Weekly Wrap Lighthouse AI Newsletter (25th July)
Another busy week in the world of AI. We’ve seen massive funding rounds, major policy shifts, and significant product developments in the AI industry. From a record-breaking $2 billion seed round to new presidential AI policies, the landscape is rapidly evolving with clear winners emerging.
AI News (Macro, Economic & Geopolitical Events)
Google users are less likely to click on links when an AI summary appears in the results
I think this may not come at as a shock to most of us who look at web stats and traffic sources, however, Pew looking in ~70,000 google searches and discovered that users were likely to be satisfied with the AI summary and therefore not visit a website. https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/
U.S. Releases AI Action Plan to Dominate AI Race
Trump’s AI plan expands exports of full-stack AI packages to allies, loosens environmental regulations for data centers, eliminates Biden-era restrictions, removes DEI requirements, and rescinds AI diffusion caps. The strategy prioritizes defeating China through deregulation, expedited infrastructure permits, and unrestricted technology sharing with friendly nations while blocking adversary access. More
Mira Murati’s Startup Raises $2 Billion, Hits $12 B Valuation
Former OpenAI CTO Mira Murati’s stealth AI company Thinking Machines Lab secured a massive $2 billion seed round, valuing the six-month-old startup at around $12 billion. The funding – led by a16z with participation from Nvidia, Accel, and others – comes despite no product launched yet. Murati hinted the first product (a multimodal AI with a major open-source component) will debut in the next few months. https://www.therundown.ai/p/mira-muratis-12b-startup-sets-launch-clock
xAI seeks additional $12 billion in funding through Valor Equity Partners
Following their $10 billion funding round earlier in July, Elon Musk’s xAI is reportedly seeking an additional $12 billion in funding. This would bring their total 2025 funding to $22 billion, representing one of the largest AI funding rounds ever assembled. https://www.theinformation.com
Yahoo Japan mandates company-wide AI adoption, targets 2030 productivity doubling
In one of the most aggressive corporate AI adoption strategies to date, Yahoo Japan mandated daily AI tool usage across all employees with productivity tracking. The company aims to double productivity by 2030, calling AI integration “no longer optional but essential.” Crescendo AI
OpenAI Model Achieves Gold Medal Level at Math Olympiad
An experimental OpenAI large language model reached “gold medal-level” performance on the 2025 International Math Olympiad, a prestigious global math competition. Under official exam conditions (two 4.5-hour tests, no tools or internet), the model solved 5 out of 6 problems, scoring 35/42, a result on par with the top human contestants. OpenAI noted the model, which won’t be publicly released immediately, demonstrates rapid progress in AI’s problem-solving and reasoning abilities, exceeding many experts’ predictions. https://www.joinhorizon.ai/p/openai-wins-gold-at-world-s-top-math-competition
US Pentagon Awards $200 Million in AI Contracts to Top Firms
The U.S. Department of Defense announced up to $200 million in contracts spread across leading AI companies Anthropic, Google, OpenAI, and Elon Musk’s xAI. The initiative aims “to accelerate DoD adoption of advanced AI” for national security uses, including developing autonomous AI workflows for defense missions. Notably, xAI is launching a “Grok for Government” service to deliver frontier AI products to federal clients.
Link: https://www.joinhorizon.ai/p/top-tech-firms-win-200-million-pentagon-ai-contracts
Anthropic release report on how to Build AI in America.
Anthropic released a report via its social channels outlining what it will take to ensure America has the energy and infrastructure it needs to maintain its leadership in AI. The company proposes two strategic pillars that will address infrastructure challenges.
Pillar 1: Building large-scale AI training infrastructure
Pillar 2: Building broad-based infrastructure for AI innovation nationwide https://www-cdn.anthropic.com/0dc382a2086f6a054eeb17e8a531bd9625b8e6e5.pdf
Tech AI News & Developments
ChatGPT-5 is coming in August
I think this may not come at as a shock to most of us who look at web stats and traffic sources, however, Pew looking in ~70,000 google searches and discovered that users were likely to be satisfied with the AI summary and therefore not visit a website. https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/
Google Notebook LM - launches featured Notebooks.
Trump’s AI plan expands exports of full-stack AI packages to allies, loosens environmental regulations for data centers, eliminates Biden-era restrictions, removes DEI requirements, and rescinds AI diffusion caps. The strategy prioritizes defeating China through deregulation, expedited infrastructure permits, and unrestricted technology sharing with friendly nations while blocking adversary access. More
OpenAI's ChatGPT Agent Fully Rolled Out to Subscribers
Former OpenAI CTO Mira Murati’s stealth AI company Thinking Machines Lab secured a massive $2 billion seed round, valuing the six-month-old startup at around $12 billion. The funding – led by a16z with participation from Nvidia, Accel, and others – comes despite no product launched yet. Murati hinted the first product (a multimodal AI with a major open-source component) will debut in the next few months. https://www.therundown.ai/p/mira-muratis-12b-startup-sets-launch-clock
Hume AI Releases EVI-3 for Advanced Voice Mimicry
Following their $10 billion funding round earlier in July, Elon Musk’s xAI is reportedly seeking an additional $12 billion in funding. This would bring their total 2025 funding to $22 billion, representing one of the largest AI funding rounds ever assembled. https://www.theinformation.com
xAI Upgrades Grok with Anime-Inspired 3D AI Companions
In one of the most aggressive corporate AI adoption strategies to date, Yahoo Japan mandated daily AI tool usage across all employees with productivity tracking. The company aims to double productivity by 2030, calling AI integration “no longer optional but essential.” Crescendo AI
Weekly Spotlight - Claude Code
Claude Code is Anthropic’s preview of an agentic command-line tool that changes how developers interact with AI assistance for coding. Rather than copying and pasting between a chat interface and their development environment, developers can now delegate entire coding tasks directly from their terminal.
All AI tools can code, but when you dig into it they dont all code alike, plus they constantly bunny hop each other with releases of their latest model, but if you spend too long in reddit like we do Claude consistently has the support of the developers using it, it has stayed the tool of choice despite credible rivals such as Gemini’s free CLI tool and Grok 4. Then firms like cursor which had an advantage with their integrated environments are losing ground.
Traditional AI coding tools require constant context-switching you describe your problem, receive code suggestions, manually implement them, debug issues, then return for more help. Claude Code eliminates this friction by operating within your development environment, understanding your codebase context, and executing complex multi-step tasks autonomously.
The game-changing aspect lies in its agency. Instead of merely suggesting code snippets, Claude Code can navigate file structures, understand project architecture, write and modify multiple files, run tests, and iterate on solutions. This transforms AI from a sophisticated autocomplete into a genuine coding partner that can handle substantial development tasks independently.
For developers, this means dramatically accelerated workflows. Complex refactoring, boilerplate generation, debugging sessions, and feature implementation that might take hours can potentially be delegated and completed in minutes. By reducing the mechanical aspects of coding, developers can focus on architecture, design decisions, and creative problem-solving—the uniquely human aspects of software development.
Vibe coding is really taking off. If you are ever sat in the office wishing you could do something to free up your time, chances are that with the tools available to you now you can (Superblocks have launched an AI to just this).
Reality check – are you going to produce production code capable of supporting 1 million actions per second? Maybe not this month, certainly not without the professionals, but productivity automations (tools like processing recordings of meetings, creating summaries and actions, emailing the attendees with the notes and actions and sending follow-ups to close-out actions, as well as creating a searchable database of meetings) are just a prompt away.