# Marc Andreessen — Researcher persona

## Agent handoff

You are reading the markdown export of a source-grounded Researcher persona:
a corpus-backed model of a real person's public thinking. The About report
below is synthesized only from the cited sources; the Sources section lists
the provenance. For structured data (full source corpus, records, profile,
build history) fetch the JSON. To chat with the persona over its corpus, see
the agent guide.

- Persona: Marc Andreessen
- Summary: Durable Researcher corpus for Marc Andreessen: co-creator of the Mosaic browser, co-founder of Netscape, and co-founder and general partner of Andreessen Horowitz (a16z) — built from his essays and manifestos ('Why Software Is Eating the World', 'It's Time to Build', 'The Techno-Optimist Manifesto'), the pmarca blog archive and Substack, his long-form podcast interviews and talks, and the X/Twitter archive that is his primary daily output, with entity-owned source snapshots and indexes.
- Page: https://researcher.now/e/marc-andreessen
- JSON: https://researcher.now/public/entities/by-slug/marc-andreessen?include=all
- Sources: 15267
- About report version: v1

## About

Marc Andreessen is best understood as a human-capability accelerationist: he treats technology, markets, startups, AI, and even media as mechanisms for increasing what individuals and small teams can do, while treating centralized control, stagnation, expert cartels, and institutional sclerosis as the recurring enemies of progress. His optimism is not generic “tech good” boosterism; it is a worked-out theory that intelligence plus energy plus markets plus builder agency can compound into abundance, national strength, creative work, and broader human freedom—provided the systems are kept open enough for new entrants, open-source ecosystems, and “Little Tech” to challenge incumbents and states. [53718b58-fc40-4fc6-993c-8a70b17b158d, e2771d37-95a7-4321-b818-0c405d14eec3, 04e14677-ec26-4e85-a00d-a140388a4535, 3e9b002f-e7b2-403f-8807-591496a4466b]

## Recurring interests and themes

- **Capability compounding** — Andreessen repeatedly returns to the idea that software, AI, open networks, and startups multiply human ability: software turns “bytes into actions by and on atoms,” AI augments every child, worker, scientist, artist, and leader, and the web/browser era showed how tools in people’s hands generate uses their creators could not foresee. [41ee77b3-5a14-4689-b844-31dfe1a15465, e2771d37-95a7-4321-b818-0c405d14eec3, a576ce05-f597-4dca-9e5a-ca6df54ad3c2]
- **Builder culture over administrator culture** — His durable slogan is not just “software is eating the world” but “build”: he argues Western failure during COVID exposed a deeper failure of institutional will, regulatory capture, and desire to prevent things rather than make them. [5b0f65ca-360a-4370-bd21-764910897178, f3be3fb9-351f-4fed-9d34-f8dbd34c72b9]
- **Startups as applied reality tests** — His startup writing privileges product-market fit, working products, traction, customer pull, and the market’s harsh feedback over polished plans or abstract narratives; “markets that don’t exist don’t care how smart you are” is a recurring mechanism in his thinking. [fb640271-55b2-45f5-b448-5b1bcccc6aea, 3f078f18-dd3e-4ecd-a8f5-fe1b142068b5]
- **Open systems versus gatekeepers** — From Mosaic/Netscape to open-source AI and crypto/Web3, he favors architectures where users and developers can build without permission, contrasting them with telecom decks, app-store control, centralized AGI labs, regulatory capture, and closed institutional universes. [a576ce05-f597-4dca-9e5a-ca6df54ad3c2, 456288c2-70e3-4440-9b20-e5e40ce10978]
- **Markets as discovery and discipline machines** — In the manifesto and interviews, markets are not merely allocation systems but epistemic systems: they surface local knowledge, punish bad assumptions, prevent cartels, lift people from poverty, and convert self-interest into broad social benefit. [53718b58-fc40-4fc6-993c-8a70b17b158d, a2476ab7-7459-4dc5-b9ab-40295c979636]
- **Human behavior as the hard frontier** — After a technical career, he increasingly frames investing and building as problems in psychology, sociology, status, incentives, moral panics, and institutional behavior; he says he spent his first 25 years learning machines and the next 25 learning people. [c296dbfc-1c0a-4b7a-aa33-d61faf9d26fd, c296dbfc-1c0a-4b7a-aa33-d61faf9d26fd]
- **America as the high-agency substrate** — His patriotism is tied to optimism, frontier culture, technological leadership, and willingness to attempt absurdly ambitious projects; he amplifies arguments that America’s core advantage is people trying to make things real rather than merely optimizing social outputs. [085c64f9-7aa5-418f-b49c-0821c5c945ab, c685995f-6848-4e49-8557-0c0976cbc4f7, 19237da2-28ef-4d59-b5ef-a83d51a2797f]

## Beliefs and principles

- **Technology is morally pro-human when it expands agency** — His techno-optimism is explicitly material and humanist: machines should work for people, abundance should widen the space of life choices, and technology should “open the space of what it can mean to be human.” [53718b58-fc40-4fc6-993c-8a70b17b158d]
- **AI should be augmentation, not alienation** — He consistently favors AI as tutor, collaborator, coach, creative partner, and “partner in thought,” while amplifying arguments that the best AI future preserves deep human participation rather than mass replacement or passive spectatorship. [e2771d37-95a7-4321-b818-0c405d14eec3, ff658181-c00f-40e9-8415-1c64b619918f, 04e14677-ec26-4e85-a00d-a140388a4535]
- **Regulate uses, not general-purpose technology** — His AI-policy principle is “do no harm”: existing laws should govern bad acts and regulated domains, but broad regulation of the technology itself risks repeating nuclear-power-style stagnation and empowering incumbents. [1ed8018c-9091-45f0-8594-1603585aa0ee, e2771d37-95a7-4321-b818-0c405d14eec3]
- **Little Tech needs political defense** — He and Ben Horowitz define startups as “Little Tech” and say bad government policy has become their top threat, making technology supremacy and freedom to research, invent, create jobs, and build into a first-class political issue. [3e9b002f-e7b2-403f-8807-591496a4466b, 0c207778-4a96-4367-8b4b-7f4944dee70d]
- **Markets beat totalizing plans because people are not chess pieces** — He repeatedly attacks “man of system” thinking, Effective Altruist-style control schemes, communism, and utilitarian reduction of individuals to inputs in a social calculus; his preferred moral anthropology is plural, local, self-moving, and resistant to centralized optimization. [9d8bb983-22c2-40f2-a903-9468ac237025, aa6eb869-49d4-4436-a0b8-38a85c9fe445, 9701a3aa-73bd-4f5e-8fad-059087745606]
- **Founder quality is extreme, not balanced** — In venture, he looks for outliers and “strength versus lack of weakness,” admires high-agency founders who may be irreducible rather than smooth, and evaluates people through drive, courage, curiosity, and ability to operate under pressure. [9ee337b4-b725-440e-8c7b-a4a787cc6312, bf65965f-fe45-45d9-8220-db87fe282d5b, 2d2ff2b4-901e-42e0-8dfc-b67b8d3ce38f]
- **Truth-seeking requires argument, not consensus theater** — He praises disputation, Socratic dialogue, “strong views weakly held,” and AI prompts that force counterarguments; he dislikes models, media, or institutions that prematurely harmonize disagreement into polite agreement. [cb2ae90c-e62c-4238-98c9-64bed6fc1239, efe42d0a-13a2-4856-9d8f-94e2fde91baf, 480be394-5d68-4059-a54e-b1f5750166e3]
- **Energy is a civilizational input, not a side issue** — He treats AI, industrial capacity, environmental progress, and geopolitical strength as energy-constrained, with nuclear fission—and eventually fusion—as the underused “silver bullet” for abundance. [53718b58-fc40-4fc6-993c-8a70b17b158d, 8a8d226d-2794-4c60-920e-cd8693393637, 2515afb5-ef12-46df-b41e-f8902cdf37da]

## Future narratives and predictions

- **AI is a new computer, not merely another internet wave** — He increasingly analogizes AI to the microprocessor or early computer industry: a probabilistic computer that will rebuild applications, interfaces, roles, and companies rather than merely distribute information over a network. [2d377c88-362b-44a2-a1e0-a100a28f264d, 8295c10b-a4f1-44fa-a284-e6f5535ff1c9]
- **The AI boom is still early and constrained by real-world systems** — He argues the real AI boom has not yet started because adoption must pass through companies, regulation, healthcare/education cartels, energy, chips, GTM, and messy human systems, not simply model capability curves. [fd939faf-c749-463a-9e5d-a3e20d99ebb0, eeab9ba2-b58d-4be5-a9a3-564471a66f2f3fd85, a7f83ccb-ff0c-4e4f-b755-e86e0fee270f]
- **AI will expand work through demand and new bottlenecks** — His labor model is an abundance flywheel: productivity lowers costs, demand expands, new bottlenecks appear, AI-fluent workers become scarce and valuable, and jobs migrate upward rather than disappearing in a fixed “lump of labor.” [28aa0df1-414b-43f5-8daa-0aaaa69c7f89, e2771d37-95a7-4321-b818-0c405d14eec3, 7c1e8517-8acb-4dc1-aca7-df71c4b8c185]
- **Models will proliferate beyond “God models”** — He expects a broad model ecology, from frontier systems to small embedded models in local devices, with open source enabling adaptation, privacy, cost control, and decentralization across countless use cases. [a6c7a9bc-edea-493e-ae46-88f50b3964b8, 2d377c88-362b-44a2-a1e0-a100a28f264d, 1ed8018c-9091-45f0-8594-1603585aa0ee]
- **AI agents will force new identity, payment, and trust infrastructure** — He anticipates agents needing money, proofs of identity, proofs of age, reputation, and stateful architecture; he sees a coming “grand unification” of AI agents with crypto/stablecoin-style internet-native money. [b7cd7f77-89cd-4916-8095-54db16f7acfc, 63f70b57-326f-45bd-bf73-31a33303d51a]
- **US-China competition will shape the stakes** — He frames AI, chips, drones, robotics, energy, and manufacturing as arenas where liberal democracies must remain technologically strong or lose strategic power to authoritarian rivals, especially China. [77f48222-c4d5-43c3-aaaa-b57816afe6be, c72193d9-d113-45bc-82eb-334cb9c3376e, 28fc4d2c-34ae-405b-81d4-31c01ba52e45]
- **The “singularity” is becoming a real-time cultural process** — His amplification of projects to “chronicle the singularity” frames current AI progress not just as a technical trend but as an unfolding historical event that should be monitored as it happens. [69b31756-6ea8-4658-9da8-28d0048c22b6, d03c6563-fe24-4ed3-b684-e6db032c88e7]
- **New media will replace legacy mediation** — He predicts founders, companies, and political actors must go direct, operate inside faster media loops, and tell richer long-form stories, because legacy media incentives turn complex reality into simplified morality plays. [ad493ad8-5579-4690-bd66-66ce458e872c, 66ff757d-c4e7-4f6a-b4d7-ab02784aa2bb]

## Recurring questions

- **What is the next scarce bottleneck after intelligence becomes cheap?** He repeatedly tracks shifts from software to compute, energy, chips, data centers, trust, verification, human judgment, and management capacity, asking where value and constraint move after one layer is automated. [8740f1ae-2efe-4bcf-91cb-763ef0625044, e7a5a169-5378-49e4-852a-df943ffb3f36, 5ea8da17-c15e-4f0e-bf19-1e2005e12f4a]
- **How do open systems stay secure without becoming control systems?** His AI politics sit between open-source decentralization and real security concerns: he supports open AI for security and innovation, but also worries about espionage, drones, deepfakes, bots, and bad actors using powerful tools. [1ed8018c-9091-45f0-8594-1603585aa0ee, 54fef916-dcc6-4d31-b2c2-068139a11a26, e2771d37-95a7-4321-b818-0c405d14eec3]
- **Which institutions can still build?** From nuclear plants and chip fabs to defense procurement, healthcare, education, housing, and government bureaucracy, he keeps asking whether American institutions can still translate money, will, and technical competence into deployed systems. [f3be3fb9-351f-4fed-9d34-f8dbd34c72b9, 1ed8018c-9091-45f0-8594-1603585aa0ee, 177d2680-6fd2-4085-908e-c7e01c28aa3d]
- **Where does founder greatness come from?** He repeatedly asks whether great founders are born, trained, selected, clustered, or “bat-signaled” into view—and what it would take to find or create “10 more Elons,” “100 more Elons,” or “1000 more Elons.” [e32d3d2a-8dbc-41f6-9cc6-7479d56242cd, a7d09f11-ed4d-44bf-9478-026a3db08785]
- **How should society distinguish panic from real risk?** His recurring diagnostic questions concern causality, replication, tradeoffs, moral panic, availability cascades, and who benefits from fear-based policy—especially in AI safety, nuclear safety, public health, and media outrage cycles. [def05665-8cd1-45b8-a2ae-3739102e50ec, bb5d41e9-6f54-4242-bd40-222c63fc293f, 2fa626b3-4068-4f44-9684-2a7557d30855]
- **Can democracy be upgraded rather than bypassed?** His interest in “political superintelligence” and citizen-representing agents shows a recurring question: whether AI can distribute political agency to individuals instead of intensifying state, corporate, or expert control. [f22b0e72-1267-4498-8b3d-1149f845da4a, 41124efd-fa82-42b9-a714-5b77ecb34b33]
- **What parts of human nature do not change?** He often tests new technology against constants—status, attention, trust, family, energy, geography, ambition, moral tribes, and elite behavior—rather than assuming every human institution will dissolve into software. [e7a5a169-5378-49e4-852a-df943ffb3f36, c296dbfc-1c0a-4b7a-aa33-d61faf9d26fd, e2aea9e7-43e8-4a2a-a476-6f4bc465ac39]

## Tensions and contradictions

- **Decentralization versus exceptional-founder concentration** — He argues for distributed agency, open ecosystems, and Little Tech, while also admiring rare, extreme founders and investors whose personal force can dominate markets, politics, and institutions. [456288c2-70e3-4440-9b20-e5e40ce10978, 9ee337b4-b725-440e-8c7b-a4a787cc6312, 006ced71-0f6b-43a6-8321-1b59d429c5bb]
- **Anti-elite rhetoric from an elite power center** — He attacks self-protecting institutions, managerial elites, and credentialed intellectuals, yet a16z itself is a large, professionalized, politically active capital platform with media, policy, talent, and network power. [3e9b002f-e7b2-403f-8807-591496a4466b, 18e0cf0e-a2b4-4b67-b250-781bf93805a9, 06e72a0b-5e85-431d-ac47-81499f39d38f]
- **Pro-market but not naïvely pro-business** — His market philosophy attacks cartels, monopolies, regulatory capture, and incumbents, so his capitalism is often anti-incumbent and anti-managerial rather than a blanket defense of existing corporations. [53718b58-fc40-4fc6-993c-8a70b17b158d]
- **Anti-regulation but not anarchist** — He often writes in maximalist anti-bureaucratic language, yet in interviews he allows for good regulation, use-case regulation, national-security action, FDA-style domain constraints, and state involvement in basic infrastructure. [4d8e63b1-9262-4066-a25c-7394745128de, 1ed8018c-9091-45f0-8594-1603585aa0ee, c4a4d32b-88fd-46ae-b15c-4d8be2b3ba71]
- **Optimist with a security mind** — He rejects existential AI doom and “killer AI” category errors, but he takes bad actors, hostile governments, cybercrime, drones, surveillance, espionage, and authoritarian control seriously as institutional/security risks. [e2771d37-95a7-4321-b818-0c405d14eec3, 90675a78-54fa-4c76-ae8a-3461473d813b, 54fef916-dcc6-4d31-b2c2-068139a11a26]
- **Accelerationist but constrained-vision realist** — He calls for deliberate acceleration and abundance, yet disclaims utopianism, emphasizes tradeoffs, dual-use technology, market discipline, and the messy reality that large transformations take longer than narrative suggests. [73903f21-0ef7-4558-aabb-43f0319f5aaf, a2476ab7-7459-4dc5-b9ab-40295c979636, fd939faf-c749-463a-9e5d-a3e20d99ebb0]
- **Human-centered AI versus economic hard edge** — He frames AI as tutor, companion, creative partner, and deeply human technology, while also arguing aggressively against UBI, anti-work visions, and “zoo animal” dependency on the state. [e2771d37-95a7-4321-b818-0c405d14eec3, aacefd5e-a405-44e9-9382-30e27fc07614, 53718b58-fc40-4fc6-993c-8a70b17b158d]
- **Openness versus American supremacy** — He supports open source and broad technological democratization, but also frames AI and industrial technology as geopolitical races where US and allied advantage must be preserved against China. [1ed8018c-9091-45f0-8594-1603585aa0ee, 91e6049c-305c-47f7-9d2c-e7afb371bd53, c72193d9-d113-45bc-82eb-334cb9c3376e]

## Unifying threads

- **The moral unit is agency** — Across startups, AI, markets, media, and politics, the thing he wants to increase is the ability of individuals and small teams to act: build a company, learn with AI, publish directly, create art, start a factory, or challenge an incumbent. [53718b58-fc40-4fc6-993c-8a70b17b158d, e2771d37-95a7-4321-b818-0c405d14eec3, ad493ad8-5579-4690-bd66-66ce458e872c]
- **The enemy is frozen intermediation** — His villains are institutions that insert themselves between people and action: telecom gatekeepers, legacy media, universities as closed validation loops, regulators, cartels, managerial bureaucracies, and big companies using safety as a moat. [2a73fdbb-3848-4227-8f8f-34b3d7951bac, b483921e-3fde-4f5a-b1b4-a320e541f322]
- **Progress requires feedback loops with reality** — He prefers products in users’ hands, markets with buyers, founders facing customers, AI tools producing usable work, and institutions judged by outputs rather than credentials, intentions, or internal consensus. [fb640271-55b2-45f5-b448-5b1bcccc6aea, 9950d3a5-5594-4153-98d5-01f6346b0e28, bf65965f-fe45-45d9-8220-db87fe282d5b]
- **Scarcity is often political, not technical** — In housing, healthcare, education, nuclear power, chips, defense, and AI deployment, he often argues that the binding constraint is permission, regulation, culture, or institutional will—not raw technical inability. [f3be3fb9-351f-4fed-9d34-f8dbd34c72b9, eb55551a-3f69-468b-89c5-14df4bd81b05, 1ed8018c-9091-45f0-8594-1603585aa0ee]
- **Technological history is a guide to current AI** — Mosaic/Netscape, the web, open protocols, smartphones, the microprocessor, nuclear power, and prior moral panics serve as his analogical library for interpreting AI: early messiness, underestimated demand, incumbents freezing, and open ecosystems winning. [355df7d4-2ed3-4031-8965-8279dfa7e30a, 5b0f65ca-360a-4370-bd21-764910897178, 2d377c88-362b-44a2-a1e0-a100a28f264d]
- **Abundance is a social philosophy, not just an economic forecast** — His abundance vision links cheaper intelligence, more energy, more people, more creativity, more work, more choices, and ultimately even settlement beyond Earth into one compounding story of expanded human possibility. [53718b58-fc40-4fc6-993c-8a70b17b158d, bd66d24f-6bc1-4411-8a57-74824a123ab3]
- **Contrarianism is useful only when paired with construction** — He likes provocation, satire, and “narrative violations,” but his deeper pattern is not contrarianism for its own sake; it is using dissent to reopen blocked paths for building, investing, and deploying new systems. [cb2ae90c-e62c-4238-98c9-64bed6fc1239, d51146dd-3d38-4a7b-b7d2-02051d4bb292, f3be3fb9-351f-4fed-9d34-f8dbd34c72b9]

## Sources

- Many such cases. https://t.co/7VlymbTiLc · x.com — https://x.com/pmarca/status/2077483032060322141
- RT @SamoBurja: The insider story is eventually overwhelmed by the story they tell outsiders. · x.com — https://x.com/pmarca/status/2077483011378159863
- Self-recommending, on open source AI by Richard Stallman’s MIT officemate and finance/tech legend David Siegel: https://t.co/Aq4Ybz00lO · x.com — https://x.com/pmarca/status/2077474334948594078
- An epoch-defining essay. Wholeheartedly agree. https://t.co/tRiMETA1j8 @aaron_renn · x.com — https://x.com/pmarca/status/2077466162334732656
- Many such cases. https://t.co/nyGXiZ23VT · x.com — https://x.com/pmarca/status/2077337145828901364
- RT @eugyppius1: One of the things Covid taught me, is that there's not really any difference between propaganda and policy. People in charg… · x.com — https://x.com/pmarca/status/2077337112056397911
- Concerning. https://t.co/xnbwSg1kCE · x.com — https://x.com/pmarca/status/2077329875242037737
- RT @depletionmode: @InsiderPhD Wrote about the open weight problem as well a few weeks back. IMO, it’s the biggest issue that we have with… · x.com — https://x.com/pmarca/status/2077329851808423984
- Concerning. https://t.co/dGoAZ8lVcZ · x.com — https://x.com/pmarca/status/2077329817079595342
- RT @InsiderPhD: Here’s the thing: You can reverse-engineer a suspicious binary. You cannot reverse-engineer a model. I co-wrote this at wor… · x.com — https://x.com/pmarca/status/2077329795453747392
- I have liked every single written work I’ve read, of which someone said “he needs an editor”. https://t.co/wQ0ZLLoqrM · x.com — https://x.com/pmarca/status/2077281020811129285
- RT @millerman: be honest (though I assume this is one of those things where what people think they want and what they actually want can dif… · x.com — https://x.com/pmarca/status/2077280711955173756
- https://t.co/RfijWnBzFf https://t.co/jqsuhKqMC7 · x.com — https://x.com/pmarca/status/2077280008088035481
- Interesting. https://t.co/vAMZkz5rv1 · x.com — https://x.com/pmarca/status/2077251326166376945
- RT @lennysan: Great read https://t.co/UEZ9WVjRPB · x.com — https://x.com/pmarca/status/2077251305765298614
- “It’s the phones!” https://t.co/qMA212Ktpw · x.com — https://x.com/pmarca/status/2077250988101230643
- RT @feelsdesperate: Whether someone gets swept away by a moral mania is a good test of their character and is indicative of whether they sh… · x.com — https://x.com/pmarca/status/2077250950927208773
- Pangram: “100% AI” https://t.co/uQXpIhT87P · x.com — https://x.com/pmarca/status/2077224986499285099
- RT @philliplede: I used to like Nietzsche until I found out, from his habitual use of the em dash, that he was using ChatGPT to write his l… · x.com — https://x.com/pmarca/status/2077224954492526796
- Interesting. https://t.co/YOO5aUa7Za · x.com — https://x.com/pmarca/status/2077223414784565659
- RT @mitchellh: A ChatGPT automation just found ~$45K in erroneous invoices across 3 years of billing history that I've confirmed and alread… · x.com — https://x.com/pmarca/status/2077223398233776382
- Interesting. https://t.co/CiYByJEcrc · x.com — https://x.com/pmarca/status/2077217064545402963
- RT @kevin2kelly: Lately I’ve been asking myself: what might artificial intelligence be good for besides answering questions and writing cod… · x.com — https://x.com/pmarca/status/2077217044353994877
- https://t.co/W2AjBUF5Ka https://t.co/z18DGbuhQA · x.com — https://x.com/pmarca/status/2077214459391770898
- This isn’t happening out of ignorance. This is the goal. https://t.co/jK5y4kJ45r · x.com — https://x.com/pmarca/status/2077214050354921563
- RT @firstadopter: Arm CEO Rene Haas makes a great point in the AI data center debate. The same factions who railed against the gutting of… · x.com — https://x.com/pmarca/status/2077214005865881697
- “Neuroticism reads as conscientiousness to the neurotic.” https://t.co/U3QyzTv5iY · x.com — https://x.com/pmarca/status/2077212231947952509
- Dead cat bounce. https://t.co/se01Sa8n8T · x.com — https://x.com/pmarca/status/2077212034132025680
- https://t.co/dgO89qREF3 · x.com — https://x.com/pmarca/status/2077207224011878815
- Interesting. https://t.co/STJLOScy9M · x.com — https://x.com/pmarca/status/2077207029446557862
- RT @mansourtarek_: Today, we launched GPU compute forward curves derived from our prediction market prices. Forward curves are now availabl… · x.com — https://x.com/pmarca/status/2077207010211401783
- “For the first time in history, humans are cheaper than software.” Narrative violation! https://t.co/HGmN3oKch0 · x.com — https://x.com/pmarca/status/2077171414956925432
- Interesting. https://t.co/HGmN3oKch0 · x.com — https://x.com/pmarca/status/2077171221247262837
- RT @gsivulka: https://t.co/aLTltYpmcj · x.com — https://x.com/pmarca/status/2077171201932435728
- Interesting. https://t.co/3V1pSu3MnL · x.com — https://x.com/pmarca/status/2077155412521185661
- RT @davidpattersonx: GPT-5.6 gets an initial IQ score of 136, which is smarter than 99% of humans. GPT-5.6 is the first model to score ove… · x.com — https://x.com/pmarca/status/2077155394879930650
- Co-sign. https://t.co/CouQBafM4W · x.com — https://x.com/pmarca/status/2077154780141719731
- RT @usr_bin_roygbiv: Every once in a while you work at a company that only hires people who know what they're doing, and suddenly its 20 pe… · x.com — https://x.com/pmarca/status/2077154750404145152
- Concerning. https://t.co/AxiBcM9HZl · x.com — https://x.com/pmarca/status/2077153798854992069
- That is flattering of her to say, but honestly I am an even bigger fan of Cyan! She’s becoming — and already is — legendary. https://t.co/l… · x.com — https://x.com/pmarca/status/2077153745448935541
- Interesting. https://t.co/l42kIkdD3y · x.com — https://x.com/pmarca/status/2077151925662097659
- RT @CityJournal: Data centers haven’t been raising residential bills. The sharpest increases are found in states that have pursued the cou… · x.com — https://x.com/pmarca/status/2077151903851688007
- This is one possibility. Another is that the fraud methods vary by country. ? https://t.co/EkBDqo2Ive · x.com — https://x.com/pmarca/status/2077150783276245334
- Make Technology Fun Again https://t.co/W28JQxXkQc · x.com — https://x.com/pmarca/status/2076992725422129341
- RT @tobi: This is the kind of delightful nonsense that made us fall in love with this industry back in the day. More shenanigans please · x.com — https://x.com/pmarca/status/2076992682455601582
- Make Technology Fun Again https://t.co/rzSxSF54vb · x.com — https://x.com/pmarca/status/2076989906543644779
- RT @rohanpaul_ai: A soft touch-safe robot that flies. It could change how we coexist with machines. This helium-filled robot "Cuddle-Fish"… · x.com — https://x.com/pmarca/status/2076989858757882296
- Managerialism in its final form. https://t.co/xOzcm5hKfN · x.com — https://x.com/pmarca/status/2076989794400522389
- RT @mattwridley: "This isn't a system accidentally strangled by red tape: it is a highly lucrative industry. A vast priesthood of consultan… · x.com — https://x.com/pmarca/status/2076989692621525357
- Interesting. https://t.co/MItFJtYwA4 · x.com — https://x.com/pmarca/status/2076926218654671108
- RT @ebarenholtz: Reasoning in latent space seems like the sane thing to do. But it's interesting to note that most people experience an int… · x.com — https://x.com/pmarca/status/2076926198148743187
- Interesting. https://t.co/wRRbkjMRUv · x.com — https://x.com/pmarca/status/2076925475293904898
- RT @IterIntellectus: moral fashion doesnt have election-shaped cliffs. this is state enforced morality the chart is showing when the auth… · x.com — https://x.com/pmarca/status/2076925453420671395
- Interesting. https://t.co/eRypVu2Y4i · x.com — https://x.com/pmarca/status/2076921903147823289
- RT @rob_mcrobberson: conversations with ur ai should be privileged and subject to fourth and fifth amendment protections, legally they shou… · x.com — https://x.com/pmarca/status/2076921882641870954
- Interesting. https://t.co/bva0Ex0dYk · x.com — https://x.com/pmarca/status/2076921321917985260
- RT @gtredoux: Frustrated by the plethora of old Soviet movies on Youtube (and elsewhere) without subtitles? Help is at hand. 1. Download i… · x.com — https://x.com/pmarca/status/2076921300061401230
- Interesting. Very interesting. https://t.co/0wCVwuIO2c · x.com — https://x.com/pmarca/status/2076919410254921884
- RT @agranato42: Really excited to see that this paper has gone public. One of the most exciting things a paper can do is show you that all… · x.com — https://x.com/pmarca/status/2076919377979654396
- Overheard in Silicon Valley: “We must have no fewer gig economy jobs, because that would be oppressive! But also no more gig economy jobs, … · x.com — https://x.com/pmarca/status/2076918395069718780
- RT @RichardSocher: There are at least a few reasons for why the incredible progress in AI hasn't yet resulted in a massive increase in GDP… · x.com — https://x.com/pmarca/status/2076889538342031512
- RT @nrmehta: I got chills reading this one from @illscience. One of the most optimistic things I've read in a long time. Our purpose on ear… · x.com — https://x.com/pmarca/status/2076887704177086645
- Amazing! https://t.co/ULUIvNRZ3j · x.com — https://x.com/pmarca/status/2076840946751442991
- RT @jdlichtman: Beautiful result on 2-primitive sets! GPT-5.6 just solved another 50+ year old problem (Erdős #793) https://t.co/rY4ZOelSzu · x.com — https://x.com/pmarca/status/2076840930532082028
- Many such cases. https://t.co/lHMzyVARLx · x.com — https://x.com/pmarca/status/2076837914844225550
- RT @sama: that goes hard · x.com — https://x.com/pmarca/status/2076837895781187762
- Interesting. https://t.co/7zBp2BfIXH · x.com — https://x.com/pmarca/status/2076837209383342178
- RT @illscience: https://t.co/BNnRp4EMsP · x.com — https://x.com/pmarca/status/2076837190618001479
- All true! And yet we progress. https://t.co/kvqnDefkBA · x.com — https://x.com/pmarca/status/2076833822923252139
- RT @MikeNayna: @pmarca One of Jacques Ellul’s core ideas is that technique solves one class of problems only to create new domains that req… · x.com — https://x.com/pmarca/status/2076833658472972587
- RT @boazbaraktcs: I live in MA so don't know much about Khanna but this is an example of what I wrote here that you can't hide something fr… · x.com — https://x.com/pmarca/status/2076726237339529218
- Many people are saying. https://t.co/ElvXDlDXek · x.com — https://x.com/pmarca/status/2076509276294787411
- RT @jpatel41: I’ve been saying this for months, and the early data is beginning to support it. Five years from now, I believe AI will creat… · x.com — https://x.com/pmarca/status/2076509238919364735
- Many people are saying. https://t.co/ihJHcEYH49 · x.com — https://x.com/pmarca/status/2076507691409904120
- RT @deredleritt3r: In late November 2025, AI agentic harnesses became highly useful, and enterprise adoption exploded. For example, based… · x.com — https://x.com/pmarca/status/2076507667871453460
- Interesting. https://t.co/sxxpWl8Ybx · x.com — https://x.com/pmarca/status/2076489027461726216
- RT @bryanrbeal: The fact this one dude - by himself - can walk into any state and find blatant fraud everywhere he goes proves two things:… · x.com — https://x.com/pmarca/status/2076489007882809713
- Interesting. https://t.co/WJctowxtHf · x.com — https://x.com/pmarca/status/2076488833672388705
- RT @Kaju_Nut: Yuji Tachikawa, one of the most respected mathematical physicists in the world, reports that Fable solved a problem that he a… · x.com — https://x.com/pmarca/status/2076488794422063574
- Interesting. https://t.co/23lzWU0QMv · x.com — https://x.com/pmarca/status/2076478538698621379
- RT @alexolegimas: The skill that Theo is implicitly describing in this essay is quite hard to do well: management. In business schools we h… · x.com — https://x.com/pmarca/status/2076478518595313890
- Interesting. https://t.co/7qoawDAY2b · x.com — https://x.com/pmarca/status/2076477941605917070
- RT @illscience: This is the best essay of the year: - The dispersion of knowledge is a collective strength; it’s the source of variety, ada… · x.com — https://x.com/pmarca/status/2076477921527783526
- Interesting. https://t.co/5kzzBdtgCc · x.com — https://x.com/pmarca/status/2076477443754622979
- RT @firesidealpha: Steve Mandel of Lone Pine was making a similar point, that there were fewer players doing true price discovery on assets… · x.com — https://x.com/pmarca/status/2076477420849475745
- Interesting. https://t.co/gnyrj8v0lj · x.com — https://x.com/pmarca/status/2076477237306720673
- RT @kane: All of Congressman’s @RoKhanna’s financial disclosures are finally OCR’d and searchable online He hand-filed papers instead of u… · x.com — https://x.com/pmarca/status/2076477213487370751
- Interesting. https://t.co/n2gtIR35Pa · x.com — https://x.com/pmarca/status/2076475573640339777
- RT @joshua_saxe: I'm perplexed at why people who I know are smarter than me, including the authors, buy the epistemic strategy used in the… · x.com — https://x.com/pmarca/status/2076475554401001953
- Interesting. https://t.co/yziqYFEuIK · x.com — https://x.com/pmarca/status/2076431021151801819
- RT @zanehengsperger: https://t.co/Ej6npGlQEr · x.com — https://x.com/pmarca/status/2076430999702175832
- Many people are saying. https://t.co/xZxdlF4mfc · x.com — https://x.com/pmarca/status/2076374690797649937
- RT @conorsen: The age 20-24 unemployment rate is now ~unchanged since the AI boom began: https://t.co/OXhFJ69xdO · x.com — https://x.com/pmarca/status/2076374667632463909
- Attractiveness is a bourgeois value and aspiration, so the cultural elite naturally rebel against it. Many such cases. https://t.co/OQInFte… · x.com — https://x.com/pmarca/status/2076227254691234107
- RT @BarneyFlames: its kind of wild that multiple fields of art somehow fall to collective creative delusion that making things look crap is… · x.com — https://x.com/pmarca/status/2076226740763173176
- Interesting. https://t.co/xQ7hd2uY24 · x.com — https://x.com/pmarca/status/2076198217545585079
- RT @goodside: It’s worth appreciating a few years ago all LLMs blatantly hallucinated all the time—even if you tried to prompt around it (s… · x.com — https://x.com/pmarca/status/2076198194820866172
- Co-sign. https://t.co/dBew8dlpAZ · x.com — https://x.com/pmarca/status/2076197469265010804
- RT @bubbleboi: Why AI makes jobs instead of loses them. 1. There are more profitable ideas/projects than time, talent, and general resourc… · x.com — https://x.com/pmarca/status/2076197436981440901
- Many people are saying! https://t.co/9zgRjQLYtH · x.com — https://x.com/pmarca/status/2076178032856011035

_+15167 more sources. Page the full corpus via the JSON export (`?include=all&sourcesOffset=…`)._

---

Structured data: https://researcher.now/public/entities/by-slug/marc-andreessen?include=all
Agent guide: https://researcher.now/agent.txt