Jim Cramer used his Mad Dash segment on June 18, 2026 to flag what he sees as a serious shift in the AI talent war. Noam Shazeer, co-lead of Google’s Gemini effort and co-founder of Character AI, is leaving Alphabet (NASDAQ:GOOG | GOOG Price Prediction ,NASDAQ:GOOGL) to join OpenAI. Cramer’s read was blunt.
He said: “This is a coup. This is a coup. And I’m not joking around… it says OpenAI still completely on its game, right?” He noted that Alphabet’s stock was not really getting hit on it, which he thought was strange. Looking at the price action, he has a point.
Why Cramer is calling it a coup
Shazeer is a singular researcher and a co-author of the 2017 paper. He is a co-author of the 2017 paper “Attention Is All You Need.” Moreover, he worked on the Transformer architecture that every modern large language model, including GPT and Gemini, is built on. He left Google to start Character.AI, then returned in 2024 to co-lead Gemini, the program Alphabet has spent the last two years building its entire enterprise pitch around.
Cramer’s framing, that Sam Altman has wanted to work with Shazeer since OpenAI’s earliest days (per a reported comment from Alphabet finance chief Ruth Porat), gives the move a personal arc beyond compensation.
The talent question matters because Alphabet is spending like the answer to AI is mostly capital. Q1 FY2026 capex more than doubled to $35.67 billion, and full-year 2026 capex guidance sits at $175 billion to $185 billion. Compute is fungible. Researchers who can shape the next architecture are scarce.
What the Gemini program actually looks like right now
Alphabet is losing Shazeer from a position of strength. Gemini 3 launched as a major milestone, the Gemini App crossed 750 million monthly active users in Q4 2025, and the platform was processing 16 billion-plus tokens per minute by Q1 2026, up 60% quarter over quarter.
Paid subscriptions hit 350 million. Google Cloud grew 63% year over year to $20.03 billion, with backlog nearly doubling sequentially to $460 billion-plus. Enterprise demand keeps showing up in the wild. HSBC said this week it expects more than $100 million in gains from deploying Google AI tools, with plans to roll out 200-plus Gemini use cases over the next two years.
The question is whether the loss of one architect, however celebrated, dents a research bench that just shipped those numbers. Probably not in the next quarter. The risk is longer-dated. Architectures move in jumps, and the people who design the jumps are concentrated in a handful of labs. You can see the full Q1 results in Alphabet’s earnings exhibit filed with the SEC, including the $109.90 billion revenue figure and $5.11 EPS that beat the $2.63 estimate.
What the market is saying, and not saying
Cramer was right that the reaction has been muted. GOOGL slipped 0.42% in premarket trading on the Shazeer news, with ticker sentiment scoring at -0.24, somewhat-bearish. But the broader market closed up. Shares finished at $368.03, up 1.17% on the day, up 2.87% on the week, and up 17.73% year to date.
Over twelve months, the stock has more than doubled, gaining 112.95%. Analysts have 14 Strong Buy and 43 Buy ratings against 7 Holds and zero Sells, with a target price of $432.83.
That is the market saying the talent story is real but the cash flow and distribution moat are bigger. Cramer’s instinct is that this is mispriced, that losing a Transformer co-author should register. The counter is that Alphabet’s compute commitments, its YouTube and Search distribution, and a moat on $422.5 billion in trailing revenue are not unwound by one resignation.
What to watch from here
Three things are worth tracking.
First, whether the Gemini release cadence slows in the second half of 2026, because that is the only objective signal of whether Shazeer’s absence is felt at the model level.
Second, whether other senior Gemini researchers follow him, since coups in research labs tend to come in clusters.
Third, the UK search rules and other regulatory pressures that were also cited as overhangs this week. The talent war is real. So is the moat. Investors weighing GOOGL against the broader AI race need both sides on the page.
Top Gaining Stocks
Top Losing Stocks