Sundar Pichai announced at Google I/O that Gemini 3.5 Pro will launch next month; attendees groaned at the model coming out later than they expected
Charles Rollet /Business Insider:
Context & Ripple Effects
Google had signaled before I/O that a new Gemini model was imminent and positioned it near the frontier, while earlier coverage framed Gemini 3 as a broad upgrade in coding and multimodal generation. The I/O announcement therefore set a near-term delivery expectation for Gemini 3.5 Pro.
Subsequent related coverage says Google fell months behind schedule while trying to improve the model, particularly for coding. That turns the audience reaction from a product-launch disappointment into an early indication of pressure around execution and capability readiness.
First-order effects
- Google must manage the gap between the launch timing Pichai set at I/O and the later reported delay, especially among developers and other users awaiting the Pro model.
- The reported effort to strengthen coding capabilities shifts Gemini 3.5 Pro’s release decision toward quality and reliability rather than simply meeting the announced timetable.
Second-order effects
- A delayed flagship release can give rival frontier models more time to establish developer usage in coding workflows, where switching costs can rise as tools and processes become embedded.
- Google’s product teams may face pressure to clarify the distinction between Gemini versions and to prioritize interim improvements to existing Gemini offerings while 3.5 Pro is unfinished.
Third-order effects
- If repeated, slippage between public model roadmaps and delivery would make frontier-AI competition less about announcement cadence and more about whether labs can reliably turn research gains into deployable products.
- The pattern also points to coding as a central proving ground for model releases: capability claims increasingly matter only when they are robust enough for sustained developer use.
The trend: Frontier-model competition is shifting from headline launch promises toward execution in high-value workflows, particularly coding, where delayed capability gains can reshape adoption.