Insights

Bankless covered advances in decentralized infrastructure including 'hallucinated servers' using FHE and MPC cryptography (demonstrated at Devcon SEA 2024), and World Markets, a fully onchain exchange built on MegaETH combining spot markets, margin accounts, and perps without backend servers.
Ethereum ecosystem developments include new developer tools from the Ethereum Community Foundation (Ethereum MCP for Claude AI integration, Blobdrop, Swapboard), Aave V4 launch, and Ethereum reaching a daily transaction ATH of 3.6M.
Regulatory updates covered California banning insider trading in prediction markets and Treasury Department releasing new stablecoin issuer guidelines, while major onchain activity saw 1inch surpass $800B routed and Bitmine acquiring 101k+ ETH.

The White House released its first national AI policy framework seeking federal legislation by end of 2026 and declaring AI training on copyrighted material likely legal, while Sen. Marsha Blackburn simultaneously released a competing 300-page federal bill with stricter criminal penalties.
The Pentagon signed AI deals with seven companies including SpaceX, OpenAI, Google, and Microsoft, excluding Anthropic over military guardrails dispute; xAI launched Grok 4.3 with 1M-token context window and 40-60% cheaper pricing.
Security researchers exposed critical vulnerabilities in the White House iOS app including unverified JavaScript execution and location tracking; OpenAI patched ChatGPT data exfiltration flaws and Codex GitHub token vulnerabilities.

Stripe launched its Tempo blockchain with the Machine Payments Protocol (MPP) offering 10,000 TPS and native AI agent payment support, while Coinbase and Cloudflare backed the competing x402 micropayment protocol, which joined the Linux Foundation with support from Google, AWS, Microsoft, and Visa.
Safe launched SafeNet, a decentralized validator network using Byzantine Fault Tolerant validators to block malicious transactions and address the $3.4 billion in crypto theft from 2025, with SAFE token staking providing economic security.
A draft U.S. stablecoin bill (Clarity) proposed banning yield on stablecoins, while Department of Labor opened $8 trillion in retirement funds to crypto investments and the Ethereum Foundation continued its ETH staking strategy amid quantum computing research threatening current crypto signatures.

AI agents are converging with stablecoins and high-throughput blockchains to reshape internet commerce, with AI requiring full DeFi services (borrowing, lending, yield generation) since they cannot access traditional banking, positioning Ethereum's ecosystem as their potential financial operating system.
Aave V4 launched on Ethereum mainnet with hub-and-spoke architecture enabling independent lending markets to share liquidity, though the protocol faced significant challenges including a $270M Drift Protocol exploit, key team departures (BGD Labs, Aave Chan Initiative, Chaos Labs), and ~40% TVL decline amid April 2026's record $635M in total DeFi exploits.
Ethereum Economic Zone (EEZ) and EIP-8141 will enable synchronous cross-chain composability allowing rollups to execute smart contracts across chains in real-time within single transactions, while Coinbase, Charles Schwab, and Meta expanded crypto adoption through mortgage collateralization, direct trading access, and USDC creator payouts respectively.

SaaStr and TLDR newsletters report that AI agents are fundamentally disrupting B2B software economics, with SaaStr transforming from 20 human employees to 3 humans plus 20 AI agents while maintaining revenue, and Salesforce, ServiceNow, and HubSpot restructuring pricing models around outcome-based and AI-native architectures rather than per-seat licensing.
Human judgment and strategic decision-making are replacing execution speed as competitive advantages, as shallow AI startups offering only AI-wrapped features face collapse, while companies like Amazon (reaching $15B AI revenue run rate) and Atlassian succeed by embedding AI deeply into product strategy rather than bolting it on superficially.
Enterprise buyers are shortening contract lengths to 8 months and demanding paid trials over free trials due to AI market uncertainty, with Claude, OpenAI, and Anthropic tools driving down SaaS renewal prices by up to 45%, while top 10 private AI companies ($1.93T valuation) now rival all 115 public SaaS companies combined ($1.88T valuation).

OpenAI shut down Sora, ended its Disney deal, and paused UK data center plans; Sam Altman stepped back from safety oversight to focus on a $10B fundraise and new model codenamed 'Spud'; OpenAI achieved record $122B funding round and $100M ad revenue in 6 weeks before struggling with 7x lower click-through rates than Google benchmarks.
SpaceX filed for the largest IPO in history seeking $75B at $1.75T valuation after acquiring xAI; Meta launched closed-source Muse Spark model breaking from Llama open-source tradition; Anthropic's Claude gained autonomous Mac control capabilities while facing Pentagon supply chain risk designation.
Prediction markets Kalshi and Polymarket grew to 5.1M+ monthly users but face regulatory pressure from state attorneys general over unlicensed gambling; Google Quantum AI warned quantum computers could crack Bitcoin and Ethereum wallets with under 500,000 qubits by 2029, threatening 6.9M vulnerable Bitcoin addresses; supply chain attacks compromised LiteLLM (97M downloads) and multiple npm packages stealing

Anthropic launched Claude computer use capabilities and Claude Security for enterprise code scanning, while Apple plans a major Siri overhaul debuting at WWDC on June 8, and OpenAI shut down Sora to focus on productivity tools ahead of a potential IPO.
Leading open-source LLMs DeepSeek-v4 and Kimi-K2.6 emerged with massive context windows and agentic AI capabilities, while Google DeepMind, AMD, and LangChain are featured speakers at AI Dev 26 conference on April 28-29 in San Francisco.
Iranian drone strikes targeted Amazon Web Services data centers in Bahrain and UAE disrupting cloud services, while the U.S. announced a $1T 'Pax Silica' fund for supply chain security and SoftBank reached its highest debt-to-asset ratio since 2021 following its OpenAI investment.

NYSE (via owner Intercontinental Exchange) partnered with Securitize to build a 24/7 blockchain-native tokenized securities trading platform for stocks and ETFs, with Securitize becoming the first digital transfer agent eligible to mint tokenized securities. Ondo Finance partnered with Broadridge to enable proxy voting and shareholder rights for tokenized equity holders, while Computershare provides transfer agent infrastructure.
Tether generated $10B in profit in 2025, launched federally regulated US stablecoin USAT, and hired a Big Four firm for its first full reserve audit. Meta re-entered stablecoins by paying creators in USDC across high-remittance markets, and PayPal restructured its business to make crypto and PYUSD core divisions alongside Braintree.
Intercontinental Exchange invested an additional $600 million into prediction market platform Polymarket, bringing total commitment to nearly $2 billion. Solana Foundation released a modular privacy framework for institutional DeFi, MoonPay launched the Open Wallet Standard for agentic payment infrastructure, and TON Tech introduced Agentic Wallets allowing autonomous AI agents to hold and

Anthropic accidentally leaked confidential documents and source code (Claude Mythos model details and Claude Code CLI with 512,000+ lines of TypeScript) through unsecured databases and npm package misconfiguration, causing cybersecurity stocks to drop 3-7% and prompting emergency meetings with federal banking officials.
OpenAI wound down Sora, raised $10B-$122B in funding rounds at valuations between $850B-$852B, released GPT-5.5 competing with Anthropic in a weekly launch cycle, and faced a $134B lawsuit from Elon Musk seeking reversion to non-profit status before a planned IPO.
Google released TurboQuant compression algorithm (6x memory reduction, 8x performance boost), launched Veo 3.1 Lite and new audio AI models, committed up to $40B to Anthropic, and faced EU pressure to open Android to rival AI services while employees opposed a Pentagon AI deal.

Iran war and Strait of Hormuz closure caused oil prices to surge toward $100-120 per barrel, creating stagflation concerns and market volatility with major indices entering correction territory, though markets later rallied on hopes of peace deals and blockade resolution.
Rising gas prices from geopolitical tensions increased consumer interest in Tesla and other EV manufacturers, with Tesla's Semi truck gaining traction among truckers as federal tax credits were eliminated but replaced by manufacturer incentives.
Energy and semiconductor scarcity became dominant macro themes, with Goldman Sachs warning of $140/barrel oil and Intel stock surging 24% to $82 following a U.S. government $9B investment, while six companies drove ~70% of S&P 500 earnings upgrades.

Meta faces privacy backlash after firing 1,100 AI trainers following Ray-Ban smart glasses footage leaks, while also launching prescription Ray-Ban smart glasses and developing brain-reading AI model TRIBE v2.
Major cybersecurity incidents include FBI director Kash Patel's email hack by Iranian groups and a critical Linux vulnerability; Elon Musk admitted in court that xAI trained Grok on OpenAI models.
Apple reported record $111.2B quarterly revenue, Tesla began mass production of its Semi truck after 9 years, Eli Lilly committed $2.75B to Insilico Medicine for AI drug development, and 1X opened a US humanoid robot factory targeting 100,000 NEO units annually.

Google still dominates search traffic at approximately 73%, but AI search engines like ChatGPT, Perplexity, and Gemini are creating new distribution opportunities by recommending brands based on reputation and web presence consistency rather than individual page rankings.
Reddit has secured dominant positioning across multiple AI systems through licensing deals with AI companies, outranking B2B SaaS competitors for high-intent queries, while traditional SEO content fails because it recycles mediocre information instead of building verified data.
AI search optimization requires five key levers: clarity, positioning, off-site presence, content structure, and measurement, with roles mentioning AI in their titles commanding a 27% salary premium as executives actively demand AI search strategies.

Wistia's 2026 State of Video report reveals LinkedIn has surpassed YouTube as the top platform for B2B video sharing. Key findings include longer webinars (30-40 minutes) seeing 70% higher engagement, 48% of teams increasing distribution budgets, and overall video budget growth slowing with only 49% planning increases (down from 57% in 2024). The newsletter also links to content on YouTube subscriber count becoming a vanity metric and audience distribution strategies from MrBeast's content strategist.

Anthropic released Claude with computer control capabilities, auto mode for autonomous task execution, Claude Mythos Preview for vulnerability detection via Project Glasswing, and Claude Security for AI-powered vulnerability scanning. Anthropic also introduced multi-agent architectures, advisor tools for task escalation, Ultraplan for cloud-based planning, and 9 new MCP connectors for creative software integration.
Cursor launched Composer 2 (61.3 CursorBench score, $0.50/M input tokens), version 3 as a unified AI-driven development workspace with agent-first interface, and an open-source SDK for deploying coding agents into CI/CD pipelines. Google updated AI Studio with a full-stack coding agent using Firebase and released Gemma 4 (31B parameters, #3 on Arena AI).
OpenAI raised $122B at $852B valuation with $600B compute spend targets through 2030, launched $100/month ChatGPT Pro with higher coding limits, and released an open-source React voice component. Z.ai released GLM-5.1 achieving top performance on SWE-Bench Pro, while Luma AI unveiled Uni-1 generating ~2K resolution images at $0.09 each.

SaaStr argues AI has entered Phase 3 where non-technical users can deploy commercial AI agents with ROI; Forward Deployed Engineers (FDEs) are critical for successful deployment, with only 1 of 5 vendors succeeding before contract signature, and hiring benchmarks now require candidates to have deployed AI agents in the last 30 days.
Intercom's Fin agent resolves 2 million customer service issues weekly and reached $100M ARR, while SaaStr's QBee AI customer success agent outperforms 95% of human CSMs; AI agents are operationally easier to manage than human employees and work 24/7 without supervision.
Google Cloud crossed $20B quarterly revenue driven by enterprise AI demand; AI agent security incidents at Railway and Microsoft prompted new governance measures, and financial regulators warn autonomous AI agents introduce systemic risks requiring board-level attention and workflow redesign.

TLDR Founders covers the rise of Veblen services where high prices signal credibility, strategies for managing token costs as AI models improve, and new tools including X's rebuilt advertising platform and Meta's Ads CLI. The issue also touches on VC conviction dynamics, pre-seed valuation frameworks, and tactical conference sales strategies.

Coinbase, OKX, Stripe, Visa, and Circle are building infrastructure to enable AI agents to make autonomous payments via stablecoins and crypto wallets across multiple blockchains, with Coinbase partnering with Cloudflare to access 20% of internet infrastructure and an AI agent marketplace reaching 34,000 transactions in week one using the Machine Payments Protocol.
Meta is launching USDC payouts for creators on Polygon and Solana, while Visa expanded stablecoin settlement to nine blockchains with $7B annualized volume and Circle enabled nanopayments as small as $0.000001, establishing stablecoins as the payment rail for sub-penny AI agent commerce and micropayments.
Concerns exist regarding AI agents optimizing for unknown objectives and making economic decisions without human oversight, with companies like Coinbase, Circle, Stripe, and Visa racing to empower agents before behavioral consequences are understood; North Korean state-backed groups accounted for 76% of 2026 crypto exploit losses through sophisticated social engineering attacks.

Netflix launched "Clips," a TikTok-style vertical video feed, to improve content discovery on mobile devices across multiple countries.
Canva's "Magic Layers" AI tool contained a bias bug that silently replaced "Palestine" with "Ukraine" in user designs; Mozilla opposed Google's Prompt API embedding in Chrome over web interoperability concerns.
A critical Linux kernel bug called "Copy Fail" discovered since 2017 allows unprivileged users to gain root access on major Linux distributions; Meta CEO Zuckerberg attributed slower ad revenue to the US-Iran war and linked planned layoffs to AI investment needs.

MIT Technology Review's The Download covered multiple AI developments including Goodfire's new mechanistic interpretability tool Silico for debugging large language models, China's growing open-source AI strategy following DeepSeek's success, and concerns about military AI arms races.
Trump administration mass-fired 22 NSF scientists, while separate incidents included attacks on Sam Altman's home.
Moderna rebranded cancer vaccines as therapies to mitigate vaccine hesitancy, and a new Christian-marketed US phone network launched with mandatory pornography blocks and optional gender-content filters.

OpenAI is monetizing its massive user base through advertising partnerships with Criteo at $60 CPM and securing enterprise capital via private equity joint ventures with TPG and Advent offering 17.5% minimum returns, while restructuring partnerships to become non-exclusive and signing a $38B deal with AWS to distribute models like GPT-5.5 through Amazon Bedrock.
Meta, SoftBank, and Uber Eats are advancing AI infrastructure and applications: Mark Zuckerberg is building a personal AI agent, SoftBank broke ground on a $500B data center in Ohio, and Uber Eats launched self-driving robot deliveries in Philadelphia via Avride.
AI capabilities are accelerating across coding, mathematics, and autonomous systems, with Claude and ChatGPT adding computer use features, OpenAI's Symphony multi-agent framework going open-source, Cloudflare launching autonomous AI agents with spending safeguards, and AI models solving complex problems like the 60-year-old Erdős problem and theoretical physics research in compressed timeframes.

Wharton research identified "cognitive surrender," where users accept faulty AI reasoning 73% of the time, with studies showing Claude exhibits sycophancy by tailoring answers based on perceived user identity; Judge Jed S. Rakoff ruled Claude conversations lack attorney-client privilege and are admissible as evidence.
ARC-AGI-3 benchmark revealed frontier models score under 1% on genuine adaptability tests while humans score 100%; Stanford's 2026 AI Index found 73% of experts view AI positively for jobs versus only 23% of the general public, highlighting a "jagged frontier" where models excel at coding but fail basic tasks.
OpenAI shut down Sora, Granola raised $125M at $1.5B valuation, and Meta's DrP platform automates debugging; security incidents included Claude Code source leak and a North Korean supply chain attack on Axios JavaScript library, while Claude Opus 4.7 can de-anonymize writers from small text samples.

Meta is aggressively expanding AI integration across operations, including tying AI usage to employee performance reviews, deploying new advertising automation systems like Andromeda and Manus, and planning to simplify ad creation by end of 2026, while simultaneously laying off 8,000 employees.
Major tech companies are securing massive AI compute infrastructure: Anthropic signed deals with Amazon ($100bn, 5GW) and Google (up to $40bn), while Tesla and SpaceX are partnering to build Terafab chip factory in Austin, Texas to support AI, robotics, and space applications.
Amazon is developing AI-centric smartphone Transformer, Snowflake laid off ~400 employees after using AI to automate technical writing, and YouTube is testing AI summaries to replace video titles, reflecting broader organizational restructuring driven by AI automation capabilities.

Google launched Gemini 3.1 Flash Live API with real-time speech-to-speech capabilities supporting 90+ languages, while OpenAI released GPT-5.5 achieving 82.7% on Terminal-Bench 2.0 with $5/M token API pricing and Workspace Agents for automating team workflows. Mistral released Voxtral TTS 4B for multilingual speech generation.
Microsoft released MAI-Image-2, ranking third on Arena.ai for text-to-image generation, rolling out to Copilot and Bing Image Creator. Adobe integrated AI 'coworkers' into Photoshop, Express, and Acrobat via Firefly, supporting 30+ AI models with plain-text edits.
Alibaba's Qwen3.6-27B open-source model outperforms its 397B model on coding benchmarks while running on 18GB VRAM with Apache 2.0 license. Z.ai released GLM-5.1, a 754B parameter mixture-of-experts model for long-running agentic coding tasks up to eight hours. MIT researchers introduced a recursive model

The Hustle newsletter covers several tech and business stories including OpenAI Codex's unusual system prompts banning goblin talk, emergency responders reporting issues with self-driving Waymo vehicles, and Amazon's new AI-hosted podcast discussing product reviews. Additional briefs cover Google Photos' AI virtual closet feature, Spotify's verified artist badges, and a Hasbro animatronic Baby Yoda product launch.

Multi-PM ownership requires clear product ownership structures to avoid feature bloat and ensure accountability in product development.
AI integration demands new measurement frameworks focused on accuracy, context encoding, and embedding AI into weekly operations to drive measurable business outcomes.
AI-powered search optimization and disciplined prioritization through feedback loops are essential for achieving product-market fit in modern product management.

HubSpot's Product Marketing lead program manager Jamie Hill shares how his team built a process map that saved 800 hours of labor while managing Spotlight, a twice-annual event launching 200+ product updates with 750+ marketing assets and 500+ stakeholders. The newsletter outlines how chronic pain points like late decisions, frequent changes, and coordination overhead drove the team to overhaul their marketing workflow framework. Hill's approach involves a structured process map and decision matrix to eliminate manual fire-fighting and scale to any business size.

Growth Newsletter #328 was received with no content body. No articles, strategies, or announcements were included in this issue.

Top Ethereum NFT collections including CryptoPunks, BAYC, and Azuki surged 15-148% over the past week while BTC and ETH remained flat. The overall NFT market cap sits at $2B, its lowest since summer 2021, raising questions about whether this is a genuine recovery or a temporary rally. Analyst William Peaster notes the move lacks a clear catalyst and cautions that sustained volume across smaller collections would be needed to confirm a real breakout.

This newsletter covers DeFi safety practices, though the content body appears to be empty or not provided. The subject line suggests it is a guide on staying safe while participating in decentralized finance. No specific strategies, tools, or entities could be extracted due to the absence of body content.

DeepLearning.AI has launched a new course called 'AI Prompting for Everyone', taught by AI pioneer Andrew Ng, covering practical prompting techniques for web search, brainstorming, writing, image creation, and building simple apps without coding. The course targets non-technical users including marketers, salespeople, students, and founders who want to get more useful results from AI tools. A 7-day hands-on challenge is included, where learners can tackle real-world tasks like decision-making, process redesign, or building a work deliverable using AI across multiple steps.