29 Jun, 2026

Log 019 - The Great Rotation

Bitcoin and Strategy weakened as risk capital chased faster AI trades. Google and Apple lost senior talent to Anthropic and OpenAI. Micron showed how AI is absorbing the memory market. And Meta launched $299 smart glasses. Capital, people, and silicon all kept moving toward AI.

In This Log

1. Crypto Loses the Bid (Crypto)

Bitcoin fell below $60,000 on June 24 and stayed under pressure the next day. Strategy dropped for seven straight sessions, its worst stretch since November 2022, while its preferred stock slid far below par. 1

The weakness was bigger than one company. Bitcoin is down more than 30% this year, spot bitcoin funds have seen roughly $5.2 billion in outflows through June 29, and momentum money has been chasing faster trades in AI-linked stocks, commodities, and the SpaceX IPO. 2

Strategy still holds a huge Bitcoin stack, but the machine around it is less clean. The premium, the preferred stock, and the constant buying habit all work better when investors are willing to pay up for the trade.

That willingness is fading. Crypto is not only fighting price pressure now. It is fighting attention pressure.

Why it matters

Bitcoin's biggest corporate buyer no longer looks like an automatic bid. If Strategy's funding costs rise while ETF money leaves, crypto loses one of the forces that used to absorb selloffs. The rotation into AI becomes a market structure problem, not just a narrative.

Reality check

This is stress, not collapse. Strategy is not forced to sell, Bitcoin has recovered from worse drawdowns, and ETF outflows can reverse. Rates, inflation, and the crypto cycle are doing real work too. AI rotation is not the only explanation.

2. Talent Chooses the IPO Labs (AI)

Two key Gemini researchers, Jonas Adler and Alexander Pritzel, are leaving Google for Anthropic. The same talent wave already includes Noam Shazeer moving to OpenAI and AlphaFold scientist John Jumper moving to Anthropic. 3

Apple took its own hit. Paul Meade, the senior hardware vice president who led Vision Pro engineering and Apple's smart-glasses work, is leaving for OpenAI's hardware division. 4

The pull is not mysterious. Google and Apple can pay very well, but their equity is tied to mature public companies. Anthropic and OpenAI can offer pre-IPO upside in the labs now seen as the center of the next platform shift.

That makes the talent drain different from normal job-hopping. The people leaving are not peripheral. They worked on the models, the science, and the hardware that the incumbents need to compete.

Why it matters

The AI race is being financed in people as much as capital. Pre-IPO equity lets the labs hire the exact researchers and hardware leaders incumbents need to keep. The talent flow points in the same direction as the money flow.

Reality check

Departures do not decide a race by themselves. Google and Apple still have deep benches, distribution, cash, and infrastructure. Startups can also overhire, misintegrate teams, or disappoint employees if the IPO math changes.

3. Micron Shows the Memory Tax (Markets)

Micron posted fiscal third-quarter revenue of $41.46 billion, up from $9.3 billion a year earlier, with adjusted earnings of $25.11 a share and gross margin of 84.6%. The company also guided to $49 billion to $51 billion in revenue for the current quarter. 5

The driver is the AI memory shortage. Micron is one of the market's hottest stocks this year, and rising memory costs are already forcing device makers to talk about higher consumer prices. 6

High-bandwidth memory has changed the business. Memory used to be a brutal commodity cycle. Now the AI data center is turning it into scarce infrastructure.

That scarcity does not stay inside data centers. When the most profitable demand is AI, supply moves there first. PCs, phones, and other devices get the leftovers at higher prices.

Why it matters

This is the rotation made physical. AI is not only taking investor dollars; it is taking wafers, memory, and manufacturing priority. The buildout raises the price of ordinary devices because the same supply chain has to feed the data center first.

Reality check

Memory is still cyclical. High margins invite capacity, and capacity can turn shortage into glut. The consumer price effect is real but uneven, and AI customers can slow spending if returns disappoint.

4. Meta Takes Glasses Downmarket (XR)

Meta launched three new Meta-branded smart-glasses frames: Adventurer, Fury, and Starfire. They start at $299, drop the Ray-Ban and Oakley branding, and are still made with EssilorLuxottica. 7

The glasses run the latest Muse Spark-powered Meta AI experience, add more languages, improve fit, and push the product closer to a normal consumer accessory than a headset. 7

Meta's bet is not that consumers suddenly want heavy AR. It is that cheap camera glasses with AI, audio, translation, and social capture can become normal before full displays are ready.

Why it matters

The consumer AI device race is splitting by price. Meta is choosing the mass-market path: useful, wearable, and cheap enough to buy casually. That gives it a much cleaner route to volume than expensive face computers.

Reality check

These are still camera glasses, not true AR. The category remains small, privacy concerns remain obvious, and Meta's hardware business still burns money. A lower price helps adoption, but it does not prove daily need.

Signals

OpenAI Starts Patching the Planet

OpenAI expanded its Daybreak cybersecurity push with Patch the Planet, offering security help to open-source maintainers and rolling out stronger cyber-focused model access for approved partners. 8

SpaceX Tests Starfall

SpaceX launched Starfall, a new uncrewed reentry capsule built to return up to 2,200 pounds of cargo from orbit for research and in-space manufacturing. 9

SK Hynix Taps Public AI Money

SK Hynix is seeking to raise more than $29 billion through a US listing, with proceeds aimed at advanced memory expansion as AI keeps demand tight. 10

Alphabet Joins the Dow

Alphabet is replacing Verizon in the Dow Jones Industrial Average, a symbolic handoff from old telecom to the AI-era platform economy. 11

Meta-Thread

The rotation had three channels: money, people, and materials. Money moved out of crypto and toward AI trades. People moved out of Google and Apple and toward the labs. Memory moved out of the general device market and toward the data center.

The pattern is not additive. AI is growing by pulling on the rest of the system. It pulls capital from other risk assets, talent from incumbents, and silicon from consumer supply chains. The open question is whether that concentration creates the next productivity boom, or leaves every other frontier thinner when the cycle turns.

Next Log drops next week.

© 2026 AELIUM // Nothing here is advice // readable by humans and agents