Microsoft-OpenAI Tensions, Infrastructure Investments, and Regulatory Shifts
While others read about AI news, you'll know exactly what to do about it
Today’s Briefing:
In today's newsletter:
Microsoft explores alternatives to OpenAI for Copilot services
Broadcom reports 25% revenue growth driven by AI chip demand
Trump administration establishes Strategic Bitcoin Reserve
Mistral launches new PDF-to-Markdown OCR API
AI search engines reduce news site traffic by 96%
Europe launches €240M RISC-V chip project for computing sovereignty
Understanding GitHub Copilot: Part 3 comprehensive guide
Sponsored by*
It's been said that investing legend, Warren Buffet's, #1 rule is: “Don’t lose money.”
And if you don’t want to lose money, investing in private credit might just be able to help, especially in a volatile market.
Here's why…
Heron Finance’s private credit portfolio has had 0 down months since the start of 2023.* Not many assets can say that.
In fact, since January 2023, the largest high-yield bond and real estate ETFs all suffered from five or more down months. Although past performance does not guarantee future returns, for investors looking for stable passive income, Heron’s Personalized Private Credit portfolio delivers consistent returns.
How does Heron do it?
Heron builds you a personalized portfolio with exposure to the world’s largest private credit funds from Ares, Apollo, and more – targeting up to 11% yield after fees.*
And when it comes to fees, Heron gives investors access to the lowest share class for all of the funds on its platform, which in most cases saves you 0.85% annually compared to directly investing in funds.
Would you like to sponsor this newsletter? View our sponsorship options
Markets
Big Tech Developments
Microsoft explores alternatives to OpenAI for Copilot - Microsoft considering replacing OpenAI models in Copilot, developing in-house alternatives for cost control and independence. Read more
Broadcom reports strong Q1 results driven by AI chip demand - Q1 revenue up 25% YoY to $14.92B; net income up 315% to $5.5B; stock jumps 10%+ after hours. Read more
CoreWeave denies Microsoft contract cancellation rumors - AI infrastructure provider refutes Financial Times reports about Microsoft withdrawing from agreements ahead of CoreWeave's IPO. Read more
Funding & Investment
Hedge Fund Triata gains 39% on DeepSeek AI excitement - Fund's strategic investments in AI software and data center stocks pay off with substantial returns. Read more
Shield AI valued at $5.3B in $240M funding round - Defense tech startup raises Series F-1 funding to enhance autonomous defense systems and expand software portfolio. Read more
Abu Dhabi weighs revival of AIQ share sale - State oil firm and Presight AI considering listing their $1.4B AI joint venture amid growing market interest. Read more
Government & Policy
Trump administration establishes Strategic Bitcoin Reserve - Executive order creates two national cryptocurrency reserves using seized assets without taxpayer costs. Read more
US plans to stem coal-plant closures for AI power needs - Market-based plan aims to prevent closures in anticipation of increased electricity demand from AI. Read more
Minnesota bill targets AI-generated deepfakes - Legislation aims to combat deepfake pornography by targeting "nudification" websites and apps. Read more
Industry Developments
Mistral launches new PDF-to-Markdown OCR API - New multimodal API efficiently converts complex PDF documents into Markdown files, outperforming competitors. Read more
AI search engines reduce news site traffic by 96% - TollBit report shows AI-powered search engines dramatically reducing referral traffic to publishers. Read more
Europe launches €240M RISC-V chip project - 38 European tech companies launch DARE project for processor units to reduce US technology dependence. Read more
ChatGPT expands direct code editing capabilities - macOS update allows direct code editing in Xcode and VS Code for Plus, Pro, and Team users. Read more
AI weather model revolutionizes energy trading - New model delivers more accurate forecasts helping traders make faster decisions in climate-affected markets. Read more
Ethics & Labor
Scale AI under labor investigation - US Department of Labor investigating potential Fair Labor Standards Act violations regarding data labeling contractors. Read more
Online harassment affects 25% of American women - Research shows significant online abuse including AI-generated deepfakes, with young and minority women most affected. Read more
Christie's AI art auction sparks artist backlash - First AI-specialized auction criticized by artists concerned about copyright infringement despite commercial success. Read more
Expert Insights
Thomas Wolf questions AI's knowledge generation - Hugging Face CSO critiques AI's inability to generate new knowledge and need for systems questioning training data. Read more
Analysis
Today's news reveals several significant shifts transforming the AI landscape:
Evolving Corporate Partnerships: Microsoft's exploration of alternatives to OpenAI for Copilot signals a potential realignment in AI development partnerships. This move, coupled with CoreWeave's need to publicly refute contract cancellation rumors, demonstrates how strategic relationships in the AI sector remain fluid despite significant investments. Major players are reassessing the balance between partnership dependencies and in-house capabilities as AI becomes increasingly central to product strategies.
Infrastructure Growth Despite Scrutiny: The continued strong performance in AI hardware (Broadcom's impressive earnings) and substantial funding rounds (Shield AI's $5.3B valuation) show robust investment momentum. However, Scale AI's labor investigation highlights growing regulatory attention to the often-overlooked human components of AI development pipelines. The industry must navigate these labor and ethical considerations as it scales.
Government Actions Shaping Markets: Today's policy developments reflect a complex balancing act by the Trump administration – embracing digital asset innovation through the Strategic Bitcoin Reserve while simultaneously protecting traditional energy infrastructure for AI power needs. This dual approach, alongside state-level initiatives targeting AI-generated deepfakes, illustrates how government actions are directly shaping market trajectories.
Content Economics Transformation: The startling 96% reduction in news site traffic from AI search engines represents perhaps the clearest evidence yet of how AI is fundamentally reshaping information economics. This disruption, combined with controversies in art markets as exemplified by the Christie's AI auction backlash, signals the beginning of significant business model evolution across creative industries.
Tool Optimization vs. Foundational Innovation: The contrast between practical tool advancements (Mistral's PDF API, ChatGPT's code editing) and Thomas Wolf's critique of AI's knowledge generation limitations highlights a tension between incremental product improvements and more fundamental AI capabilities. As the industry matures, balancing these dual innovation paths will become increasingly important for sustainable progress.
Understanding GitHub Copilot: A Comprehensive Guide Part 3
Part 3: The Future of AI-Powered Development
Welcome to the final part of our GitHub Copilot guide! After exploring how to get started in Part 1 and diving into advanced strategies for teams in Part 2, we're now looking ahead at how Copilot is reshaping software development and what this means for your career.
Recent Innovations and Impact
By March 2025, GitHub Copilot has introduced game-changing features that elevate it from a code suggestion tool to a true development partner:
Agent Mode (now in preview) represents a significant leap forward, allowing Copilot to autonomously execute commands and iteratively improve code. Imagine asking it to "debug this API call and implement retry logic" — it will write the code, test it, identify failure points, and refine until it works. This feature alone has reduced debugging time by up to 60% in early testing.
Code Review capabilities (also in preview) now provide detailed feedback on pull requests, catching potential bugs and suggesting test cases. It's like having a virtual senior developer reviewing your code 24/7, significantly improving code quality before human review.
The integration of more powerful models like Claude 3.7 and GPT-4o has substantially improved Copilot's ability to understand complex requirements and generate appropriate solutions, even for newer frameworks and libraries.
A real-world example: I recently built a React dashboard with Copilot handling the data fetching layer. After describing the API endpoints and caching requirements, Copilot generated a robust implementation with fetch, error handling, retry logic, and a Map-based cache system—about 90% of what we needed in seconds. We delivered the project 40% faster than estimated, with fewer bugs.
However, it's worth noting a recent privacy concern: In February 2025, researchers discovered that Copilot occasionally exposed private repository code in suggestions, affecting over 16,000 organizations. GitHub quickly patched the issue, but it highlights the ongoing tension between functionality and privacy in AI tools.
The Collaborative Future
The most productive relationship with AI coding assistants isn't about replacement—it's about amplification. As Fei-Fei Li recently observed, "Artificial intelligence is not a substitute for human intelligence; it is a tool to amplify human creativity and ingenuity."
Copilot excels at handling the mechanical aspects of coding—the implementation details, boilerplate, and repetitive structures—freeing developers to focus on system architecture, user experience, and problem definition. The cognitive load reduction is substantial: developers report spending up to 30% more time on creative problem-solving and significantly less time consulting documentation.
What's Next for GitHub Copilot?
Looking ahead, we can expect several important developments:
Deeper ecosystem integration - Tighter connections with GitHub Actions, allowing Copilot to automatically generate tests and deployment workflows based on your code.
Personalized learning - Copilot will increasingly adapt to your personal coding style and patterns, making its suggestions feel more like they came from you.
Cross-repository understanding - The ability to analyze and suggest code based on patterns across multiple repositories within your organization.
Enhanced security measures - In response to recent privacy concerns, GitHub is implementing stronger controls around data usage and suggestion sourcing.
Staying Ahead of the Curve
To maximize your effectiveness with current and future Copilot features:
Embrace iterative prototyping - Use Agent Mode to quickly generate and test multiple approaches to a problem, then refine the most promising one.
Focus your expertise on what AI can't do - Invest in skills like system architecture, performance optimization, and ethical considerations where human judgment remains essential.
Experiment with emerging frameworks - Copilot can help you quickly learn and implement new technologies, giving you a significant advantage as the landscape evolves.
Final Thoughts
GitHub Copilot isn't just changing how we write code—it's transforming the entire development process. Teams using Copilot effectively report 32% faster onboarding for new developers and up to 55% reduction in time spent on repetitive tasks, as we noted in Part 2.
As AI coding assistants continue to evolve, the most successful developers won't be those who resist change, but those who skillfully collaborate with these tools—using them to amplify their creativity, accelerate implementation, and deliver more value.
Dive in, experiment boldly, and let GitHub Copilot help you build the future.
Featured Prompt
GitHub Copilot Future-Proofing Prompt
I'm a [your role] working on [project type] with [frameworks/languages]. To prepare for emerging AI-assisted development, please help me:
1. Create a structured approach for leveraging Copilot's Agent Mode to automate complex tasks in my workflow
2. Design advanced comment patterns that will guide Copilot to generate more sophisticated solutions for [specific challenge areas]
3. Develop a strategy for balancing Copilot assistance with skill development to ensure I'm growing technically while using AI
4. Build a template for documenting where and how Copilot was used in projects to maintain transparency and facilitate knowledge sharing
5.Suggest methods to evaluate and measure Copilot's impact on my productivity and code quality over timeAI Tools Roundup
As AI coding assistants become essential for modern development, the ecosystem is rapidly evolving beyond GitHub Copilot's initial breakthrough. Here are nine powerful tools shaping software creation:
Amazon Q Developer: Helps with code generation, debugging, and software transformation, integrated into IDEs like VS Code and JetBrains.
Gemini Code Assist: Offers real-time coding assistance, code completions, and explanations within IDEs, powered by Google's AI.
Cline: Autonomous AI coding agent for your IDE that creates/edits files, executes commands, and debugs applications with your permission at each step.
GitLab Duo: Provides AI-driven features for code understanding, generation, and management within the GitLab platform.
JetBrains AI Assistant: Integrates AI into JetBrains IDEs for code generation, completion, and chat, enhancing developer workflows.
CodeRabbit: Automates code reviews with context-aware feedback, reducing manual review time and improving code quality.
Qodo Merge: Assists with pull request reviews, offering automated feedback and suggestions for better code quality.
GitHub Copilot Workspace: An AI-native environment for starting projects, planning, and building code using natural language.
GitHub Spark: Enables creating and sharing micro web apps using natural language, making app development accessible to non-coders and developers.
Get Involved
Submit news or Share insights: hello@futureweb.vc
Sponsor
Would you like to sponsor this newsletter? View our sponsorship options
*Disclosure: This is a paid advertisement. The opinions expressed in this advertisement are strictly those of Heron Finance. The information in this advertisement does not constitute an offer to sell securities or a solicitation of an offer to buy securities. Further, none of the information contained in this advertisement is a recommendation to invest in any securities. Please note there are no material conflicts of interest related to this advertisement. 12% based on the Aggressive portfolio. Any investment target interest rate presented here is intended for informational purposes only and does not guarantee future performance results. This model assumes no variability, including no loan defaults, no fluctuation in interest rate, no customer withdrawal requests, no late payments, and unchanged management fees throughout this projection. Please be aware that all investments involve inherent risks. Customers are advised to consult their own legal and tax advisers before investing. Returns are not a guarantee of future results. Please consider all risk factors before investing.