Google Tests Ads in AI Chatbots, Apple Partners with Anthropic
It's Wednesday, May 6, 2025, and you're reading the Agentive Daily Report.
Busy People's Section
Google Begins Testing Ads in AI Chatbots
Google has quietly begun testing ads in third-party AI chatbots like iAsk using its AdSense platform. This marks the first significant step toward monetising conversational AI interfaces, providing a potential revenue stream for developers beyond subscription models.
The move signals that ad-based business models are likely to become a dominant force in the AI economy, just as they did in early web and social media. For users, this represents the beginning of a shift from purely utilitarian AI interactions to commercial ones, potentially altering how people use and trust these systems. The development is especially significant for smaller AI chatbot developers who need sustainable business models but lack the resources to build and train their own models.
Apple and Anthropic Partner on "Vibe-Coding" Platform
Apple has teamed up with Anthropic to create an AI-powered coding platform that will integrate Claude Sonnet into Xcode for generating, editing, and testing code. The system, which is focused on what's being called "vibe-coding," represents Apple's most significant AI development collaboration to date.
This partnership highlights Apple's strategic approach to AI—partnering with specialists rather than building everything in-house. While initially intended for internal use, this collaboration could eventually reshape how developers interact with Apple's development environment. It also indicates Anthropic's growing enterprise strategy, positioning Claude as a premium AI model for high-value business applications rather than competing directly with ChatGPT in the consumer space.
Claude's Research Mode Gets Major Time Extension
Anthropic has significantly upgraded Claude's research capabilities, tripling its investigation time limit to 45 minutes before delivering comprehensive reports with full citations. The new system can analyse complex queries, search hundreds of sources, and compile document-style responses with embedded references.
This enhancement positions Claude as a powerful tool for knowledge work, particularly in academic, legal, and business contexts where thorough research and proper attribution are essential. By focusing on source quality and proper attribution, Anthropic appears to be differentiating Claude with a "trust-first" approach in contrast to competitors. The upgrade, along with their new Integrations feature for connecting third-party services, indicates Anthropic's push to make Claude a central workspace tool rather than just a conversational assistant.
Fast Forward
- Midjourney's Omni-Reference Tool: Midjourney released a new experimental feature called "Omni-Reference" that replaces the older Character Reference tool, giving users significantly more control over visual details in AI-generated images through a slider system that can be fine-tuned from subtle inspiration to precise detail.
- DeepMind on Context Windows: DeepMind's Nikolay Savinov explained that despite the trend toward million-token context windows, focusing on relevant content yields better results as models must divide attention across all tokens, potentially diminishing focus on important details. This research challenges assumptions about simply adding more context.
- Space-Based Data Centres: Eric Schmidt reportedly acquired Relativity Space to enable launching significant computing infrastructure into space, powered by solar panels and using the vacuum of space for cooling. This represents a novel approach to the growing energy demands of AI infrastructure.
- Google's Antitrust Case: Google's antitrust case has evolved from focusing on search monopoly to centering on AI market control, highlighting regulatory concern about dominant players leveraging existing advantages to control emerging technologies.
- Duolingo's AI Transformation: Duolingo's implementation of AI has reportedly led to the quiet replacement of up to 100 writers and translators, providing a concrete example of how AI is reshaping knowledge work and raising concerns about entry-level job opportunities.
New Tools Discovered
- Dash: A workflow automation tool that allows users to control their entire workflow through prompts, eliminating the need to switch between applications.
- WebThinker: A deep research framework powered by large reasoning models that enables autonomous web searches and exploration for comprehensive research reports.
- Daytona: A secure infrastructure platform specifically designed for running AI-generated code safely, addressing a key concern for developers working with AI coding assistants.
- mrge: A code review tool built specifically for the AI development era, streamlining collaboration on AI-generated or enhanced code.
- Credal: A platform that deploys secure AI agents that can collaborate, automate workflows, and act across enterprise tools while maintaining proper data permissions.
Discover more tools at Agentive.Directory
That's a wrap for today! Thank you for reading this report.
Have thoughts on today's edition? Hit reply and let us know what you're thinking. Or if you've discovered a cool AI tool we should feature, drop us a line.
Until tomorrow,
Hak from Agentive.Studio
