May 2026
Prudhvi Vatala, Snap’s head of engineering platforms, argues that the company’s 10-plus-petabyte daily experimentation pipeline became a cost and scale problem that could not be solved by adding more CPUs. In an NVIDIA AI Podcast interview, he says Snap cut job costs by 76% by moving Spark workloads to NVIDIA GPU-accelerated infrastructure on Google Cloud, reusing idle inference GPUs overnight, and doing so without application code changes.
William Costello argues that cross-sex friendship is not best understood as a fragile exception to romantic interest, but as one common route into relationships and a way for men and women to understand each other better. In a discussion with Chris Williamson, Freya India, and Tania Reynolds, he says romantic and sexual interest inside opposite-sex friendships is common, often misread, and not necessarily evidence that the friendship is fake. The group’s central tension is that the same ambiguity that can turn friendship into partnership can also produce wishful thinking, jealousy, backup-mate dynamics, and confusion over what counts as truly platonic.
In an a16z conversation with Erin Price-Wright, former Tesla executives Turner Caldwell and Drew Baglino argue that America’s AI ambitions depend on rebuilding the physical systems beneath them: critical minerals, refining, power electronics, manufacturing and the grid. Caldwell, now CEO of Mariana Minerals, says the US is decades behind China in minerals capacity and must use automation and vertical integration to speed mining and refining. Baglino, CEO of Heron Power, says outdated mechanical grid equipment should be replaced with silicon- and software-based power electronics, backed by durable industrial policy and coordinated infrastructure planning.
Paul Graham told Swedish founders in Stockholm that the way to strengthen the city’s startup ecosystem may begin with leaving it. The Y Combinator founder argued that Silicon Valley remains the startup world’s dominant center because it concentrates ambitious peers, fast-moving investors, chance encounters and a culture of help; Stockholm’s opportunity, he said, is for founders to absorb those advantages and return with stronger companies, capital and habits that can compound locally.
Madison Faulkner and Hugo Santos of Namespace argue that traditional CI/CD is organized around human-paced pull requests, and starts to fail when autonomous agents generate continuous, overlapping streams of code. Their proposed replacement keeps validation inside a stateful agent loop, uses caching and orchestration to avoid cold starts, and moves completed work into a pre-merge layer where humans review intent and outcome rather than every diff. The underlying CI functions remain, but the pull request stops being the system’s basic unit of work.
ElevenLabs is introducing Workflows, a visual editor for its Agents Platform that lets builders design routed conversation flows instead of placing all business logic inside one agent prompt. The company argues that specialized subagents, each with their own instructions, tools, knowledge bases and model choices, give teams more control over cost, latency and accuracy. The product is positioned as a way to combine AI interpretation with predefined actions, verification steps and human handoffs on the same design surface.
Anthropic CFO Krishna Rao argues that the company’s rise is best understood through compute: a scarce capital asset that must be bought years ahead and constantly reallocated across model training, customer demand, internal automation and future products. In an interview with Patrick O’Shaughnessy, Rao says ordinary forecasting and software-margin frameworks break down when model capability, adoption and revenue compound together, leaving Anthropic to manage growth through scenarios rather than point estimates.
Suno founder and CEO Mikey Shulman argues that AI music should not be understood as a cheaper substitute for streaming catalogs, but as a new form of active consumer entertainment. In a conversation with Sequoia’s Sonya Huang, he says Suno’s technical choices — modeling raw sound, prioritizing full songs, and using preference data rather than conventional benchmarks — support a product thesis that making music can be as much the point as listening to it. Shulman also frames partnerships with labels such as Warner as central to building new participatory music formats, not as a concession to incumbents.
Sam Parr and Shaan Puri use a discussion of Jim Ratcliffe, Palmer Luckey and other unconventional founders to argue that edge is usually formed indirectly: through childhood obsessions, adversity, risk and the confidence that comes from surviving unfamiliar situations. On My First Million, they extend that idea from parenting to company-building, saying the same odd specificity that produces billionaire side quests can also show up in scrappy consumer brands, TikTok affiliate machines and businesses that begin with janky tactics before becoming durable.
Google DeepMind, OpenAI, Vercel, SAP, Adaptive ML, and CME’s compute-futures plan all point to the same applied-AI shift: agents are being designed around the conditions that let them operate safely in real workflows. The open questions are less about whether models can act and more about reference, permissions, memory, business context, feedback, evaluation, and compute exposure.
Google DeepMind researcher Adrien Baranes argues that the mouse pointer can become more than a tool for selecting and clicking. In an experimental prototype, he presents the cursor as an AI-mediated reference layer: a way for Gemini to connect words such as “this,” “that,” and “here” to the precise objects, app data, and screen content a user is indicating. The aim is to make pointing function as shared context between a person and an AI system across documents, calendars, maps, and images.
In a Hoover Institution discussion on civil-military relations, retired generals Joseph Dunford and Christopher Cavoli, with H.R. McMaster moderating, argued that the U.S. military’s democratic role depends on a hard boundary: senior officers give candid, private, nonpartisan advice; civilian leaders make policy; and the military carries out lawful orders. Their case was that this discipline is not Washington etiquette but a practical safeguard for the force, ensuring Americans are not sent into danger without an achievable political objective, clear risks, and accountable civilian decisions.
OpenAI’s trial, Anthropic’s secondary-market flare-up, and two media deals are read on Diet TBPN as fights over control, enforceability, and credibility. John Coogan argues that Musk v. OpenAI is increasingly not only about whether OpenAI betrayed its nonprofit mission, but whether Elon Musk accepted a for-profit path only if he controlled it; Jordi Hays frames the Anthropic panic as a test of whether private-company transfer restrictions can hold against demand for AI exposure. Coogan and Hays treat Thinking Machines’ demo separately, as a bet that real-time interaction should be native to AI models, while eBay’s rejected GameStop bid and Byron Allen’s BuzzFeed investment turn on market confidence.
Condé Nast chief executive Roger Lynch argues in a TBPN interview that publishers should plan for a media market in which search traffic is no longer a reliable foundation and generic AI content is not a defensible advantage. His case is that brands such as Vogue and The New Yorker can become more valuable if they rely on direct audience demand, subscriptions, events, editorial authority and human-reported work, while using AI mainly to make product and technology teams faster.
Charles and Chase Koch used an All-In interview to explain Koch Industries’ rise from a 300-person company in 1961 to a private conglomerate they say is worth 9,000 times more today. Their central argument is that Koch’s refusal to go public was not incidental but essential: private ownership let the company build around transferable capabilities, long-cycle culture change, values-first talent, and experiments whose learning could matter more than near-term earnings. They extend the same framework to education, philanthropy, politics, and AI, arguing for bottom-up contribution over centralized control.
AI and media incumbents are being forced to respond to systems changing faster than their strategies, regulations or business models. Sriram Krishnan, Aarthi Ramamurthy and Condé Nast chief executive Roger Lynch make that case across AI regulation that may miss the next generation of products, private AI investing repackaged through SPVs, and media businesses built on platform traffic that is disappearing. Lynch’s counterpoint is that media companies can still endure if they move away from click incentives and toward authority, direct audience relationships and human creative work.
OpenAI’s Ari Weinstein argues that computer use turns Codex from a coding agent into a system that can operate local Mac applications by seeing interfaces, clicking, typing and continuing work in the background. In a demonstration with Romain Huet, Weinstein presents the feature as distinct from a full-desktop takeover: Codex uses a separate cursor, combines screenshots with macOS accessibility data, and requires app-by-app permission before it can see or type into local software.
A South Korean policy chief’s proposal to return part of AI-related gains to citizens jolted the country’s chip market, with Samsung and SK Hynix closing down around 5% after Kim Yong-beom argued that profits from the AI infrastructure era should be shared more broadly. Bloomberg reported that the presidential office later described Kim’s post as personal opinion, while the same program pointed to related pressure points in the AI boom: CME’s plan with Silicon Data for compute futures and Nvidia CEO Jensen Huang’s absence from Trump’s China delegation as approval for Blackwell sales looked unlikely.
Nico Albanese, a Vercel engineer working on the AI SDK, argues that agents become more reliable when they are given a persistent sandboxed computer, not just a runtime and tools. In his workshop, he builds that pattern with AI SDK 6, Vercel’s named sandboxes, a bash tool, and a file-backed memory system, showing how an agent can plan in files, preserve context across sessions, and create reusable scripts without a separate memory layer.
CME Group and Silicon Data are trying to make computing power tradable as a futures product, Bloomberg’s Katherine Doherty says, using an index of compute prices as the basis for contracts that would let companies and investors hedge future price moves. Doherty frames the plan as an effort to treat GPU processing capacity less as a procurement cost and more as a commodity exposure, though the market still needs regulatory approval and enough liquidity to function.
SAP chief executive Christian Klein used Bloomberg Technology to frame the company’s new autonomous enterprise platform as a bet that AI agents need business context more than proprietary models. He argued that SAP’s advantage is its access to ERP data and process knowledge, which can make agents reliable enough to coordinate work across finance, commerce, inventory, procurement and supply chains. Pressed on competition from partners such as AWS, Klein said SAP’s role is to provide the enterprise context layer while working with hyperscalers and data platforms to harmonize data beyond SAP systems.
A-Star co-founder and general partner Bennett Siegel argues that venture-style outperformance still depends on small, early checks rather than joining the largest AI financings. In a Bloomberg Tech interview, Siegel said A-Star’s new $450 million fund is designed to preserve its seed-stage discipline: backing a limited number of companies before products, markets and consensus are fully formed, then following on selectively as winners emerge. He frames billion-dollar formation rounds as a separate market, not proof that traditional seed investing has lost its relevance.
Alessandro Cappelli, co-founder and chief customer officer of Adaptive ML, argues that enterprise generative AI pilots fail to reach production because companies lack a systematic way to turn defects, user feedback, business metrics and production signals into model improvement. In a talk on Fortune 500 deployments, he says prompting and instruction fine-tuning can produce credible demos, but reinforcement learning is the mechanism needed to train models and agents against enterprise-specific environments, rewards and KPIs. His case is that agents make this feedback loop more urgent, because they consume more tokens, touch live systems and leave less room for error.
In a Stanford CS153 Frontier Systems lecture, General Matter chief executive Scott Nolan argues that AI’s infrastructure constraint is moving upstream from chips and data centers to electricity. For high-uptime, low-carbon data-center power, Nolan says the long-term answer points toward nuclear, but the decisive U.S. bottleneck is not reactors themselves; it is uranium enrichment, a capability he says the country has largely lost and that General Matter was founded to rebuild.