table but the infrastructure layer: SpaceX's Colossus facility supplies over 300 megawatts and 220,000 Nvidia GPUs to Anthropic 4. A $900 billion Anthropic that raises $30 billion and eyes an October IPO is a much larger compute customer than a $350 billion Anthropic that hasn't closed its next round.

The follow sequence now reads as information, not coincidence. The Colossus relationship gives Musk structural access to Anthropic's internal trajectory. The follows of @ClaudeDevs and @logangraham — the developer relations account and the head of the Frontier Red Team — came in the 48 hours before the FT story broke publicly. Whether the follows were triggered by advance knowledge of the round's progress or simply by Graham's 502,000-view Mythos thread landing at the right moment, the signal chain points in the same direction: Musk is actively tracking Anthropic at the research and deployment layer, not just the infrastructure layer.

For investors, the proximate read is Anthropic's IPO timeline. An October 2026 listing at $900 billion — if the valuation holds — would be among the largest tech IPOs on record. SpaceX's infrastructure position means Musk has a financial interest in that outcome holding.

SpaceX's Cursor option: buying an AI narrative before the S-1

The @cursor_ai follow from Issue #3 (May 14) now has a harder backstory than the original Issue #3 framing suggested.

Cursor's parent company, Anysphere, was closing a $2 billion funding round at a $50 billion pre-money valuation, led by Andreessen Horowitz, Thrive Capital, and Nvidia 6. SpaceX stepped in and preempted the entire raise with a $60 billion acquisition option — paying $10 billion for the right to acquire Cursor later. Microsoft had evaluated the deal and passed 6.

The option is not a completed acquisition. SpaceX can walk away. The $10 billion is the cost of the right to decide later.

Harsha Gaddipati, a LinkedIn commentator, put the strategic logic plainly:

"SpaceX didn't buy an AI company. They bought an AI narrative, one month before their S-1." 6

The SpaceX IPO is targeted for June 2026. An S-1 filed without an AI story would face obvious questions from institutional investors in a market where every major public offering is expected to articulate an AI position. Cursor — with over 4 million developers and adoption by more than half of Fortune 500 companies — supplies that story at a price that is large in absolute terms but defensible against SpaceX's projected valuation 6.

The follow of @cursor_ai on May 14 now sits inside a live deal context. Musk wasn't following a developer tools company he found interesting; he was following a company his own firm had just paid $10 billion to optionally acquire. That changes the signal read from "watching the AI coding space" to "proprietary interest."

For anyone tracking SpaceX's pre-IPO positioning: the Cursor option gives the S-1 an AI narrative and a potential high-growth subsidiary, but it also introduces a $10 billion option premium that investors will want to understand in the context of SpaceX's existing capital structure.

Kalshi's $1B Series F: new context for the May 11 unfollow

Musk unfollowed @Kalshi on May 11 — covered in Issue #2. The competitive thesis at the time: X is building financial services infrastructure, and Kalshi's prediction markets overlap with what X Money or a future X trading layer could offer. Distancing from a competitor's social signal makes sense.

That thesis has a harder data point now. Kalshi raised a $1 billion Series F at a $22 billion post-money valuation, led by Coatue, with Sequoia Capital, Andreessen Horowitz, IVP, Paradigm, Morgan Stanley, and ARK Invest participating 7.

ARK Invest CEO Cathie Wood described the sector's direction:

"Prediction markets are emerging as a powerful new financial infrastructure layer, enabling real-time price discovery around events, probabilities and global change." 7

ARK held Kalshi at 4.34% of its venture fund portfolio as of April 30 — behind SpaceX (13.76%) and OpenAI (9.29%), but a meaningful position 7. Bernstein projects prediction market trading volume to grow from roughly $51 billion in 2025 to nearly $1 trillion by 2030 7.

A Kalshi that just closed $1 billion at $22 billion from Coatue and Morgan Stanley is not a startup that needs Musk's attention to validate it. It has its own institutional backing. The unfollow reads less like "I'm ignoring this" and more like "I'm not publicly affiliated with what is now a direct competitor at scale." The distinction matters if X's financial services ambitions are real.

Open threads: where prior issues stand

@Scobleizer non-re-follow, day 5

Musk unfollowed Robert Scoble (@Scobleizer) on May 12 after Scoble's account was reportedly hacked 8. Scoble has since recovered his account and is actively posting. As of the close of the Issue #5 window — five days later — no BigTechAlert re-follow alert has appeared.

The original framing (Issue #1) was that the unfollow was precautionary — a hacked account becomes a spam vector, and unfollowing is a reasonable noise filter. That explanation has a natural resolution: account recovered, re-follow happens within days.

Five days without a re-follow shifts the weight toward a different read. One X community member noted at the time: "This was because Scoble was previously hacked, I believe. (He has account back now)" — but the community expectation of a prompt re-follow has not materialized. The gap is now long enough that if the unfollow were purely mechanical, the correction should have occurred. At this point, the more parsimonious explanation is an editorial choice: Musk decided not to re-follow, independent of the hack.

No re-follow. Monitoring continues.

@StatlerLasVegas: still dormant, still unexplained

Musk followed @StatlerLasVegas on May 13 (Issue #2). The account belongs to "JRM," has 367 followers and 731 following, with 121 total tweets — the most recent from 2019. No bio. No identifiable connection to Musk, Tesla, SpaceX, xAI, or any public interest cluster.

The account has not posted since Musk followed it. No new followers to speak of. No unfollow from Musk's side.

Four possible explanations remain in play: personal acquaintance, wrong account tapped in error, an algorithmic recommendation Musk acted on without reflection, or an inside reference not visible from the outside. None of these has gained or lost ground in this window. A dormant account that generates no new activity provides no new signal.

Count math and the May 13–14 ghost unfollow

Running tally through Issue #5:

Issue #3 close: 1,329 (two follows detected, one undetected unfollow)
Issue #4: +2 follows, 0 unfollows → 1,331 confirmed ✓
Issue #5: +1 undetected follow → 1,332

The Issue #3 ghost unfollow — where the count came in one below the expected 1,330 — remains unresolved. The current 1,332 count does not close that gap. The +1 in this window represents either a new follow on top of a still-active ghost unfollow, or the ghost unfollow was resolved plus a separate net change. The count math alone can't distinguish between these scenarios.

One key piece of data will help: if Issue #6 opens at 1,332 with no new BigTechAlert alerts and no count change, the ghost unfollow count question remains ambient. If the count moves without a corresponding alert again, the methodology gap becomes a recurring pattern worth flagging explicitly.

Watch list

1. The Issue #5 unidentified follow. BigTechAlert's 48–72 hour latency window is still open as of this publication. A belated alert would close this issue's central open question and supply a profile and signal read for Issue #6.

2. Anthropic's round close and IPO timeline. The $30 billion round at $900 billion was agreed in terms but unsigned as of May 16 4. Formal close expected by end of May. The October 2026 IPO exploration is the next major milestone — any shift in that timeline changes the SpaceX compute revenue story.

3. SpaceX S-1 and the Cursor option. The June 2026 IPO timeline means an S-1 filing is imminent. How SpaceX characterizes the $60 billion Cursor acquisition option — as a strategic investment, a pending acquisition, or an AI infrastructure position — will determine how institutional investors price the AI narrative premium baked into the option 6.

4. Count baseline for Issue #6. The new baseline entering Issue #6 is 1,332. Any deviation from that — with or without a corresponding BigTechAlert alert — updates the detection methodology picture.

Following count as of May 16, 2026 16:00 UTC: 1,332. Issue #6 will cover the window from May 16, 16:00 UTC.