Debating The Open Internet: Cloudflare vs. Perplexity
Research Rundown #148, plus new memos on Bubble, Filevine, and more.
Research Rundown
A few weeks ago, we published a piece called An Internet Built For AI. In it, we unpacked the ongoing evolution of the internet from an open knowledge commons to a privatized, AI-intermediated information ecosystem. At the end of the piece, we came to this conclusion:
“[The open web will become] a training ground for AI companies, increasingly synthetic and AI-generated, less economically viable for human creators, and accessible only through AI intermediaries for most users. The open web is becoming a raw material source, driving a shift away from foundational values of openness and accessibility and toward closed systems, paywalls, and machine-to-machine content loops. The speed and scale of this shift mean the next few years will determine whether the internet preserves its role as an open platform for independent creators, diverse voices, and knowledge-sharing or becomes primarily infrastructure for machine learning.”
What was, just a month ago, a theoretical overview of how the web is evolving has now broken out into a full-on war of words just over the last week or so. In our latest deep dive, we drill into how Cloudflare and Perplexity have broken out as two contenders in the debate for what the internet should look like in the age of AI.
On the one hand, Cloudflare is an OG defender of the open web. It believes that every site should have the tools necessary to determine who is allowed to use its content and when.
On the other hand, Perplexity is building its entire business around AI access and summarization. It doesn’t want publishers to stand in the way of it serving content to Perplexity users.
The outcome of this debate may very well break the social contract that has dictated the business model of the internet for decades.
Read the full deep dive here.
Inside Bubble, the no-code engine accelerating product launches across industries. To find out more, read our memo here.
How Filevine is automating the legal world—without lawyers noticing. Read our memo here for everything you need to know.
What makes Redpanda the go-to choice for low-latency, high-throughput event processing? We break it down in our memo here.
Vercel is in early talks with investors about a funding round that could value the nine-year-old cloud hosting and AI tools provider at $8–9 billion, nearly triple its May 2024 valuation, after doubling ARR to $200 million with 76% gross margins on core services and launching its V0 coding assistant.
Cognition has raised nearly $500 million led by Founders Fund at a $9.8 billion valuation, more than double earlier this year, to scale its Devin AI coding assistant, following its Windsurf acquisition, as competition in enterprise code-generation heats up against rivals like Microsoft, Anysphere, Anthropic, and OpenAI.
Apple is planning a major AI hardware push, including a 2027 tabletop robot with lifelike Siri, a 2026 smart display, and new home-security cameras, as part of a broader strategy to revive its innovation reputation and compete in smart-home and AI devices against Amazon, Google, Meta, and Samsung.
Researchers at Google DeepMind introduced PH-LLM, a Gemini-based model fine-tuned for sleep and fitness coaching using wearable data, which outperformed human experts on domain exams, matched or exceeded them on personalized case study recommendations, and accurately predicted self-reported sleep quality via multimodal sensor integration.
Sam Altman is reportedly co-founding Merge Labs, a brain–computer interface startup valued at about $850 million that could receive significant funding from OpenAI’s ventures arm, positioning it to compete with Elon Musk’s Neuralink, which is testing implanted chips to help paralyzed patients control devices.
General Catalyst CEO Hemant Taneja outlined the firm’s shift into a “strategic conglomerate” combining venture capital, industry partnerships, and incubations to transform sectors, starting with healthcare. Initiatives include acquiring a hospital system via Hatco, pursuing AI-driven rollups in labor-arbitrage industries, and expanding seed investing through acquisitions in Europe and India.
In Breakneck: China’s Quest to Engineer the Future, analyst Dan Wang argues the US is overrun by lawyers while China is led by engineers, urging America to cut rulemaking red tape and revive large-scale building to compete in fields like infrastructure, manufacturing, and advanced energy, while warning that current policies risk widening strategic gaps.
In a recent essay, Superhuman co-founder, Conrad Irwin, argues that while LLMs can generate and update code, they can’t truly build software because they lack the ability to maintain clear mental models, leading to confusion, recency bias, hallucinations, and poor handling of omitted context in complex projects.
In another essay, Chris Paik evaluates Cursor’s “unlimited” subscription model and warns that without pricing proportional to variable costs tied to frontier AI models from OpenAI and Anthropic, the company risks negative gross margins, user churn from price hikes and caps, and an unproven product-market fit disguised by subsidized usage.
At Contrary Research, we’ve built the best starting place to understand private tech companies. If you're interested in researching and writing about tech, apply here for our Research Fellowship!