Back to feed
Ars Technica· Tech· Mon, 08 Jun 2026 19:00:30 Heat 5

Gemini 3.5 and Antigravity come to Google NotebookLM

NotebookLM is getting a big upgrade, but it's only for AI Ultra and enterprise accounts right now.

Read at Ars Technica

Hidden Truths · AI Analysis

Mainstream Narrative

Google is rolling out Gemini 3.5 AI model and a feature called "Antigravity" to its NotebookLM product, but access is limited to premium AI Ultra subscribers and enterprise customers rather than free users.

Missing Context

NotebookLM launched in 2023 as Google's AI-powered research and note-taking tool that can synthesize information from uploaded documents. The "tiered rollout" pattern reflects Google's broader strategy shift toward monetizing AI products after burning billions on development. Gemini 3.5 represents an incremental model update in Google's competitive race against OpenAI's GPT-4 and Anthropic's Claude. The term "Antigravity" is not explained in the summary—context about what this feature actually does is conspicuously absent, which is critical for evaluating the announcement's significance.

Bias Analysis

Ars Technica typically maintains tech-enthusiast/early-adopter framing with mild pro-innovation bias. The headline's breathless naming of "Antigravity" (likely a marketing term) without explanation suggests acceptance of corporate branding at face value. The phrase "getting a big upgrade" implies positive framing without critical assessment of whether these changes materially improve user experience or simply gate-keep features behind paywalls. No apparent political slant, but mild deference to Big Tech product narratives.

Counter-Narratives

**Consumer rights advocates** would frame this as another example of "enshittification"—rolling out free products to build user dependency, then restricting features to premium tiers. **Privacy critics** might question what data collection practices support these AI models and whether enterprise customers receive different privacy protections. **Open-source AI proponents** would argue this demonstrates why proprietary AI development concentrates power in corporate hands rather than democratizing access to tools.

Alternative Angles (Speculative)

Some tech critics speculate that major AI companies are creating artificial scarcity around model access to justify subscription pricing while actual computational costs have decreased. Fringe commentators suggest proprietary AI development allows corporations to control information synthesis and potentially shape narrative formation at scale. **These remain unproven concerns** but reflect anxieties about centralized AI control.

Fact-Check Flags

**What does "Antigravity" actually do?** The feature name reveals nothing about functionality—readers need concrete specifications
**Pricing structure**: What's the cost difference between free/Ultra/enterprise tiers?
**Model performance claims**: Are there independent benchmarks comparing Gemini 3.5 to previous versions and competitors?
**"Only for" limitation**: Is this permanent tiering or a staged rollout plan?

What To Read Next

**Google's official NotebookLM documentation** to understand actual feature specifications beyond marketing language
**Independent AI model benchmarking sites** (like Artificial Analysis or Chatbot Arena) for performance comparisons
**Tech policy analysis** from outlets like The Markup or EFF regarding AI product pricing structures and their implications for equitable access to emerging tools
⚠ Alternative angles are speculative · Always verify with primary sources

Made with Emergent