The Frontier

A demand crunch for AI compute tokens is driving a business model shift from flat-rate subscriptions to usage-based billing. GitHub Copilot cited unsustainable inference costs in its pricing change.

GPU rental prices have increased 40% over the last six months due to real token demand. Oguz Erken notes the top two AI labs generate nearly $60B in aggregate annual revenue, driven by fundamentals.

Nathaniel Whittemore argues we're entering a phase where AI is critical global economic infrastructure, not a startup-era novelty. This is reflected in Big Tech earnings and massive private valuations.

Big Tech cloud earnings showed massive AI-driven growth: AWS up 28% year-over-year, Microsoft Azure up 40%, and Google Cloud up 63%, beating estimates and causing a historic market cap jump.

Anthropic is in talks to raise at a valuation above $900B, surpassing OpenAI's last valuation of $825M. Secondary market trades suggest some investors value Anthropic near $1T.

Microsoft and OpenAI restructured their deal, removing the AGI clause and granting Microsoft free access to OpenAI models for a half-decade. OpenAI is now free to sell models through AWS and Google Cloud.

The White House considers unwinding Anthropic's supply chain risk designation. However, US government officials oppose Anthropic's planned Mythos model rollout, citing national security and compute constraint concerns.

Dean Ball argues the US government restricting model rollouts like Mythos constitutes an informal, improvised licensing regime. He states the 'trial runs are over' for AI policy.

Product development is shifting focus from raw models to the 'harnesses' or interfaces that deliver AI capabilities. New tools like Cursor's SDK and OpenAI's updated Cursor aim to simplify agent deployment.

Anthropic split technical and non-technical work into Claude Code and Claude Cowork, while OpenAI's Cursor bets on a single interface for all knowledge workers, promoting technical skill acquisition.

OpenAI traced a 'goblin' bug in Cursor to a personality reinforcement learning artifact from GPT-5 models. The quirk highlights how model interdependencies can amplify unusual behaviors.

The viral 'MS Paint' image prompt instructs AI to redraw an image in a 'clumsy, scribbly, and utterly pathetic way' to mimic low-quality mouse-drawn art. It spread on Threads and Asian social channels.

Nathaniel Whittemore is skeptical of Silicon Valley predictions like a 'permanent underclass,' arguing builders often misunderstand real-world technology diffusion and broader economic forces.

Nathaniel Whittemore argues developers are skipping the AI assistant phase entirely, building 'digital employees' with job titles, IDs, and human-style performance policies like the three-strike termination rule at Myze.

The goal of these agentic hierarchies is to find the minimum human intervention needed to run a business. Whittemore says removing the human safety net exposes the friction points in current AI reasoning.

Whittemore identifies a massive infrastructure gap: agents have no native long-term memory. Builders currently duct-tape solutions with markdown 'brain files', vector databases, or manual context injections for every session.

Some developers resort to extreme measures to combat agent amnesia, like running autonomous systems three times a day just to keep internal logic fresh, highlighting the field's critical need for persistent memory.

The plummeting cost of software creation enables markets of one. Whittemore highlights non-technical experts building hyper-specific agents, like a person with Graves' disease creating a thyroid flare detector from personal health data.

These niche agents solve problems traditional VC-backed companies ignore. An Arkansas kayaker built a rain-fed creek predictor, showing agentic coding is shifting who gets to build useful software.

Google released an agentic suite called Gemini Intelligence for Android before its IO event, featuring an upgraded Gemini assistant and a personal intelligence memory system rolling out across devices this summer.

Google is exploring orbital data centers through talks with SpaceX and other rocket companies, with prototypes targeted for orbit by next year, following similar interest from Anthropic and Nvidia.

Google launched an AI consulting group within Google Cloud hiring hundreds of forward-deployed engineers to help enterprise customers with agent development, mirroring moves by OpenAI and Anthropic.

Anthropic expanded its vertical strategy with Claude for Legal, adding connectors to tools like DocuSign, Trellis, and Harvey alongside 12 pre-built agents, after legal professionals became their most engaged knowledge worker users.

Nathaniel Whittemore argues the backlash against token maxing, incentivizing employees to use more AI tokens, stems from two flawed narratives: that AI isn't good and that AI is a bubble, both increasingly disconnected from reality as models improve and revenue soars.

Incentivizing token usage is a practical response to the shift from assisted to agentic AI, where experimentation is necessary to figure out new workflows, because managing agents is a new work primitive without established best practices.

Token incentive critics commit logical fallacies by generalizing from isolated reports of gaming - like Amazon employees automating unnecessary tasks - to claim most token consumption lacks economic value, ignoring selection bias and the traceability of actual usage.

Nathaniel argues that experimentation inherently produces token consumption with delayed financial gain but immense learning value, and that companies with token incentive programs will outpace those afraid of waste, even with some gaming.

Companies are developing alternative metrics to token consumption, exemplified by Salesforce introducing agentic work units to measure output and impact directly.

Whittemore argues Q1 2026 marked AI's second moment, shifting from viable chatbots to workable agentic systems with higher economic and corporate stakes.

Industry AI capex is projected to reach $650 billion in 2026, triple the spending from a couple of years ago and exceeding the inflation-adjusted cost of the US interstate highway buildout.

Claude Code's annualized revenue grew from $1 billion to $2.5 billion in about two months during Q1 2026, while its use expanded beyond coding to preview broader agentic trends.

Cursor doubled its annualized revenue to $2 billion this quarter, Lovable hit $400 million ARR with a $100M monthly jump, and Replit projects $1 billion ARR by end of 2026.

Anthropic reached a $19 billion run rate and captured 70% of first-time enterprise AI buyers per Ramp data, while OpenAI held 25% and about $25 billion in annualized revenue.

Gartner predicts 40% of enterprises will have working agents in production by end of 2026, fueled by products like agent credit cards from Ramp and Stripe.

AI usage surveys show 71% of practitioners 'vibe coded' in the past month, and 62% used automation or agentic AI, with the average user employing 3.5 different models.

Survey data indicates a shift from efficiency AI to opportunity AI, with time-saving use cases dropping from 19.9% to 13.6% of ROI between January and February 2026.

HR AI deployment grew 320% in 12 months from 19% to 61% adoption, while seven states have enacted AI employment regulations.

The Pentagon designated Anthropic a supply chain risk after a dispute over using Claude in military operations, while OpenAI's subsequent DoD deal triggered a 775% surge in one-star ChatGPT reviews.

The Generative Engine Optimization market was valued under $1 billion in 2025 but is projected to reach nearly $34 billion by 2034 as AI referrals convert better than traditional search.

Whittemore notes a widening capability overhang where the gap between potential AI value and actual deployed value increases costs, creating a larger divide between leading and lagging companies.

McKinsey's AI transformation manifesto identifies 12 themes separating AI leaders from laggards, arguing technology alone creates no advantage; enduring systems and capabilities built around tools drive success.

Leading companies focus AI on economic leverage points within their business model, not just productivity. McKinsey found AI transformations in 20 leading companies delivered a 20% EBITDA uplift.

McKinsey states AI leaders see break-even on investments in 1 to 2 years and generate $3 of incremental EBITDA for every $1 invested, framing AI as a growth and opportunity technology.

George Zarkadakis argues institutional AI requires a coordination layer to align individual AI outputs, preventing chaos from thousands of unmanaged agents with divergent styles and outputs.

Ramp built an internal AI workspace called Glass to solve coordination and setup pain. It auto-configures with all company tools via SSO and includes a marketplace of over 350 reusable skills.

Seb's essay on Ramp's Glass argues against simplifying AI tools for non-technical users. The goal is to make complexity invisible while preserving full capability, enabling power-user workflows for everyone.

Ramp's Glass system includes an AI guide called Sensei that recommends relevant skills from the marketplace based on a user's role and connected tools, aiming to surface the five most useful skills on day one.

Seb argues internal AI infrastructure is a competitive moat; companies that make every employee effective with AI will compound advantages competitors cannot match. Ramp built Glass in-house for speed and direct feedback.

McKinsey argues that agentic engineering is the next capability to master, involving ingesting unstructured data, extending AI platforms with agentic capabilities, and automating guardrails.

Guillermo Rauch announced Vercel is open-sourcing a reference platform for cloud coding agents, aiming to help others build the type of internal systems companies like Stripe, Ramp, Spotify, and Block use.

Nathaniel Whittemore identifies Q2 2026 as AI's "second moment," marking a shift from viable assistant chatbots to workable agentic systems, which he deems the most consequential period since ChatGPT's launch.

The "AI second moment" signifies dramatically scaled capabilities, with weekly active users reaching billions and economic stakes involving $650 billion in projected capital expenditure this year, signaling a major corporate reorientation.

Nathaniel Whittemore points to Q4 2025/Q1 2026 as an inflection point, driven by new models like Opus 4.5 and GPT 5.2, alongside transformative capabilities from Claude Code and Codex, leading to record frontier model releases.

Q1 2026 was the "quarter of Open Claw," an agent project that evolved from Claude bot, became GitHub's most starred open-source project, and was subsequently recruited into OpenAI. Nvidia's Jensen Huang deemed it a highly significant software release.

Nathaniel Whittemore describes Anthropic and OpenAI converging in strategy; Anthropic gained 70% of first-time enterprise AI buyers, even as OpenAI, with higher overall annualized revenue of $25 billion, sought to consolidate its products.

Enterprise AI adoption is shifting from pilots to production agents; Gartner forecasts 40% of enterprises will have working agents by year-end 2026, supported by new financial tools for agents to spend money.

Q1 2026 saw a "SAS apocalypse" as investor concern shifted from AI's potential insufficiency to its disruptive power, evidenced by Block cutting 40% of its staff, alongside substantial AI company revenue growth.

Leading AI companies reported immense revenue growth, with Claude Code reaching $2.5 billion annualized revenue, Cursor doubling to $2 billion, and Anthropic achieving a $19 billion run rate.

Practitioner surveys show widespread AI usage, with 71% engaging in "vibe coding" and average users employing 3.5 models. The perceived value of AI shifted from time savings to increased output and new capabilities.

AI adoption varied across sectors: HR deployments grew 320% in 12 months, and sales emerged as the most mature function with 63% of use cases "primetime." Finance adoption was high, but 91% reported low impact due to data quality.