That Was The Week

This week’s headline: When Your AI Breaks Your Heart.

GPT-5 arrived “better” by every metric—yet users begged for GPT-4o back. It wasn’t about accuracy. It was about personality. People felt like they lost a friend. OpenAI listened, backtracked, and gave them their companion back.

But should it have? Progress is messy, and heartbreak may be the price of change.

The Pain of Change

Users bond with AI like colleagues or partners—and revolt when those bonds are broken.

OpenAI faced its first true PR crisis, forcing it to act like a consumer company, not just a lab.

But longing for “the old AI” is as unrealistic as yearning for Windows 95. Change is the only constant.

The Shifting Web

Cloudflare’s Matthew Prince warns: AI is killing the Web.

Perplexity’s $34.5B bid for Chrome shows the fight for browser control—but the browser itself may be obsolete.

Just as Spotify freed music from CDs, AI is unbundling content from URLs and tabs. The web isn’t dying—it’s being liberated.

Inputs vs. Manipulation

AI’s real weakness? Databases. Models still can’t query live inventory, prices, or transactions.

“SEO for AI” tries to paper over this by gaming prompts—just like spammers gamed Google.

But the future isn’t tricks. It’s context engineering: clean data + authentic inputs.

Winners & Losers

40% of VC money is going to just 10 AI deals. The power law rules: winners take almost everything.

Geoffrey Hinton warns of AI “alien beings,” but others argue that fear distracts from real infrastructure challenges—like power grids, chips, and data quality.

The Real Opportunity

Startup of the Week: Torch, a health AI that turns a decade of medical records into personalized insights.

This is the real future—integrating trustworthy data into AI, not re-skinning old personalities.

The controversy this week is simple:

Do we cling to the familiar—or embrace the heartbreak that comes with progress?

While some mourn GPT-4o, the real story is far bigger: AI is rewriting law, health, energy, and the web itself. And it’s happening whether we’re ready or not.

Contents

Venture Capital

Data is actually not a great VC-backed business

Why Every International Founder Should Spend Three Months in the U.S.

A Path Forward for Seed VCs

Why Seed Rounds Are Growing as Startups Shrink

Why our access is improving

On 9 data-driven tips for YC startup founders

Ultra-Unicorn Investors: These Firms Have Amassed The Largest Portfolios Of $5B+ Startups 

Landscape of VC-Backed M&A

Essays

A Summer of AI in San Francisco

The Satya of Satya’s Layoff Memo

Ads are inevitable in AI, and that's okay

The AI Search Tipping Point

The Slow Death of Social Networks

10 Things I Wish I Knew Before Vibe Coding

PageRank in the Age of AI

European Weakness

Wise shareholders back plan to move listing from the UK to the US

AI

As Anthropic goes, so goes the generative AI trade, says Big Technology's Alex Kantrowitz

a16z GP, Martin Casado: Anthropic vs OpenAI & Why Open Source is a National Security Risk with China

Balaji Srinivasan: How AI Will Change Politics, War, and Money

AI that was inevitable

Iconiq set to lead $5bn funding round for AI start-up Anthropic

The Evolution of AI Agents: Navigating the “Fog of AI” in Rapidly Changing Foundations | Stanislas Polu and Harrison Chase

What will it take for robotaxis to go global?

The AI SDR Reality Check: How To Actually Make It Work

OpenAI’s IMO Team on Why Models Are Finally Solving Elite-Level Math

#259: Why Data Is the Stack

Google Lands $1.2 Billion Cloud Contract From ServiceNow

Pew Study: Google Users Click Less When AI Summaries Appear in Search Results

Loveable and Replit Both Hit $100M ARR in Record Time. The Vibe Coding TAM: How Big Can This Market Really Get?

New AI architecture delivers 100x faster reasoning than LLMs with just 1,000 training examples

Tesla signs a $16.5 billion chip contract with Samsung Electronics

The Making Of Dario Amodei

Chinese Tech

Kimi

China’s AI Gambit: Code as Standards

How Hangzhou Spawned Deepseek and Unitree

Zhipu crushing benchmarks

Xi Jinping is the main thing holding China back

China Prepares to Unseat US in Fight for $4.8 Trillion AI Market

Media

New Media: Podcasts, Politics & the Collapse of Trust

IPO

Figma’s Auction-Like IPO Set Up to Capitalize on Strong Demand

Education

Why You Should Still Study Computer Science

Interview of the Week

"AI Is Too Busy to Take Your Job: The Electrifying Truth about our AIgorithmic Future

Regulation

Boston city council members introduce a bill to require drivers in Waymos

M & A

Palo Alto Networks agrees $25bn takeover of CyberArk

Key Trend 1: Hyper-Accelerated Scaling and New Venture Capital Dynamics

Significance:

AI innovation is driving hypergrowth that shatters traditional timelines. Companies now catapult from zero to hundreds of millions in ARR in months, not years. Venture capital must adapt to this new reality with massively larger and risk-tolerant funding rounds focused on parallel scaling — raising as much capital as revenue grows to capture market share rapidly.

Why it matters:

The "burn rate is a feature, not a bug" mentality defines funding strategies today, making capital intensity a necessity in winner-take-all AI markets. Companies ignoring this shift risk falling behind or being outspent by competitors who build moats with talent, infrastructure, and data first.

Key Trend 2: Talent and Data as Critical Moats in the AI Arms Race

Significance:

Talent wars are escalating, with massive compensation packages used to acquire top AI researchers and engineers, reflecting a strategy to build defensible moats beyond pure technology. Additionally, proprietary data pipelines and reinforcement learning processes are becoming crucial competitive advantages, often trumping model architectures alone.

Why it matters:

As AI models become commoditized and easier to replicate, the real differentiation lies in costly-to-copy human capital and exclusive data ecosystems. Companies investing in these defensive layers will sustain leadership and fend off rapidly emerging competitors.

Key Trend 3: Redefining Content Economics in the AI Era

Significance:

AI’s reliance on vast amounts of web content to train models, often without compensation or permission, is triggering a fundamental rethink of content ownership, access, and monetization. Cloudflare’s new policies signal a shift toward pay-for-access models that require AI companies to compensate content creators, disrupting the previous “free crawl” economic bargain.

Why it matters:

This reshapes incentives for publishers, creators, and AI businesses alike. Content providers gain leverage to set terms and generate revenues from AI models, while AI companies must adapt business models to accommodate these new costs, potentially accelerating AI ad monetization.

Key Trend 4: The Great Differentiation — Building Hard-to-Copy Moats in an AI World

Significance:

As AI makes imitation easy and replicable, companies must differentiate through costly signals, authentic experiences, and unique assets that competitors cannot copy cheaply. This includes physical infrastructure, branding, cultural elements, and deep human expertise — all forming sustainable moats in a landscape of digital abundance.

Why it matters:

In a world where digital replication is trivial, the economic value shifts toward rarity and authenticity. Companies adopting this mindset can build lasting competitive advantages that resist commoditization.

Key Trend 5: The Geopolitical and Regulatory Landscape of AI

Significance:

AI development is not just a technology race but a geopolitical contest, with varied national approaches balancing innovation speed and regulation. Europe’s AI Act exemplifies efforts to govern AI but faces pushback for potentially stifling competitiveness compared to the US and China’s growth-first posture.

Why it matters:

The regulatory environment shapes where and how AI innovation flourishes. Diverging standards and delayed coordination may influence global market leadership, investment flows, and the speed of AI adoption.Discussion Questions

How does the new model of “parallel scaling” of funding and revenue fundamentally change startup growth strategies in AI compared to traditional SaaS? What risks and benefits does this introduce?

With talent and data becoming primary moats, is the AI market at risk of consolidating power among a small set of firms? How can startups compete in such an environment?

Cloudflare’s “Pay Per Crawl” aims to rebalance value between content creators and AI companies. Will this model incentivize innovation or hamper the open data flows AI depends on?

In a world where AI makes copying easy, what are the most viable forms of costly signals for differentiation? Can digital firms realistically replicate physical or cultural moats?

Given the divergent regulatory approaches between the US, Europe, and China, how might geopolitical competition affect the speed and ethics of AI adoption globally?

How do the controversies around tokenization and digital asset legitimacy, like OpenAI’s rejection of Robinhood tokens, reflect broader regulatory challenges for blockchain-based financial innovation?

Is the venture capital industry prepared to adapt investment models to AI’s capital intensiveness and growth patterns? How might smaller VCs or new investors respond to the concentration of “ultra-unicorns”?

Overview

This newsletter issue commemorates 20 years of TechCrunch, reflecting on its landmark influence in shaping the startup ecosystem and tech journalism since its launch in 2005. Beyond nostalgia, the content reveals key ongoing shifts in technology, venture capital, AI innovation, and market dynamics that continue to define the industry’s present and future.

Listeners will gain perspective on how TechCrunch grew from a simple Web 2.0 weblog to a foundational startup network hub, alongside insights into current critical trends such as AI’s evolving role in venture capital and software development, Apple’s design and AI strategy, evolving IPO markets, and debates around AI ethics. The combination of historical context and forward-looking analysis makes this a compelling episode for anyone interested in the tech industry's trajectory.

Key Trend 1: The Enduring Influence and Evolution of TechCrunch as a Startup Network

TechCrunch’s founding vision was not only to report new Web 2.0 companies but to serve as a connective platform for entrepreneurs, investors, and innovators globally.

It emerged as the definitive startup network akin to how Facebook shaped social networks, fundamentally influencing tech culture, funding, and ecosystem formation.

Today, TechCrunch remains a vital resource, expanding its global footprint with strategic partnerships and deeper engagement in key startup hubs like Europe.

Key Trend 2: AI’s Growing Impact on Venture Capital, Software Development, and Industry Structure

AI continues to reshape venture capital with strong focus on B2B operational tooling, platform/API-first startups, and developer-centric innovation.

Large models and AI coding tools (e.g., vibe coding, integration in Xcode) signal a shift towards AI-assisted software creation workflows.

However, challenges remain in reasoning capabilities of AI models, skeptical internal debates on AI safety, and ethical implications within leading tech firms.

Strategic investments and valuation surges of AI companies, such as Anysphere’s rapid growth and Meta’s big bet on Scale AI, highlight intense competition for AI supremacy.

Key Trend 3: The Resurgence of Public Markets and Shifting Investment Dynamics

2025 has marked a reopening of the IPO window, especially favoring growth-stage B2B SaaS companies and innovative tech firms with strong fundamentals.

High-profile IPOs like Circle and CoreWeave demonstrate renewed investor appetite, with smaller deals sometimes outperforming large ones.

Secondary markets in venture capital are becoming primary liquidity sources, with record transaction volumes and large funds specializing in venture secondaries addressing liquidity constraints.

AI and defense tech sectors continue attracting major funding rounds and valuations, underpinning strategic industry shifts.

Apple’s new “Liquid Glass” design language and UI changes blur lines between iPad and Mac, signaling acknowledgment of evolving user expectations.

AI-driven interfaces are moving beyond traditional input methods to embrace natural language, voice commands, and conversational experience.

Voice AI technologies, such as “Voice in a Box” and true speech-to-speech models that incorporate prosody and emotion, are poised to revolutionize both consumer and enterprise interfaces.

The future of devices will increasingly be defined by AI assistance quality rather than hardware aesthetics, with “legacy” hardware becoming less relevant.

Key Trend 5: Ethical, Social, and Political Implications of AI and Tech Platforms

Major tech companies wrestle internally with AI safety, privacy risks, and ethical governance amid fierce innovation pressures.

AI’s societal impact carries dual potentials for utopia or dystopia, prompting calls for governance frameworks balancing innovation with responsibility.

Social media platform changes, such as X’s transformation and decentralized alternatives like Bluesky, reveal ongoing tensions in moderation, community cohesion, and political discourse.

Criticism of Big Tech growth focus and user experience degradation shows persistent cultural dissatisfaction despite transformative potential.

Discussion Questions

How has TechCrunch’s role as a startup network reshaped the venture capital ecosystem compared to traditional tech media? What lessons does this hold for emerging platforms today?

Given the dominance of B2B and automation-focused AI startups in YC’s recent accelerator cohorts, what does this suggest about the future directions of AI entrepreneurship versus consumer applications?

Apple is pushing hard on design and controlled AI integration, while Meta invests heavily in superintelligence labs—how do these divergent strategies reflect different visions of AI’s role in society and technology?

What are the implications of the IPO resurgence and growing secondary markets for startup founders, investors, and public market investors in the current economic cycle? Does this signal a sustainable tech market rebound or potential volatility?

With ethical concerns rising within companies like Apple and voices like Vinod Khosla warning of AI’s societal risks, what governance or regulatory frameworks should be prioritized to ensure safe and equitable AI development?

How do changes in social media dynamics—such as the rise of decentralized platforms like Bluesky and the transformation of X under Musk—impact political communication and community building in the digital age?

What does the evolution of voice AI and UI convergence (e.g., iPadOS blending with macOS, ‘vibe coding’ tools) mean for how individuals will interact with technology in the near future? Could these trends reduce technical barriers or introduce new challenges?

Closing Segment

TechCrunch’s 20-year journey exemplifies the power of dedicated media to build ecosystems and influence innovation rhythms. As we stand on the threshold of AI-driven transformation, the themes resonate: human connection remains central even as machines advance; technology for good requires intention amid rapid change; and markets and devices evolve to meet new realities while grappling with legacy and complexity.

Our final thought: The future will not be defined solely by the most advanced algorithms or sleekest designs, but by how well the industry balances innovation, ethics, human values, and global inclusion to craft a truly transformative technology landscape.

How can societies balance the undeniable economic benefits of AI-driven abundance with the urgent need to address political and social inequalities in distribution?

Given the increasing capital intensity and funding polarization in AI startups, what strategies should founders adopt to succeed in this evolving venture capital landscape?

What are the ethical and legal implications of AI companies using user-generated online content without explicit consent, and how might this shape AI development and regulation?

How does the increasing integration of major tech companies with military and government agencies affect public trust and innovation trajectories?

What role should governments and regulators play in managing the power of tech giants, especially in light of ongoing antitrust cases and geopolitical economic pressures?

To what extent can emerging social media platforms like Bluesky reshape the media ecosystem, given the persistent dominance of entrenched networks such as X?

How might educational institutions integrate AI tools ethically and effectively to enhance learning without fostering overreliance or academic dishonesty?

Show Notes: What is Abundance? And is it a Good Thing?

Overview

This newsletter explores the concept of abundance, particularly in the context of technology, energy, and capital. It challenges debates about whether abundance is real or manageable, presenting it instead as an unstoppable force rapidly reshaping business, society, and governance. The content spans trends from explosive AI-driven startup growth and energy breakthroughs to shifts in media, venture capital, and political dynamics.

Listeners will find this collection compelling because it connects broad macro forces—technology advances, energy cost collapses, investment flows—with societal and economic changes. It offers a nuanced view that acknowledges friction and obstacles but maintains optimism that abundance is already here, accelerating, and fundamentally altering the rules across multiple domains.

Key Trend 1: Explosive Growth and Changing Dynamics in AI-Driven Innovation and Funding

The emergence of AI as a multiplier of human capability is driving unprecedented revenue growth in startups, reshaping the venture capital landscape, and redefining what “scale” means. Late-stage funding surges and monumental investments in AI infrastructure reflect growing confidence in AI’s commercial potential.

The top 10% of B2B AI startups are achieving astronomic 236% ARR growth, marking a departure from the efficiency-first era to rapid expansion and capturing “escape velocity.”

The size of top-1% venture-backed exits is nearly doubling every five years, signaling massive future capital returns at the intersection of cloud, mobile, and AI platforms.

Late-stage AI investments dominate funding, including mega rounds like Anthropic’s $3.5B Series E, underscoring belief in scalability and profitability.

Oracle’s $40B commitment to Nvidia chips for OpenAI’s new data center exemplifies the scale of capital pouring into AI infrastructure needed to power trillion-parameter models.

The explosion of AI integration across tools, like Perplexity Labs generating complex work products or multiple AI agents collaborating on code, highlights multi-layered adoption in workflows.

Key Trend 2: Energy Abundance as the Prerequisite for Sustainable Technological and Societal Growth

Energy is the foundational “subsidy” enabling societal complexity, climate action, and advanced AI. Rapid advances in solar, nuclear, and fusion research herald a future of "energy too cheap to meter," which will be a game changer as demand explodes.

Energy breakthroughs have historically powered leaps in human development—from fire to fossil fuels—and solar energy is the latest, with plummeting costs creating a tipping point.

Government reform, grid modernization, and deregulation are essential to accelerate adoption and infrastructure buildout.

Upcoming nuclear small modular reactors and fusion research (including AI-assisted catalyst discovery) represent critical next steps along the energy innovation trajectory.

Meeting urgent energy needs is critical for climate solutions, AI’s soaring compute demand, and sustaining democratic institutions and economies.

Key Trend 3: Institutional Friction vs. Market-Led Speed and Execution in Technology and Government

While abundance forces press forward, friction remains, especially rooted in institutional inertia, regulatory complexity, and political coalitions. However, the market champions the agile and fast-executing players who prioritize speed over bureaucratic safety.

Biden administration’s infrastructure rollout illustrates government slowness: trillions in spending with slow or no visible results.

Companies achieving rapid ARR growth routinely bypass “progressive coalition politics” favoring execution and iteration.

Elon Musk’s brief tenure in government showed the challenges of applying private-sector efficiency models to public institutions, ending with his resignation amid political conflicts.

The “Abundance Agenda” calls for governance reform but faces entrenched interest-group resistance, reflecting recurring liberal factional rivalries.

Key Trend 4: The Shifting Media and Information Ecosystem—From Screening to Summarizing, and the Challenge to Web Content

AI-powered search is transforming how users access information, moving from link-based discovery to AI-generated summaries that threaten traditional web traffic patterns and publisher revenue models.

Google’s AI Overviews and AI Mode prioritize summarization over link retrieval, reducing user clicks to websites, shifting how “the web” is monetized and accessed.

This shift generates tensions as content creators face reduced traffic even as their editorial authority becomes more valuable to AI training.

New protocols like Microsoft’s NLWeb aim to make websites more accessible to AI agents, signaling an evolution toward AI-powered conversational interfaces.

Publishers’ survival depends increasingly on establishing verified fact-based content and new business models compensating their data contribution to AI.

Key Trend 5: The Democratization and Accelerated Meme-ification of Venture Capital and Culture

The VC landscape, and culture at large, is increasingly shaped by rapid narrative cycles, social media algorithms, and AI-driven content creation, emphasizing hype and viral content while paradoxically increasing the importance of genuine personal connection and location.

Meme coins like Fartcoin, driven by AI-generated hype, show how narrative velocity can create rapid yet often ephemeral market spikes influencing capital flows.

Social media and AI have democratized “taste,” with rapid cycles of trend formation driven by platform algorithms favoring engagement over depth.

Venture capital branding is evolving to embrace memes and out-of-home advertising to reach broader retail investor audiences, challenging traditional LP communication.

Despite hyper-meme culture, location and face-to-face networks remain crucial as a grounding force amid accelerated, digital-first trends.