Welcome to #53

Which means this is our 1 year anniversary of the DAS newsletter 🥳

In this issue:

The Agentic Web: Is This the End of SaaS as We Know It?

Are We Losing Ourselves as AI Takes Over Our Lives?

Meeker’s Data Drop: The Numbers Every Business Leader Needs on AI

Plus, we discuss AI love on the subway, Super Micro is acting like Super Mario, the problem with infinite content, and all the news we found interesting this week.

It’s Sunday morning!

Don’t worry if you forgot our anniversary.

AI doesn’t hold grudges.

…..maybe

The DAS Crew - Andy, Beth, Brian, Eran, Jyunmi, and Karl

The Agentic Web: Is This the End of SaaS as We Know It?

A new wave of AI agents is changing how businesses use software. Instead of logging into platforms and clicking through dashboards, users will soon describe their goals and let agents handle the work. Microsoft CEO Satya Nadella recently outlined this shift, predicting that intelligent agents will manage complex workflows, connect to any back-end data source, and deliver exactly what users request, all with a single prompt.

In this “Agentic Web,” legacy SaaS platforms will become just another source of data. New protocols like MCP (Model Context Protocol) and native connectors allow agents to access and aggregate information across the enterprise. Satya’s vision puts the focus on results, not data silos or proprietary interfaces. Any tool that cannot connect to these agents risks falling behind.

This transformation requires more than new technology. Companies must rethink customer experience, security, and business models from the ground up. Businesses that adapt quickly, improve orchestration, and enable seamless human-agent collaboration will win.

WHY IT MATTERS

The Application Layer Is Changing: AI agents will handle full business processes across multiple systems, which reduces the need for working in vertical SaaS apps .

Legacy Data Moats Are Disappearing: Software providers that do not open up and support new Agent access protocols will lose relevance as agents choose the fastest, most accessible sources.

Security and Permissions Matter: As agents automate more tasks, companies must enforce strong access controls and clear data boundaries to protect sensitive information.

The User Experience Evolves: People will stop clicking and searching and copying and pasting. They will describe their needs, and agents will coordinate the rest in the background.

Speed and Flexibility Win: Companies that move fast and connect easily through agentic protocols will take the lead as software and workflows change.

Are We Losing Ourselves as AI Takes Over Our Lives?

As AI transforms the workplace, it’s quietly changing something deeper—how we find meaning, purpose, and belonging. In cultures where identity often comes from what we do, the rise of automation and intelligent agents is forcing people to ask: If my job isn’t the center of my life, what is?

For generations, purpose came from three main sources: work, family, and community (including religious groups or other “third spaces”). Now, as AI reshapes or even replaces work, many are struggling to rebuild that sense of belonging elsewhere. Communities are being fractured by the speed of change, leaving some people searching for new anchors or resisting AI entirely.

Even the pursuit of mastery, long seen as a path to meaning, is evolving. Some find joy and connection in using AI creatively, while others feel it erodes what made their expertise special. AI doesn’t just automate tasks; it can simulate companionship, guidance, and even spiritual presence. For some, that’s enough. For others, the simulation feels hollow, raising hard questions about what is real and what is just easy.

With loneliness rising and traditional communities in decline, the challenge is clear: AI’s impact is as much about the heart as the economy. Individuals and organizations must rethink how to build real connection, share purpose, and stay grounded, even as technology becomes a bigger part of daily life.

WHY IT MATTERS

Purpose and Identity Are Evolving: As AI automates or redefines jobs, people must seek meaning beyond traditional work, finding new ways to contribute and belong.

Communities Need Reinvention: The decline of physical workplaces and interest-based local communities leaves a gap. New kinds of communities built around AI, or built around rejecting it, are emerging, but many people risk falling through the cracks.

Mastery Feels Different: The satisfaction from mastering a skill is changing as AI takes over creative, intellectual, and even social tasks. For some, AI opens new doors. For others, it erodes the purpose endowed by experienced craftsmanship.

Belonging Can’t Be Automated: AI can simulate connection, but genuine meaning comes from real relationships, mutual support, and purposeful giving to others.

Responsibility Remains Human: Ultimately, it’s up to individuals and communities to decide how to use AI in ways that support human dignity and shared goals, not just efficiency or convenience.

Meeker’s Data Drop: The Numbers Every Business Leader Needs on AI

Legendary internet analyst Mary Meeker just released a 300+ page report that lays out the state of AI with more data, speed, and clarity than anything else this year. Her message is simple: AI’s growth and impact are not just unprecedented, they’re reshaping the economy and our daily lives at a pace that leaves even the internet boom in the dust.

The numbers back it up. AI adoption is already moving from “experiments” to deployment in real businesses. At Kaiser Permanente, over 10,000 doctors use AI medical scribes. In San Francisco, a third of all ride hailing trips are now handled by autonomous vehicles. The FDA approved more than 220 AI-powered medical devices in 2023 alone, compared to just five a decade ago. And at the developer level, 63% globally now use AI tools, up from 44% just last year. Google processes 480 trillion AI tokens monthly, a 50x year-over-year jump.

Meeker’s report is about strategy. She lays out a clear three-stage playbook for leaders: immediately integrate AI into current products and workflows, start building AI-native products in the mid-term, and make long-term bets on truly transformative technology like AGI. Her optimism is striking, urging companies to prioritize speed and innovation over caution and to view AI as a pivotal shift on par with the printing press or the internet itself.

The report also surfaces new risks. Gaps remain around AI equity, governance, copyright, and environmental impact. And Meeker’s analysis is more “catch-up” than forward-looking, offering a snapshot of where we are today rather than a prediction of what’s next. Still, her data-driven approach gives business leaders a powerful map for understanding and navigating AI’s explosive evolution.

WHY IT MATTERS

Unprecedented Speed and Scale: The pace of AI adoption is faster than any previous tech revolution, and waiting to adapt could mean being left behind.

AI as Core Infrastructure: Meeker sees AI becoming as fundamental as the internet, changing how every industry operates and how every individual interacts with technology.

A Clear Playbook for Action: Companies should move quickly to adopt, experiment, and iterate, integrating AI today, planning for AI-native offerings, and investing in long-term bets.

Major Gaps Still Exist: Issues like equity, governance, copyright, and environmental impact need urgent attention to ensure that AI’s benefits are widespread and sustainable.

Data-Driven, Not Hype-Driven: Leaders finally have a comprehensive, numbers-first roadmap for understanding and planning around real AI impact and not just headlines or speculation.

Just Jokes

News Story: A viral photo showing a man tenderly conversing with ChatGPT on a New York City subway has stunned the internet and sparked widespread debate over AI companionship.

Did you know?

Super Micro Computer, a company previously under the radar, has surged up the Fortune 500 list by 206 spots, now ranking at number 292. This remarkable ascent is attributed to its strategic focus on artificial intelligence, cloud computing, and data center infrastructure. The company doubled its revenue to $14.99 billion and recorded $1.15 billion in earnings, marking the highest one-year profit growth in its industry. Super Micro's partnerships with tech giants like Nvidia and Intel have positioned it as a preferred vendor for AI-optimized infrastructures.

Notably, Elon Musk’s xAI team selected Super Micro to develop a 750,000-square-foot data center in Memphis. The company's commitment to innovation, including investments in green computing and energy-efficient systems, underscores its rapid growth in the AI sector.

Source

This Week’s Conundrum
A difficult problem or question that doesn't have a clear or easy solution.

The Infinite Content Conundrum

AI is quickly moving past simple art reproduction. In the coming years, it will be able to reconstruct destroyed murals, restore ancient sculptures, and even generate convincing new works in the style of long-lost masters. These reconstructions will not just be based on guesswork but on deep analysis of archives, photos, data, and creative pattern recognition that is hard for any human team to match.

Communities whose heritage was erased or stolen will have the chance to “recover” artifacts or artworks they never physically had, but could plausibly claim. Museums will display lost treasures rebuilt in rich detail, bridging myth and history. There may even be versions of heritage that fill in missing chapters with AI-generated possibilities, giving families, artists, and nations a way to shape the past as well as the future.

But when the boundary between authentic recovery and creative invention gets blurry, what happens to the idea of truth in cultural memory? If AI lets us repair old wounds by inventing what might have been, does that empower those who lost their history or risk building a world where memory, legacy, and even identity are open to endless revision?

The conundrum
If near-future AI lets us restore or even invent lost cultural treasures, giving every community a richer version of its own story, are we finally addressing old plunder, cultural displacement and injustices, or are we quietly creating a world where the line between real and imagined is impossible to hold? When does healing history cross into rewriting it, and who decides what belongs in the record?

Want to go deeper on this conundrum?
Listen/watch our AI hosted episode

Anthropic Cuts Windsurf’s Access to Claude Models Amid Acquisition Rumors
Anthropic abruptly restricted Windsurf’s access to Claude 3.7 Sonnet and Claude 3.5 Sonnet, just as rumors swirled about Windsurf being acquired by OpenAI. This follows Anthropic’s push for developers to move to its own Claude Code platform, now gaining traction among leading AI coders.

Deeper Insight:
The fight for developer mindshare is heating up. As Claude Code grows and providers assert more control, expect ongoing competition, shifting alliances, and more lock-in risks for teams that depend on third-party AI assistants.

Character.AI Expands with Scenes and Avatar FX for User-Created Conversations
Character.AI rolled out new multi-modal features, including Scenes and Avatar FX, to all users. These tools let people script and share avatar-driven conversations, adding an extra layer of immersion and creativity to the platform.

Deeper Insight:
This opens the door for even more personal and viral AI-generated content. As these avatars start appearing in shorts and TikToks, the line between real interaction and wish fulfillment will blur further, raising new questions about identity and digital storytelling.

Epic Games Brings Generative AI NPCs to Fortnite
Epic Games introduced generative AI-powered NPCs in Fortnite, starting with a Darth Vader character. The beta rollout was paused after users got the AI to say unscripted lines, but the underlying technology is now available to all Unreal for Fortnite developers.

Deeper Insight:
Generative AI is changing how game stories evolve in real time. This shift will let games create more unique, unpredictable experiences, but it will also force studios to rethink content moderation and player safety.

Sakana AI Unveils the Darwin Gödel Machine, a Self-Improving AI Agent
Sakana AI announced the Darwin Gödel Machine, an agent that evolves and modifies its own code after deployment. Inspired by evolutionary biology, it continuously enhances its own reasoning and capabilities.

Deeper Insight:
Self-improving agents mark a major shift from today’s static AI systems. The promise is more adaptable and creative agents, but this also brings new safety and oversight challenges as these systems grow beyond their original design.

Manus AI Agent Adds Video Generation for Complete Story Creation
Manus now allows users to generate video stories from prompts, automating complex, multi-step creative tasks within a single workflow.

Deeper Insight:
AI agents are no longer just helpers—they’re taking over full content pipelines. Users will increasingly direct, not produce, digital projects, making hands-off creation the new normal.

Meta Announces AI-Generated Ads By 2026 Without Required Disclosure
Meta revealed that, starting in 2026, businesses can have AI generate ad campaigns automatically, with no guarantee that AI-generated ads will be labeled as such.

Deeper Insight:
This could make targeted advertising cheaper and more effective, but a lack of disclosure will likely fuel transparency debates. The pressure is now on regulators to set clear guidelines for AI-generated marketing.

Anthropic Launches 'Claude Explains,' a Blog Section Written by AI
Anthropic’s new Claude Explains blog is fully AI-generated, with human editors curating the output. The move follows a growing trend as more companies test AI-written news and educational content.

Deeper Insight:
This portends a new model of product documentation, but readers will need to pay closer attention to source and trust signals. As machine writing becomes standard, the line between editorial voice and algorithmic output will blur even further.

Claude Deep Research Adds Zapier and Other Third-Party Integrations
Claude’s deep research feature now lets users integrate with Zapier, making it easy to pull data from thousands of external apps for analysis and reporting.

Deeper Insight:
Plug-and-play integrations make LLMs true workflow orchestrators. The more business apps LLMs can touch, the more effective they become, moving closer to being indispensable for day-to-day productivity.

WebBench Benchmark Ranks AI Browser Agents
WebBench, a new open-source benchmark, lets developers test browser-based AI agents on navigation, form filling, and automation tasks. Anthropic’s Sonic 3.7 is currently the top performer, just ahead of Skyvan and OpenAI.

Deeper Insight:
Standardized testing will accelerate improvements in AI web agents. With clear scores and leaderboards, developers can quickly spot and close gaps in usability and reliability.

RunnerH and Genspark Push the Boundaries of Agentic Automation
RunnerH and Genspark, two emerging agentic platforms, demonstrate live how they can select and scrape websites, generate campaign plans, book meetings, and even place AI conversational phone calls, all with minimal user input.

Deeper Insight:
This signals a shift from single-step chained automation to fully orchestrated, multi-task workflows conceived and conducted by Planning Agents. Expect more teams to experiment with agents that can carry out complex sequences in minutes, not days.

Palantir Contract Stokes Privacy and Surveillance Concerns
Palantir secured a US government contract to use AI for mass detection and identification tasks. Critics warn that such capabilities could be used for surveillance beyond military targets, including immigration enforcement or political profiling.

Deeper Insight:
AI surveillance is becoming a policy and civil rights flashpoint. The tools built for security can easily be repurposed, and debates over acceptable use will intensify as capabilities grow.

States’ Rights Debate Over Federal AI Regulation
A proposed federal law in the current reconciliation bill would block states from making their own AI rules for a decade. Detractors like Marjorie Taylor Greene argue this is a dangerous overreach and undermines the protections of State sovereignty over such matters.

Deeper Insight:
Centralizing AI regulation may speed up innovation, but it also risks ignoring unique regional needs. The tug-of-war between national uniformity and state-level experimentation and self-regulation will only grow as AI becomes more embedded in society.

Marc Andreessen Urges US to Build National Robot Supply Chains
Marc Andreessen called for the US to do the intelligent design of robotic supply chains on the coasts and manufacture robots in interior cities, setting up a distributed approach to mass-producing general-purpose robots.

Deeper Insight:
This vision frames robots as the new backbone of American industry. If adopted, it could define a new era of distributed, high-tech manufacturing and workforce transformation.

OpenAI Rolls Out Memory for Free Users
OpenAI is giving free ChatGPT users access to a basic version of its memory feature, allowing for more persistent and context-aware conversations.

Deeper Insight:
Universal memory brings new privacy and expectation challenges. Users will need to weigh convenience against control of their own data in everyday AI use.

Hollywood Greenlights Sam Altman and OpenAI Biopic, “Artificial”
A feature film titled “Artificial” is in the works, focusing on Sam Altman and the dramatic events at OpenAI during Thanksgiving 2023. The script will be written by a Saturday Night Live alum and is expected to have comedic moments.

Deeper Insight:
AI leaders are now pop culture figures, and their stories will shape public opinion about the industry. Expect more tech biopics and documentaries as AI continues to move from niche to mainstream.

Tokyo University of Science Reveals Self-Powered Synapse for Color Vision
Scientists at Tokyo University of Science created a solar-powered artificial synapse that mimics human color vision, offering near-human spectral recognition without external power.

Deeper Insight:
Breakthroughs like this can make machine vision cheaper, more efficient, and self-sustaining, opening new doors for autonomous vehicles and bio-inspired hardware in everyday life.