Tech Signals: Apple picks Google's Gemini to power Siri
Apple's decision to integrate Google's Gemini into Siri marks a significant shift in AI strategy, potentially enhancing voice assistant capabilities. This collaboration between tech giants could redefine user expectations, with 66.5% approval from nine analyzed signals.
#1 - Top Signal
Apple and Google signed a multiyear partnership to use Google’s Gemini models and cloud technology to power an AI-upgraded Siri expected later this year. Apple says Google provides the “most capable foundation” for Apple Foundation Models, while models will still run on-device and via Apple’s Private Cloud Compute. The deal follows reports Apple may pay ~$1B/year for Google AI and comes as Google briefly topped $4T market cap and surpassed Apple in market value for the first time since 2019. This shifts the near-term Siri “LLM layer” toward a multi-provider architecture (Gemini + existing ChatGPT fallback), creating immediate demand for privacy, evaluation, and reliability tooling around voice-agent workflows.
Key Facts:
- Apple is teaming up with Google to use Gemini models for an AI-powered Siri upgrade expected later this year.
- The partnership is described as multiyear and will lean on Google’s Gemini and cloud technology for future Apple foundational models.
- Apple states the models will continue to run on Apple devices and Apple’s Private Cloud Compute.
- Apple declined to comment on deal terms; Google referred to a joint statement.
Also Noteworthy Today
The collaboration between Apple and Google on integrating Gemini into Siri highlights the increasing trend of major tech companies leveraging AI to enhance their services. This trend is further exemplified by the introduction of Claude Code by Cowork, designed to optimize workplace tasks, and the innovative tools like NanmiCoder and MediaCrawler, which are reshaping content creation and media analysis.
Cowork: Claude Code for the rest of your work
Hacker News · Read Original
Anthropic launched “Cowork” (research preview) to bring Claude Code-style agentic workflows to non-developers by granting Claude scoped access to a user-selected local folder on macOS. Cowork can read/edit/create files, make a plan, execute tasks with higher autonomy than chat, and queue tasks in parallel while asking before “significant actions.” Early community reaction is dominated by safety concerns: irreversible filesystem actions, lack of rollback/snapshots, and prompt-injection/exfiltration risk when an agent can browse and touch local data. This creates a near-term product gap for “safe agentic workspaces” (sandboxing, versioned operations, policy controls, and auditable action logs) that can sit between LLM agents and real user files.
Key Facts:
- Cowork is available as a research preview for Claude Max subscribers in the macOS Claude app (announced Jan 12, 2026).
- Cowork differs from chat by granting Claude explicit access to a chosen folder; Claude can read, edit, and create files within that folder.
NanmiCoder / MediaCrawler
Github Trending · Read Original
[readme] NanmiCoder/MediaCrawler is a multi-platform “self-media” public data collection tool covering major Chinese platforms (Xiaohongshu, Douyin, Kuaishou, Bilibili, Weibo, Tieba, Zhihu) with features like keyword search, post-ID crawling, creator homepage crawling, and comment wordcloud generation. [readme] It uses Playwright to preserve login sessions and avoids heavy JS reverse-engineering by extracting signature parameters via JS expressions in an authenticated browser context. Recent GitHub issues show both security maintenance pressure (a “CRITICAL” CVE fix request) and reliability/compatibility pain points (dependency import errors, partial crawl limits). The repo’s traction on GitHub Trending plus recurring operational issues suggests an opportunity for a more reliable, compliance-aware, “crawler-as-a-product” stack (observability, resumable jobs, account/proxy orchestration, and safer dependency management).
Key Facts:
- Signal source is github_trending for NanmiCoder/MediaCrawler.
- [readme] MediaCrawler positions itself as a multi-platform self-media data collection tool supporting Xiaohongshu, Douyin, Kuaishou, Bilibili, Weibo, Tieba, and Zhihu.
Market Pulse
The recent market movements around Google's valuation highlight a significant moment for tech founders to consider the implications of large-scale cloud computing and AI developments. Google's brief surpassing of a $4 trillion market value underscores investor confidence in its stability and deep resources, despite subsequent share pullbacks. For tech founders, this signals a market environment that rewards robustness and scale, suggesting that stability and financial strength are increasingly valued over sheer innovation. This trend may encourage founders to prioritize building resilient business models that can withstand market fluctuations.
Discussion on platforms like Hacker News reveals a cautious yet pragmatic perspective among tech experts regarding Google's strategic moves. There is considerable attention on privacy and data governance, especially if Google's AI, Gemini, utilizes Apple's Private Cloud Compute. This focus serves as a reminder for founders to ensure their AI products address data privacy concerns, as these factors are becoming central to user trust and adoption. The skepticism around Siri's reputation suggests that founders should also be aware of brand perception and its impact on product success, particularly in AI where user experience and trust are critical.
Attention to safety and data governance, especially concerning prompt injection and exfiltration risks, indicates an evolving landscape where security is paramount. The risk of irreversible local actions without rollback capabilities presents a cautionary tale for founders developing AI agents with browser and file access. This underscores the need for robust safety protocols and rollback mechanisms in product design to prevent destructive actions. Founders should prioritize these aspects to maintain user trust and avoid potential liabilities associated with data breaches or unintended destructive actions.
Despite these concerns, there is optimism about the potential productivity gains from AI, as demonstrated by the positive feedback on agentic debugging and productivity tools. A specific example is the use of Claude Code tooling for rapid debugging, which highlights the practical benefits of AI in enhancing development efficiency. Founders should watch for opportunities to leverage AI for productivity gains, ensuring their products not only mitigate risks but also deliver tangible operational benefits to users. Balancing innovation with safety and privacy will be crucial as AI continues to integrate into more aspects of business operations.
Explore the full intelligence dashboard
Open Intelligence Dashboard