AI Week in Review: Brains, Browsers, and Billions
I dig through hundreds of articles (and my agents crawl through thousands more) to bring you the most important stories shaping the future of AI.
381,172 AI headlines dropped this week. Here's three shifts that were impossible to avoid:
OpenAI is backing Merge Labs, a Neuralink rival—an announcement that landed the same week Altman sparred with Musk online. The contest is moving beyond chatbots into wiring AI directly to human thought.
Perplexity’s $34B bid for Chrome signals the browser—the gateway to billions of daily interactions—is now the most valuable distribution channel in AI, underscoring why they’ve been tirelessly building Comet.
Ex-OpenAI researcher Leopold Aschenbrenner raised $1.5B for his hedge fund, showing how proximity to AI breakthroughs can outmaneuver Wall Street itself.
The takeaway: the AI race is no longer about one model—it’s about control over minds, distribution, and capital. Let’s dive in...
SAM ALTMAN & OPENAI 🧠 OpenAI to back Neuralink competitor
News: OpenAI is reportedly in talks to fund Merge Labs, a brain-computer interface startup valued at $850M, co-founded by Sam Altman, aiming to compete with Musk’s Neuralink.
Details:
Alex Blania, head of Altman’s iris-scanning Worldcoin, will oversee the project; Altman is co-founder but not operational.
This marks OpenAI’s first major investment in brain-computer interfaces.
Musk predicts Neuralink will serve 20,000 people annually by 2031, hitting $1B revenue.
Altman previously discussed such tech in his 2017 blog “The Merge”.
Why it matters: With the Musk-Altman feud dominating headlines, news of OpenAI funding a Neuralink rival is sure to add fuel to the fire. But focusing only on ChatGPT misses the real story—Altman is quietly assembling an expansive, highly competitive portfolio that spans AI, biotech, sustainable energy, and fintech. He’s co-founded and invested in ventures like Worldcoin, which has raised over $135M in token sales from backers such as a16z and Bain Capital Crypto to bring biometric crypto identity to millions. His track record includes significant stakes in companies like Helion Energy, Retro Biosciences, Coalition, and Reddit—making OpenAI just the tip of an iceberg of ecosystem leverage designed to keep him at the top of frontier tech for years to come.
PERPLEXITY 💰 Perplexity’s $34B bid for Chrome
News: Perplexity reportedly made an unsolicited $34.5B offer to buy Google’s Chrome browser, amid antitrust pressure that could force a divestiture.
Details:
Pitch made directly to Alphabet CEO Sundar Pichai
Bid nearly doubles Perplexity’s $18B valuation
Chrome has 3.5B users and 60% global market share
Judge Amit Mehta to decide on forced sale this month
Why it matters: The $34.5B offer is a bold strategic move—unsolicited and non-binding—meant as much to elevate Perplexity’s profile as to position it for a potential antitrust-driven Chrome divestiture. With its AI-first browser Comet gaining traction, Perplexity is signaling readiness to invest heavily in Chrome’s future while betting on the strategic importance of browser distribution in AI-powered search competition.
Ex-OpenAI researcher raises $1.5B for AI hedge fund
News: Former OpenAI researcher Leopold Aschenbrenner has reportedly secured over $1.5B in funding for his ‘Situational Awareness’ AI-focused hedge fund — despite having no formal background in professional investing.
Details:
Aschenbrenner was part of OpenAI’s Superalignment team and was one of two employees dismissed in April 2024 after being accused of leaking sensitive information, a claim he has publicly addressed.
He later authored a viral essay titled ‘Situational Awareness’ (the fund’s namesake) outlining his predictions for AGI and the trajectory of AI development.
In the first half of 2025, Aschenbrenner’s fund delivered a 47% return, outperforming the S&P 500’s 6% gain, despite his lack of prior investing experience.
The fund concentrates on AI-adjacent sectors, such as semiconductors, infrastructure, and power companies poised to benefit from the AI boom, along with selective short positions in industries likely to be disrupted.
Why it matters: The AI surge is transforming the hedge fund space, giving those closest to the technology a competitive edge over traditional finance veterans. Many want exposure to the AI gold rush, but few possess the insight to anticipate where the industry is truly headed.
COHERE 💰 Cohere’s $500M funding round boosts valuation to $6.8B
News: Cohere, a Canadian AI foundation model provider, has raised $500 million in an oversubscribed round co-led by Radical Ventures and Inovia Capital, with backing from AMD Ventures, Nvidia, PSP Investments, Salesforce Ventures, and others. The investment lifts its valuation to $6.8 billion, up 24% from last year’s $5.5 billion.
Details:
Funds will accelerate Cohere’s expansion and development of secure agentic AI solutions for enterprise and government clients.
Focus areas include data sovereignty , regulatory compliance, and automating routine workflows.
Joelle Pineau, former head of Meta’s FAIR team and McGill professor, joins as Chief AI Officer to lead research and product development.
Former Uber exec Francois Chadwick joins as Chief Financial Officer.
Cohere specializes in custom large language models for enterprises, targeting highly regulated sectors like finance and healthcare.
Why it matters: The enterprise AI segment is a relatively small niche in the broader AI market, but by prioritizing security, privacy, and on-premises deployments, Cohere has carved out a leadership position. In fact, their focus and execution arguably make them a more prominent and specialized enterprise AI partner than OpenAI or Anthropic, especially for regulated industries with strict compliance needs.
GOOGLE 👏 Google’s new Gemma model is smaller than ever
News: Google launched Gemma 3 270M, the smallest yet in its open-source model family—capable of running on smartphones, browsers, and other consumer devices with surprising efficiency and designed for advanced on-device AI applications.
Details:
Ultra-compact: At 270 million parameters, Gemma 3 270M beats similarly small AI systems at instruction-following despite being far smaller than most models.
Energy efficient: In tests, it handled 25 conversations on a Pixel 9 Pro while using less than 1% battery.
Fast customization: Developers can fine-tune it in minutes for specific tasks; Google demoed a Bedtime Story Generator running fully offline.
Why it matters: The big story here is that smaller models like Gemma 3 270M are now delivering outsized performance. This matters because the future of AI for most people will be driven by compact, efficient systems that can run directly on personal devices, revolutionizing on-device AI. As Liquid AI’s LFM2 also advances, this paves the way for what Mark Zuckerberg recently described as “personal superintelligence”—where general users who don’t need the absolute highest performing models still gain powerful, private, and always-available AI in their pocket. This shift could mark a seismic change for the general public.
Thanks for reading this far! Stay ahead of the curve with my daily AI newsletter—bringing you the latest in AI news, innovation, and leadership every single day, 365 days a year. See you tomorrow for more!