Two researchers. Two exits. One brutal week for Google.
On June 18, Noam Shazeer — co-author of the landmark 2017 "Attention Is All You Need" paper, the foundational work behind every major AI model you use today — announced he was leaving Google DeepMind for OpenAI.
Shazeer had returned to Google in 2024 after the company paid roughly $2.7 billion to bring him back from Character.AI, the chatbot startup he co-founded. He lasted less than 22 months.
Sam Altman called it a hire he had "wanted since the very beginning of OpenAI." Shazeer's role is Lead for Architecture Research — the person responsible for the neural network structures underlying all of OpenAI's models.
Then it got worse.
Just two days later, AlphaFold lead John Jumper departed Google DeepMind for Anthropic — meaning two foundational AI researchers left Google for direct competitors in the same week, at a pace the industry has not seen before.
Bloomberg reported that Alphabet stock fell as much as 7.2% intraday on the news, the steepest single-day drop since February 2026.
Here's why that number stings: it's not just a talent story. It's a credibility story.
As of June 24, Gemini 3.5 Pro has still not launched. Google CEO Sundar Pichai committed explicitly at Google I/O in May to a June 2026 release date.
A miss past June 30 would mark the second consecutive I/O commitment the company failed to deliver on schedule.
Google is running out of June fast, and now it's doing so without two of its most recognizable research names.
Gemini 3.5 Pro's confirmed specs are significant — a 2-million-token context window (the largest of any production frontier model), a Deep Think reasoning mode, and full multimodal support.
If it actually ships, it would be a genuinely powerful tool: a 2-million-token context window means you could drop an entire novel, a year of emails, or a massive codebase into a single prompt and reason across all of it. That's not a benchmark number — that's a workflow transformation. But you can't use a model that isn't out yet.
Meanwhile, the broader competitive picture is shifting in ways that matter to anyone who uses these tools daily.
ChatGPT's market share fell below 50% for the first time — to 46.4% — while Gemini rose to 27.7%.
And on the geopolitical front,
China unveiled a $295 billion, five-year government AI infrastructure plan, representing roughly $59 billion annually in state-directed AI spend.
The talent exodus from Google may not doom Gemini —
the company's $4.46 trillion market cap gives it structural resilience
— but it is a real signal about where the most ambitious researchers think the frontier is being built right now. That's worth paying attention to, even if you never look at a benchmark in your life.
---
**Sources:**
- https://www.buildfastwithai.com/blogs/ai-news-today-june-24-2026
- https://www.buildfastwithai.com/blogs/ai-news-today-june-22-2026
- https://www.buildfastwithai.com/blogs/ai-news-today-june-23-2026
- https://www.infotech.com/software-reviews/vendor-technology-notes/big-5-ai-vendor-roundup-week-of-june-15-2026
- https://www.crescendo.ai/news/latest-ai-news-and-updates