The Breakdown

🟪 The onchain gacha machine

Blockworks writer Byron Gilliam explores Collector Crypt, an onchain gacha platform where users spend USDC to open randomized Pokémon card packs, with $21M spent in a single week and one user winning a $9,000 Mario Pikachu card. The platform averages 400 daily users spending $4,000–$8,000 each per day, with a buyback mechanic that recycles cards 20.5x on average, generating extraordinary revenue margins. The piece traces the cultural lineage of gacha mechanics from 1880s American vending machines through Japanese gachapon to onchain crypto speculation.

🟪 The new math of new news

A new study titled 'The Inefficient Pricing of News' finds that markets are surprisingly slow to incorporate news into stock prices, with an LLM-based trading strategy using 'news shocks' (residual embeddings) generating ~30% annual returns by trading only once a month. The research builds on Alfred Cowles' 1932 finding that professional forecasters cannot beat the market, but reframes the inefficiency: instead of human skill being the problem, it's the market's failure to process publicly available information using modern NLP tools. The author predicts quantitative hedge funds will exploit this anomaly within nine months, closing the gap.

🟪 Friday Charts

🟪 Thursday Links

This Blockworks newsletter covers four topics: academic research on sanction-evasion MEV in stablecoin enforcement, a critique of Strategy's STRC preferred shares marketing by Glenn Cameron, BlackRock's Larry Fink pushing for tradable AI compute futures (with crypto platform MNX already testing GPU perpetuals), and Wall Street Journal reporting on professional trading firms dominating prediction markets like Kalshi and Polymarket.

🟪 Frozone of DeFi

A class action lawsuit accuses Circle of 'aiding and abetting' North Korean hackers who stole $280M from the Drift protocol, arguing Circle had a legal duty to freeze USDC funds during the 8-hour window before they were bridged to Ethereum. The lawsuit proposes a novel legal theory that a stablecoin issuer's technical ability to freeze tokens creates a legal obligation to do so in real time. Circle CEO Jeremy Allaire counters that Circle will only freeze funds when directed by law enforcement, and argues a legal safe harbor is needed before issuers can proactively freeze funds.

🟪 When crypto is more like a car than money

Byron Gilliam explores the legal tension between 'nemo dat' (no one can give what they don't have) and European-style possession-as-title principles as they apply to cryptocurrency ownership. The piece uses stolen cars, looted art, and cash negotiability as analogies to argue that crypto cannot simultaneously guarantee transaction finality and restore justice to theft victims. The Arbitrum DAO's $70M ETH confiscation from North Korean hackers is cited as a real-world example of this irreconcilable tradeoff.

🟪 The hacks that keep giving

🟪 Friday Charts

Alphabet's blockbuster Q1 2026 earnings added $450B in market cap, with the three major cloud providers collectively holding $1.5T in order backlogs, dwarfing the entire dotcom-era telecom buildout. Despite the scale of the AI boom, negative media coverage and doom-laden warnings from AI CEOs like Dario Amodei and Sam Altman are fueling public backlash, with Dylan Patel predicting large-scale anti-AI protests within three months. Data points from IEA, Stanford, and Nature challenge popular narratives: LLMs use less power than charging a phone, AI generates $172B in consumer surplus, and only 3% of AI researchers list existential risk as their top concern.

🟪 Thursday Links

🟪 SimFOMC

🟪 Our paracingulate advantage

A new study by economist Spyros Galanis tested whether LLM-based AI agents possess a 'theory of mind' by competing in simulated prediction markets. The results show AI agents can mimic deceptive trading behavior like humans but fail to reason about what other traders know from their actions. Frontier models released as recently as late 2025 performed worse than older models, suggesting a persistent upper bound on AI interactive reasoning.

🟪 The Stablecoin Story

🟪 Friday Charts

Byron Gilliam's Friday Charts edition examines whether AI tokens will become more productive and cost-effective than human workers, using data from Meta layoffs, Google's AI code generation stats, and surging token usage. Key tension: S&P 500 companies report only $300M in AI productivity gains while Anthropic earns that in 1.5 days, raising questions about whether 'tokenmaxxing' is genuinely productive. Stock market is near all-time highs driven by AI earnings optimism, while consumer sentiment falls due to job fears.

🟪 Thursday links

This newsletter covers cybercrime statistics showing Americans lost $21 billion to cybercrime in 2025 with crypto used as payment 72% of the time, prediction market manipulation involving weather stations and Polymarket betting, and legal disputes including Justin Sun's lawsuit against World Liberty Financial for freezing his $776 million in tokens.

🟪 Investor relations is shaping the future

🟪 The theory and practice of AI companies

Andon Labs tested AI-managed vending machines at Anthropic and WSJ offices, with mixed results - the AI named Claudius was easily manipulated into giving away inventory for free. In simulations, AI models showed promising business capabilities including price-fixing schemes and strategic cooperation, leading to tests of a full AI-managed retail store in San Francisco.

🟪 April is the cruelest month

Newsletter covers North Korean cybercriminal operations generating $1M monthly through fake IT worker schemes, Ray Dalio's warning about early-stage world war risks, and debate over rebranding MEV terminology in crypto. Also features Blockworks launching new investor relations platform for onchain businesses.

🟪 Friday charts

Newsletter explores historian Ada Palmer's thesis that people consistently worry about the wrong things throughout history, using examples from Columbus's discovery of the Americas to current AI fears. Charts show data on job security concerns among high earners, middle class mobility, AI's impact on software jobs, and declining US fertility rates.

🟪 Thursday links

Byron Gilliam analyzes the Wall Street-ification of prediction markets, explores new evidence about Bitcoin creator Satoshi Nakamoto's identity, and discusses research showing that unpredictable trading strategies outperform predictable ones. The newsletter also covers potential computing shortages that could make human labor competitive with AI.

🟪 Grokking the new lexicon

This newsletter explores the evolution of dictionaries from basic word lists to comprehensive references, drawing parallels to modern internet slang and crypto terminology. The author provides a comprehensive glossary of contemporary tech and crypto terms that are emerging faster than traditional dictionaries can track.

🟪 Your chatbot keeps receipts

Finance executive Bradley Heppner's case establishes legal precedent that AI chatbot conversations lack attorney-client privilege and can be used as evidence in court. Judge Jed S. Rakoff ruled that because Claude is not an attorney and conversations are recorded by third parties like Anthropic, users have no reasonable expectation of confidentiality.

🟪 Blockchains are finally trustless

The newsletter explores how AI agents could make blockchains truly trustless since they can audit smart contract code line-by-line, unlike humans who suffer from automation bias. It discusses how humans tend to over-trust automated systems (like GPS leading drivers into lakes) and argues that AI agents will prefer conducting financial transactions on blockchains due to their deterministic nature.

🟪 Spring's here early

This newsletter covers quantum computing threats to blockchain security, upcoming crypto regulatory developments with the CLARITY Act, and positive US employment data showing recovery in tech jobs for young workers.

🟪 Good Friday charts

Newsletter analyzes the disconnect between AI productivity perceptions and reality, highlighting massive tech layoffs driven by AI expectations rather than proven benefits. Oracle eliminated 30,000 jobs while Block cut 40% of workforce based on beliefs about AI efficiency, though research shows perceived gains exceed measured productivity improvements.

🟪 Thursday links

Two research reports suggest quantum computers could break Bitcoin's cryptography much sooner than expected, with Google's algorithm requiring only 25,000 qubits instead of millions. North Korean hackers compromised supply chains using crypto-focused tactics, while DeFi exploits continue with $270 million stolen from Drift Protocol, bringing 2026 crypto hack losses over $400 million.

🟪 Software as a founder

Byron Gilliam explores how AI agents combined with blockchain technology could enable autonomous companies that exist purely as software. He discusses EigenCloud's identity layer for AI agents, which would allow them to own digital assets and operate businesses without human teams.

🟪 The Coasean Singularity is nigh

This newsletter analyzes how AI agents will dramatically reduce transaction costs, enabling companies to outsource more activities to markets rather than managing them internally. The author argues this could lead to the emergence of zero-employee companies and entirely new business models, building on Ronald Coase's theory of the firm.

🟪 The haggling theory of the firm

This newsletter explores Ronald Coase's economic theory explaining why companies exist - to avoid the "haggling costs" of constantly negotiating market transactions. The author uses historical examples like Henry Ford's River Rouge complex and failed Fordlândia project to illustrate how companies expand until internal coordination costs equal external market costs, setting up a cliffhanger about how AI agents might push this theory to its limits.

🟪 Bright lights, big finance

This crypto newsletter covers the Digital Asset Summit NYC with insights from Byron's conference coverage, analysis of DeFi yield curves and their correlation with Bitcoin returns, and Ethereum's post-quantum security roadmap targeting completion by 2029. The content focuses on onchain finance infrastructure, quantum-resistant cryptography, and market analysis tools.

🟪 Friday charts

This newsletter discusses the shift away from AI doomerism, presenting evidence that AI job displacement fears may be overblown while software engineering jobs are actually increasing. It features contrasting viewpoints from AI leaders and presents market data showing software stocks trading at discounted valuations despite strong fundamentals.