← All Reports

The Quiet Alpha Map

What 80 leading venture firms are betting on for the AI era — and where the smart money quietly disagrees.

📅 June 8, 2026 · v3 (re-validated) 🔭 Galileo Research

Executive Summary

We read every forward-looking market claim published over the last twelve months by 80 of the most credible venture firms — from Sequoia, a16z, Founders Fund and Bessemer to specialist voices like Paradigm, NfX, Emergence and Pebblebed — and mapped where they converge and where they break apart. This v3 edition re-validated every attribution by reading the underlying claim rather than keyword-matching, and corrected the citation layer so each firm now links to the specific essay where it voiced each thesis.

The method is deliberately simple: a claim's signal is inversely proportional to how many firms already hold it. We sort every thesis by distinct-firm convergence into five bands — from Consensus (priced-in orthodoxy) down to Lone Signals (a single firm willing to say it out loud). The crowded top tells you what's already in the water. The sparse bottom is where edge, if it exists, is hiding.

Three things stand out. First, the consensus has hardened around two poles: software commoditizing up-stack, and atoms — robotics, defense, reindustrialization — the field has collectively decided the next decade is physical and that the app interface is no longer the moat. Second, the agent economy has consolidated into a single crowded bet: commerce, identity, and security for autonomous agents are now one converged thesis at the top of the table. Third, the most interesting disagreements are no longer about whether AI matters but about where the value settles: incumbents vs. insurgents, apps vs. infrastructure, capability vs. trust — and whether the whole thing is mistimed.

The five-second version:

“Maps to” sector key — pills are grouped into colour families:
Digital / AIAgentsAI appsAI labsSoftwarePhysical / atomsRoboticsEnergyDefenseNatural ResourcesHealthHealthcareMoney / market / metaFintechConsumerServicesMacroVC modelMoatsCryptoCryptoUnmatchedOther
Auto-derived from each claim by keyword — a glance aid for cross-referencing sectors, not a ranking.
① Consensus — the priced-in orthodoxy (21+ firms)

What the field has collectively decided. These are real and large — but if you're hearing it from 20+ firms, it's in the price. Useful as the baseline to measure everything else against.

ThemeFirmsThe claim, distilled — and why it mattersMaps toVoiced by
Embodied AI is the next platform 31 Robotics, humanoids, and "physical AI" are treated as the inevitable second act once language is solved. The framing has shifted from if to which form factor — and the capital is already moving. Robotics AStar1, Bessemer2, BoostVC3, Caffeinated4, Contrary5, Conviction6, Cowboy7, DayOne8, Emergence9, Endeavor10, Felicis11, First Round12, Founders Fund13, Haystack14, Index15, Interplay16, Khosla17, Lightspeed18, Pantera19, Pear20, Redpoint21, Riot22, SPC23, Sequoia24, Spark25, Starburst26, Susa27, Thrive28, USV29, YC30, a16z31
Software is commoditizing — value moves up-stack 31 If a competent model can generate the interface on demand, the UI moat erodes. The crowded conclusion: defensibility migrates to workflow ownership, proprietary context, and the business model — not the software itself. The single most-voiced thesis in the corpus. SoftwareAI appsAI labsMoats 500Global32, 8VC33, BasisSet34, Benchmark35, Bessemer36, BoxGroup37, Contrary38, Conviction39, DesignerFund40, Emergence41, Eniac42, Felicis11, First Round43, Floodgate44, Founder Collective45, Greylock46, Haystack47, Headline48, IAVentures49, Lightspeed50, Maveron51, NfX52, Notable53, Pebblebed54, SPC55, Slow56, Spark57, Streamlined58, Work-Bench59, XFund60, a16z61
Consumer AI: proactive, voice-first, context-aware 26 The "death of the prompt box": consumer AI that observes, anticipates, and acts rather than waiting to be asked. Passive context capture is the agreed-upon unlock; the open question is who owns the context. Consumer BasisSet62, Benchmark63, Bessemer36, BoxGroup64, Browder65, CRV66, CollabFund67, Cowboy68, Flybridge69, Forerunner70, General Catalyst71, Goodwater72, HustleFund73, Index15, Interplay74, Khosla75, Lightspeed76, Maveron77, NfX78, Notable53, Pear79, Redpoint80, SPC81, Spark82, USV83, a16z31
Proprietary data and workflow depth are the real moat 26 As models commoditize, the defensible asset is data and workflow nobody else has. Greylock sharpens it: not input data (commoditizing) but human-in-the-loop feedback — which outputs users accept, reject, or override — producing a hard-to-copy reasoning graph of how an organization thinks. SoftwareAI appsMoats 2048 Ventures84, 500Global32, Bessemer36, BoxGroup37, CRV85, Contrary38, Cowboy68, Eniac86, Felicis87, First Round43, Founder Collective88, Greylock46, Haystack47, Headline89, HustleFund90, IAVentures91, Index92, Maveron77, NfX93, Pear94, Redpoint21, SV Angel95, Spark96, Streamlined58, Thrive97, a16z31
Healthcare is AI's first trillion-dollar vertical 25 Near-universal conviction that healthcare's labor budget — not its software budget — is the prize. The disagreement is only on the wedge: clinical documentation, diagnostics, or owning the delivery entity outright. SoftwareHealthcare 2048 Ventures84, 8Bit98, 8VC99, Anthology100, Bessemer101, BoostVC102, BoxGroup37, CRV103, CollabFund67, Conviction104, Cowboy7, Eniac105, Felicis11, Floodgate44, Founder Collective45, Fusion Fund106, Goodwater107, Headline48, Khosla108, Pear79, Redpoint21, Susa109, Work-Bench110, XFund111, YC112
Defense and dual-use go mainstream 23 American Dynamism graduated from thesis to consensus. Counter-UAS, autonomy, and a software-defined industrial base are now table-stakes venture categories rather than contrarian bets. SoftwareDefense 8090113, 8VC114, Bessemer115, Caffeinated4, Contrary5, DCVC116, DayOne8, Endeavor10, Felicis11, Founders Fund117, General Catalyst118, Haystack119, HumanCapital120, Initialized121, Pear20, Riot22, SPC23, Starburst122, Susa27, Thrive123, USV124, YC125, a16z126
Agents as doers — and outcome pricing kills the seat 22 The copilot-to-autopilot migration as the real value frontier, with monetization shifting from per-seat to output- and outcome-based pricing. General Catalyst's framing: outcome-delivery builds a learning-by-doing flywheel point-solutions can't match. The contrarian wrinkle (Work-Bench): value these companies on absolute gross-profit dollars per customer — a labor-budget TAM 4-5x a SaaS seat — not on margin. SoftwareAgents 8Bit127, AStar1, Anthology100, BOND128, BasisSet34, Bessemer129, Browder130, CRV103, Contrary131, Emergence132, Felicis11, First Round133, Founder Collective134, General Catalyst135, Headline89, Notable136, Pear137, Redpoint80, SPC23, Sequoia138, USV139, a16z31
Reindustrialization and the industrial base 22 Onshoring, supply-chain sovereignty, and the "machine that makes the machine." A macro-political tailwind that has become a default lens for hard-tech investing across the field. DefenseNatural ResourcesMacro 8090113, Bessemer115, Caffeinated4, Contrary140, Cowboy7, DCVC116, DayOne8, Felicis11, Founders Fund117, Headline48, Initialized141, Interplay16, Khosla108, Redpoint21, Riot22, Sequoia142, Starburst26, Susa27, Thrive28, USV29, YC143, a16z126
The agent economy & its trust layer 21 Three facets of one structural shift, now converged into a single crowded bet. Commerce: when the buyer is an agent, the whole funnel — marketing, SEO, pricing, negotiation — re-architects for a non-human customer (Maveron's "ask-and-decide" agents kill the funnel; TenOneTen bets the real volume is unglamorous B2B back-office execution). Identity: proof-of-humanity and know-your-agent become prerequisites once machines transact. Security: spend controls, write-access limits, and shadow-AI governance become the differentiator — TenOneTen predicts an embarrassing agent-spend blow-up sorts the market. AgentsMoats Browder144, General Catalyst71, Interplay145, Paradigm146, Pebblebed147, Redpoint148, Sequoia149, TenOneTen150, USV151, a16z31, 8Bit152, Anthology100, Conviction6, Cowboy153, Lightspeed154, Notable155, Pear137, Susa156, Work-Bench110, DayOne157, Initialized158
② Strong — conviction forming, not yet crowded (15–20 firms)

Enough independent firms to be more than noise, few enough that the trade isn't done. This is where a thesis is becoming actionable before it becomes obvious.

ThemeFirmsThe claim, distilled — and why it mattersMaps toVoiced by
Coding is the killer app 19 The one application everyone agrees has crossed the threshold. The frontier has moved from autocomplete to autonomous background agents — and coding is the proving ground for everything agentic that follows. AgentsAI apps AStar1, Anthology100, Conviction159, Cowboy160, DesignerFund40, Emergence161, Felicis162, Greylock163, Headline89, Index164, Lightspeed50, Pear94, Pebblebed54, Redpoint148, SPC23, Sequoia138, Spark165, Work-Bench59, YC166
Atoms over bits — the physical bottlenecks return 18 After a decade where bits scaled faster than atoms, several firms argue the binding constraints are physical again: hardware, materials, manufacturing, and the un-virtualizable. The next generational companies work with atoms, and the moat is the physical thing that can't be copied to zero marginal cost. Moats 1517167, 8090113, AStar1, BOND128, Bessemer168, BoostVC169, DCVC116, DayOne8, Founder Collective88, Founders Fund170, Lightspeed18, Liquid2171, Redpoint21, Riot22, SPC23, Spark25, Susa27, USV124
This is a bubble — and it will reprice 18 A surprisingly broad skeptic caucus. Redpoint's analogy is sharpest: AI adoption is moving an order of magnitude faster than cloud, which makes it look more like the internet era — complete with Cisco-style "air pockets" where eyeballs don't convert to revenue, and where today's leaders may not be the eventual giants. Macro AStar1, Abstract172, Anthology100, Bessemer101, BoxGroup37, CRV85, CollabFund173, Contrary38, Endeavor174, Eniac105, Felicis87, Founder Collective45, Founders Fund13, HustleFund90, Notable175, Pantera176, SPC177, Thiel Capital178
Agent infrastructure is the new middleware 17 Picks-and-shovels for agents: orchestration, eval, memory, auth, observability. The bet is that today's agents are fragile, and the durable opportunity is the unglamorous tooling that makes them reliable in production. SoftwareAgents 8Bit152, Bessemer115, Cowboy153, Flybridge69, Index164, Lightspeed50, Neo179, Notable180, Paradigm181, Pear137, Redpoint21, SPC23, Sequoia149, USV139, Work-Bench59, YC166, a16z31
Biology is the next AI substrate 17 AI-x-bio as a generational platform: protein design, AI-run wet labs, and an "AlphaFold moment" for engineering simulation (Conviction). Crowded enough to be real, novel enough that the winning form factor is still contested. Healthcare 2048 Ventures182, 8VC99, BoostVC102, Contrary140, DayOne183, Eniac184, Founders Fund13, Fusion Fund106, Greylock163, Headline48, Index15, Initialized185, Pear186, SPC23, Susa27, Thiel Capital178, YC30
Talent and judgment become the scarce inputs 17 As AI commoditizes execution, the bottleneck moves to taste, judgment, and the rare people who have them. Floodgate's framing: AI "empowers order-makers and threatens order-takers." The corollary several firms draw: back the person, not the market. ConsumerMoats 1517187, AStar188, Abstract189, Browder190, Felicis191, Floodgate192, Founders Fund13, Haystack47, HumanCapital193, Neo179, NfX78, SPC55, Sequoia142, Soma194, Spark195, Susa27, XFund111
Models commoditize — architect for multi-model 16 The model layer keeps improving while becoming MORE competitive and multi-vendor, not less. Founders should architect for a multi-model world (open models, routing) and treat raw capability as a commoditizing input the market competes to supply. AI labs Anthology196, BOND128, BoxGroup37, Contrary38, Conviction39, Emergence41, Fusion Fund106, Greylock46, IAVentures91, Lightspeed50, NfX93, Redpoint21, Spark96, Susa109, USV83, XFund60
③ Emerging — early clusters worth tracking (9–14 firms)

Emerging structure. A handful of credible firms converging on something the broader field hasn't fully named yet.

ThemeFirmsThe claim, distilled — and why it mattersMaps toVoiced by
Incumbents & labs win the apps 13 A sober counter to startup romanticism: distribution, data and capital advantage the incumbents, while frontier labs march up the stack into applications. Pebblebed: SaaS incumbents won't be toppled overnight. Headline: vendor-native tooling from hyperscalers and labs becomes the quiet default. The consensus on this side is that the model — and increasingly the app — layer is not where most startups should plant a flag. SoftwareAI labsMoats 500Global197, AStar1, BOND128, BoxGroup64, Cowboy68, Felicis87, Flybridge198, Haystack199, Headline89, Pebblebed54, SPC200, Slow56, Thrive123
Non-AI fundamentals still win deals 13 A quiet contrarian steadiness: distribution, retention, unit economics and survival still decide outcomes. Haystack's two-rubric market — "native AI" companies judged on speed vs. everyone else needing an explicit reason to exist — is the sharpest articulation. Moats 1517167, Afore201, Anthology202, Bessemer36, CRV85, CollabFund173, Endeavor203, Eniac105, Founder Collective204, Haystack205, HustleFund90, Initialized206, Thiel Capital207
The VC model itself is breaking 13 The field turning the lens on itself: capital is no longer the differentiator; trust, listening and selection are. A reflexive admission that the mega-fund "capital-as-a-weapon" playbook may be self-defeating in a world of abundant money — and that the firm itself becomes software. SoftwareVC modelMoats 500Global197, AStar1, Abstract172, BoxGroup64, CRV85, Founder Collective208, Haystack47, IAVentures49, Neo209, NfX210, Pebblebed54, SV Angel95, Susa156
Energy is the binding constraint on AI 11 Compute is gated by power. The field has absorbed that data-center demand collides with grid reality, making generation, transmission, and efficiency first-order AI investments rather than adjacent ones. Energy 8090113, BOND128, BoostVC211, Contrary212, Conviction213, DayOne8, HustleFund214, Initialized215, Khosla108, Liquid2216, USV217
Inference is the runtime; value migrates to deployment 11 The bet that the action moves from training to serving: edge deployment, KV-cache economics, throughput-first clouds, and small specialized models that out-economize frontier models on unit cost. Value accrues to whoever deploys, not whoever builds. AI appsAI labs 8090218, BOND128, CRV85, Conviction219, Emergence161, Felicis162, Fusion Fund106, Index220, Redpoint21, Sequoia142, YC221
Brand, taste and trust as the durable moat 10 As capability converges across models, the differentiator becomes intangible: brand, taste, emotional attachment. Pebblebed's framing — once vibe coding makes the first 80% of a product nearly free, taste and trust become the scarce contested assets — is the cleanest statement of where consumer defensibility goes. ConsumerAI appsMoats CollabFund173, DesignerFund40, Floodgate222, Founder Collective134, HustleFund214, Maveron51, Pebblebed54, Spark57, Tuesday223, a16z31
Seed compression and younger founders 9 Series A bifurcating into $50M "native-AI" rounds vs. $10M everything-else. The middle hollows out; capital barbells toward category-defining bets and lean survivors, and several firms move allocation toward younger, non-credentialed founders the old playbook would have screened out. MacroVC model 1517167, 2048 Ventures224, AStar188, Afore201, Browder225, CRV85, Haystack47, HustleFund90, IAVentures226
④ Sharp — the almost-nobody bets (2–8 firms)

Two to eight serious firms looking at something most of the field is ignoring. Highest risk, highest potential edge — and where the most falsifiable, specific claims live.

ThemeFirmsThe claim, distilled — and why it mattersMaps toVoiced by
Liquidity reshapes — secondaries and the long private life 8 With companies staying private far longer and the IPO window erratic, secondaries and structured liquidity become a primary path to returns rather than a sideshow. A plumbing change most LPs haven't fully priced. Macro Abstract172, BoostVC227, Endeavor10, Haystack199, IAVentures49, Pantera228, Slow229, Streamlined230
Permanent capital, rollups and AI-enabled buyouts 8 Buy the boring business, install AI, expand the margin. The PE-venture convergence: own-and-operate service businesses and roll up fragmented verticals where AI is the operational alpha rather than a product feature. Services AStar1, CollabFund173, Contrary131, Felicis11, General Catalyst231, NfX210, SPC200, Thrive232
Crypto, stablecoins and tokenization come back 6 After a brutal altcoin year, a focused caucus bets the forward setup is asymmetric: stablecoins become necessary rails for machine-speed payments (TenOneTen), tokenization moves IPOs and equities on-chain (Paradigm), and privacy becomes the winner-take-most moat in crypto itself (a16z). CryptoFintechMacroMoats BoostVC3, Interplay145, Pantera233, Paradigm234, TenOneTen150, a16z235
Science institutions are broken — fund the heterodox 5 A conviction that academia and grant science are structurally failing, opening room for privately-funded, heterodox research (focused research orgs, independent labs). 1517 takes it furthest — risk culture itself, not capital, is the binding constraint on breakthroughs. Other 1517236, Founders Fund170, Initialized237, XFund238, a16z31
Space / orbital infrastructure as an investable frontier 5 Cheap reusable launch turns orbit from a moonshot into a venture stage. A small specialist cluster bets that in-space manufacturing, lunar resource extraction, and orbital compute/defense infrastructure are leadable seed/Series A categories now — not 2050 science fiction. The strangeness is the claimed timing, not the destination. DefenseNatural ResourcesMacroVC model DCVC239, Initialized240, Starburst122, Soma194, YC241
Emerging markets as the under-priced lane 4 While US capital crowds into horizontal AI, several firms point at lower-competition international and vertical markets. The steer: durable returns come from verticals with workflow moats and geographies with far less competition. SoftwareAI appsMacroMoats 500Global242, Endeavor10, HustleFund243, Index92
Permitting and interconnection as the real bottleneck 4 Not generation — connection. The unsexy regulatory and grid-interconnection queue is the actual constraint on powering AI, making the investable opportunity political and infrastructural rather than purely technological. Energy Contrary212, Conviction244, Initialized215, TenOneTen150
Skilled trades and the physical-labor premium 4 As knowledge work commoditizes, the un-automatable physical trades gain economic and cultural value. An under-explored corollary of the embodiment thesis — the humans who work with atoms get more valuable, not less. Robotics Conviction245, Cowboy7, Emergence9, Index246
Insurgents / AI-natives out-execute incumbents 3 The mirror image of the incumbents bet — and a genuine minority view. Anthology's data: at the application layer, AI-native startups now earn ~$2 for every $1 incumbents earn (63% share, up from 36%). DesignerFund and Spark argue speed and step-change UX let AI-natives rewrite workflow norms faster than entrenched distribution can defend. A small, falsifiable contrarian wager that the disruptors win this round. SoftwareAI appsMoats Anthology100, DesignerFund247, Spark96
⑤ Lone Signals — a single firm willing to say it (curated tier)

One firm, on the record, against the grain — the strongest standalone bets in the corpus, hand-picked from the validated lone-signal set. Serious, uncrowded, and investable. Most will be wrong. The ones that aren't are where outsized returns come from.

The betFirmWhy it mattersMaps to
AI breaks resource discovery, not scarcity Khosla108 Khosla inverts the scarcity narrative: the binding constraint on lithium, cobalt, steel and copper is discovery capacity, not the resource itself — and AI breaks the detection barrier, outpacing consumption by finding and designing new materials. Natural Resources
Insiders are the worst-placed to see the next breakthrough 1517248 1517's "inside context problem": the people closest to a field are systematically blind to its next discontinuity, because their identities and incentives are bound to the current paradigm. The investable corollary — back the outsider, the dropout, the teenager — runs against every credentialing instinct in venture. Other
Restaurants are mis-priced entertainment-fintech assets USV217 A genuinely strange reframe: restaurants as premium entertainment venues that will support native loyalty and payments layers — paired with USV's long-standing bet that credit-card interchange economics finally die. FintechConsumer
Liability is the moat the labs won't take Emergence41 Emergence's wedge: the frontier labs will not assume legal liability for autonomous actions, so the durable enterprise moat goes to whoever does — owning the outcome, the indemnity, and the accountability the model vendors refuse. AgentsAI labsMoats
Micro-entrepreneurs replace corporate employment Khosla75 Khosla's labor thesis: AI tooling lets individuals run company-scale operations, dissolving the default of corporate employment into a world of leveraged solo and micro-entrepreneurs — a structural shift in how work itself is organized. Other
Natural language becomes the API to humans Floodgate249 Floodgate's primitive: as agents proliferate, plain language becomes the universal interface between machines and people — the API to humans — and value accrues to whoever owns that conversational layer. SoftwareAgents
Prediction markets replace polls for power Paradigm250 Paradigm bets prediction markets move from novelty to civic infrastructure: priced, liquid forecasts displace polling as the way institutions read public will and allocate attention — with a new "opportunity markets" structure for scouting talent and research before it's signed. Fintech
$1 trillion of fraud is about to be exposed BoostVC3 BoostVC's contrarian forensic call: AI-grade detection surfaces roughly a trillion dollars of latent fraud across finance and commerce — a short thesis on the businesses quietly built on undetected deception. Other
Venture capital for people, not companies Slow251 Slow's signature heterodoxy: finance individuals directly — a share of a person's future earnings or a holding company over their output — rather than single companies. Legally novel, productized into a dedicated Creator Fund. VC model
78% of AI failures are invisible and model-immune Susa109 Susa's striking forensic claim: most production AI failures generate no complaint, no rating, no escalation — and ~94% would persist even with a much better model, because they're driven by interaction dynamics, not capability. If true, "scale fixes everything" is wrong, and an entire AI-supervision category is needed now. Other
🛸 Off the Deep End — bets from the far tail

Pulled out for their own spotlight: the strangest, most science-fiction-sounding single-firm theses in the corpus. We make no claim they're right. We note only that genuinely non-consensus returns, by definition, start out sounding like these.

The betFirmWhy it earns a spotlightMaps to
Space resources are investable today — asteroid, lunar, microgravity Caffeinated4, YC221 Not a 2050 bet. Caffeinated argues microgravity manufacturing (Varda) and energy "abundant from Earth to the asteroid belt" are category-defining seed/Series A markets to lead now; YC's RFS calls for startups extracting silicon, aluminium, iron and titanium from molten lunar regolith. The strangeness isn't the destination — it's the claimed timing: orbital industry as a current venture stage. Natural ResourcesEnergyMacroVC model
Cryogenic brain preservation is a fundable category DayOne183 DayOne's longevity tail: cryonics and brain preservation treated as a backable venture category, not a fringe — betting the science and the market both mature enough to make it a real fund line. Healthcare
Computational hardness caps superintelligence Pebblebed252 A structural anti-doom argument from complexity theory: if hard problems stay hard (P≠NP), there are limits no amount of scaling transcends, bounding what AI can ever do. A rare thesis that uses theoretical computer science to argue the superintelligence runaway is mathematically constrained. AI labs
AI-generated NSFW content goes mainstream Slow253 Slow's blunt consumer call: AI-generated adult content crosses from niche to mainstream as models and product experiences mature — a large, under-discussed demand pool most firms won't name on the record. Consumer
"Apocalypse-grade tech is good — accelerate into it" Founders Fund254 Founders Fund's accelerationist tail: catastrophic-risk technologies are inseparable from their inverse upside — climate doom implies weather control, pandemic risk implies nanobot immunity, AGI risk implies abundance. The rational posture is to accelerate into apocalyptic-grade technology. AI labs
⚔️ Sharp Contradictions — where serious firms directly disagree

The highest-value reading in the report. When credible firms reach opposite conclusions from the same evidence, the disagreement itself is the signal — it marks the questions still genuinely open. Expanded to seven two-sided debates, each cited on both sides.

  1. Do incumbents win the apps, or do AI-natives out-execute them? The single sharpest split in the corpus. Pebblebed, Headline and Felicis argue distribution, switching costs and vendor-native tooling let incumbents (and the labs moving up-stack) keep the application layer. Anthology counters with data — AI-natives now earn ~$2 per $1 incumbents earn at the app layer — and Spark and DesignerFund bet step-change UX and adoption speed let insurgents rewrite the rules first. Same evidence, opposite conclusions.
  2. Is this a bubble, or a super-cycle? Contrary, Anthology and HustleFund present data that usage is slipping, productivity gains are small, and a timing air-pocket is near; Redpoint warns of Cisco-style repricing. Against them, the Robotics and Software-commoditizing consensus bands assume relentless adoption. Both can't be right about 2026-27.
  3. Does value accrue to apps or to infrastructure? Floodgate and Fusion Fund bet value migrates to the application/intent layer; Spark and BOND bet the opposite — value concentrates into the deepest infra layers (models, datacenters, chips), where a few platforms absorb most of the return. The defining structural disagreement of the cycle.
  4. SaaS: does outcome-pricing kill the seat, or does seat-based still win? Bessemer, Emergence and General Catalyst bet monetization moves to output- and outcome-based pricing, dragging gross margins down. Work-Bench reframes it — value on gross-profit dollars per customer, not margin — while Pebblebed argues incumbent seat-based system-of-record revenue stays sticky via switching costs. The unit of account itself is contested.
  5. Open or closed web? USV and a16z bet permissionless, composable, agent-readable protocols beat closed frontier labs over time; the Incumbents-win caucus (Headline, Pebblebed) bets the closed, vendor-native stack stays the default. A bet on the shape of the next internet.
  6. Is capital a weapon or a commodity? Haystack sees $50M native-AI rounds as mega-funds wielding capital as an offensive weapon; CRV and again Haystack argue capital is now a non-differentiator and judgment is the scarce edge. The same firms hold both — a real tension about whether money still buys advantage.
  7. Is brand a durable moat, or does the agent bypass brand entirely? Spark, Pebblebed and FounderCollective bet that as capability converges, brand and taste become the last moat. But the agent-economy camp (Maveron, TenOneTen) argues an "ask-and-decide" agent buying on your behalf doesn't see brand at all — collapsing the marketing funnel that brand was built to win. Either brand is the moat, or the agent routes around it.
Methodology

We harvested every forward-looking market claim we could find from 80 venture firms — their essays, theses, annual reports, partner Substacks, and podcast transcripts — published primarily between June 2025 and June 2026. A small number of standing signature theses outside that window are retained and noted where a firm still actively acts on them.

Themes were derived bottom-up from the claims themselves, then each was scored by distinct-firm convergence — the number of separate firms independently voicing it. The five bands (Consensus 21+, Strong 15–20, Emerging 9–14, Sharp 2–8, Lone Signals) reflect how crowded a bet is, on the premise that signal is inversely proportional to how many people already hold it. Crowded themes are real but priced-in; sparse themes are where edge, if any, hides. Thresholds were recalibrated to the genuine post-validation distribution.

v3 changelog — what changed and why

This edition is a full re-validation, not a refresh. Counts were re-validated by reading each underlying claim rather than keyword-matching: roughly 27% of prior firm-attributions were keyword false-positives and were removed. In the other direction, 280 previously-untagged claims were mined — recovering 133 lone signals and 108 theme additions that the first pass missed. The citation layer was rebuilt: each firm now links to the specific essay where it voiced a given theme (drawn from per-(theme,firm) provenance), fixing the v1 defect where a single keyword matcher could point a firm at the wrong post; essays sharing a collapsed URL (YC, Conviction, Pear) are disambiguated by exact title.

Themes were also restructured: the old "incumbents-win" bucket was split by claim polarity into Incumbents & labs win the apps vs. Insurgents / AI-natives out-execute incumbents; three overlapping agent themes (commerce, identity, security) were merged into The agent economy & its trust layer; a new Space / orbital infrastructure theme was promoted from scattered lone signals into a five-firm cluster; and opposed-polarity bears were stripped from the Crypto and Brand counts. This is the audit trail — the report is the same map as June 5, corrected and made honest, not a different document.

This is a map of what credible investors are saying, not an endorsement of any claim. Every thesis is attributed to its specific source below.

📇 Firm Index — who shows up, how often, and where

Every firm in the write-up, alphabetical. Themes = how many distinct convergence themes the firm appears in. Articles = distinct source essays/posts cited for that firm. Sectors touched = the union of sector tags across the themes it voices.

FirmThemesArticlesSectors touched
151757ConsumerMacroMoatsVC model
2048 Ventures43AI appsHealthcareMacroMoatsSoftwareVC model
500Global55AI appsAI labsMacroMoatsSoftwareVC model
809054AI appsAI labsDefenseEnergyMacroMoatsNatural ResourcesSoftware
8Bit45AgentsHealthcareMoatsSoftware
8VC44AI appsAI labsDefenseHealthcareMoatsSoftware
a16z124AI appsAI labsAgentsConsumerCryptoDefenseFintechMacroMoatsNatural ResourcesRoboticsSoftware
Abstract42ConsumerMacroMoatsSoftwareVC model
Afore27MacroMoatsVC model
Anthology85AI appsAI labsAgentsHealthcareMacroMoatsSoftware
AStar103AI appsAI labsAgentsConsumerMacroMoatsRoboticsServicesSoftwareVC model
BasisSet34AI appsAI labsAgentsConsumerMoatsSoftware
Benchmark24AI appsAI labsConsumerMoatsSoftware
Bessemer126AI appsAI labsAgentsConsumerDefenseHealthcareMacroMoatsNatural ResourcesRoboticsSoftware
BOND61AI appsAI labsAgentsEnergyMoatsSoftware
BoostVC75CryptoEnergyFintechHealthcareMacroMoatsRoboticsSoftware
BoxGroup84AI appsAI labsConsumerHealthcareMacroMoatsSoftwareVC model
Browder56AgentsConsumerMacroMoatsSoftwareVC model
Caffeinated31DefenseMacroNatural ResourcesRoboticsSoftware
CollabFund65AI appsConsumerHealthcareMacroMoatsServicesSoftware
Contrary127AI appsAI labsAgentsDefenseEnergyHealthcareMacroMoatsNatural ResourcesRoboticsServicesSoftware
Conviction102AI appsAI labsAgentsEnergyHealthcareMoatsRoboticsSoftware
Cowboy106AI appsAI labsAgentsConsumerDefenseHealthcareMacroMoatsNatural ResourcesRoboticsSoftware
CRV96AI appsAI labsAgentsConsumerHealthcareMacroMoatsSoftwareVC model
DayOne76AgentsDefenseEnergyHealthcareMacroMoatsNatural ResourcesRoboticsSoftware
DCVC43DefenseMacroMoatsNatural ResourcesSoftwareVC model
DesignerFund44AI appsAI labsAgentsConsumerMoatsSoftware
Emergence76AI appsAI labsAgentsMoatsRoboticsSoftware
Endeavor64AI appsDefenseMacroMoatsRoboticsSoftware
Eniac66AI appsAI labsHealthcareMacroMoatsSoftware
Felicis136AI appsAI labsAgentsConsumerDefenseHealthcareMacroMoatsNatural ResourcesRoboticsServicesSoftware
First Round47AI appsAI labsAgentsMoatsRoboticsSoftware
Floodgate47AI appsAI labsConsumerHealthcareMoatsSoftware
Flybridge37AI labsAgentsConsumerMoatsSoftware
Forerunner13Consumer
Founder Collective910AI appsAI labsAgentsConsumerHealthcareMacroMoatsSoftwareVC model
Founders Fund84ConsumerDefenseHealthcareMacroMoatsNatural ResourcesRoboticsSoftware
Fusion Fund41AI appsAI labsHealthcareSoftware
General Catalyst55AgentsConsumerDefenseMoatsServicesSoftware
Goodwater24ConsumerHealthcareSoftware
Greylock55AI appsAI labsAgentsHealthcareMoatsSoftware
Haystack107AI appsAI labsConsumerDefenseMacroMoatsRoboticsSoftwareVC model
Headline83AI appsAI labsAgentsDefenseHealthcareMacroMoatsNatural ResourcesSoftware
HumanCapital22ConsumerDefenseMoatsSoftware
HustleFund84AI appsConsumerEnergyMacroMoatsSoftwareVC model
IAVentures66AI appsAI labsMacroMoatsSoftwareVC model
Index99AI appsAI labsAgentsConsumerHealthcareMacroMoatsRoboticsSoftware
Initialized912AgentsDefenseEnergyHealthcareMacroMoatsNatural ResourcesSoftwareVC model
Interplay56AgentsConsumerCryptoDefenseFintechMacroMoatsNatural ResourcesRobotics
Khosla57ConsumerDefenseEnergyHealthcareMacroNatural ResourcesRoboticsSoftware
Lightspeed811AI appsAI labsAgentsConsumerMoatsRoboticsSoftware
Liquid224EnergyMoats
Maveron46AI appsAI labsConsumerMoatsSoftware
Neo33AgentsConsumerMoatsSoftwareVC model
NfX78AI appsAI labsConsumerMoatsServicesSoftwareVC model
Notable69AI appsAI labsAgentsConsumerMacroMoatsSoftware
Pantera44CryptoFintechMacroMoatsRobotics
Paradigm312AgentsCryptoFintechMacroMoatsSoftware
Pear103AI appsAgentsConsumerDefenseHealthcareMoatsRoboticsSoftware
Pebblebed66AI appsAI labsAgentsConsumerMoatsSoftwareVC model
Redpoint124AI appsAI labsAgentsConsumerDefenseHealthcareMacroMoatsNatural ResourcesRoboticsSoftware
Riot41DefenseMacroMoatsNatural ResourcesRoboticsSoftware
Sequoia87AI appsAI labsAgentsConsumerDefenseMacroMoatsNatural ResourcesRoboticsSoftware
Slow39AI appsAI labsMacroMoatsSoftware
Soma22ConsumerDefenseMacroMoatsNatural ResourcesVC model
Spark107AI appsAI labsAgentsConsumerMoatsRoboticsSoftware
SPC137AI appsAI labsAgentsConsumerDefenseHealthcareMacroMoatsRoboticsServicesSoftware
Starburst43DefenseMacroNatural ResourcesRoboticsSoftwareVC model
Streamlined34AI appsAI labsMacroMoatsSoftware
Susa106AI labsAgentsConsumerDefenseHealthcareMacroMoatsNatural ResourcesRoboticsSoftwareVC model
SV Angel23AI appsMoatsSoftwareVC model
TenOneTen31AgentsCryptoEnergyFintechMacroMoats
Thiel Capital34HealthcareMacroMoats
Thrive610AI appsAI labsDefenseMacroMoatsNatural ResourcesRoboticsServicesSoftware
Tuesday12AI appsConsumerMoats
USV1010AI labsAgentsConsumerDefenseEnergyMacroMoatsNatural ResourcesRoboticsSoftware
Work-Bench57AI appsAI labsAgentsHealthcareMoatsSoftware
XFund55AI appsAI labsConsumerHealthcareMoatsSoftware
YC91AI appsAI labsAgentsDefenseHealthcareMacroNatural ResourcesRoboticsSoftwareVC model

Sources

  1. [167] 1517 — A drop in the ocean ↩
  2. [236] 1517 — Flux Capacitor 2026: Applications Open for 1517 Angel Check Program ↩
  3. [248] 1517 — Inside v. Outside Context Problems ↩
  4. [187] 1517 — Why Bell Labs Worked ↩
  5. [182] 2048 Ventures — 2048 Thesis — Software is Eating Bio ↩
  6. [84] 2048 Ventures — 2048 Thesis — Vertical AI / Frontier Tech / Health / Bio ↩
  7. [224] 2048 Ventures — Introducing 2048 Ventures Pre-Seed Fast Track ↩
  8. [242] 500Global — Christine's Quarterly Letter: Q2'25 Venture Review & Outlook ↩
  9. [197] 500Global — Rewriting the Venture Capital Playbook for the AI Era ↩
  10. [32] 500Global — The Moat Hasn't Disappeared. It Has Been Redefined. ↩
  11. [113] 8090 — Reindustrialization and Deep Tech Investing with 8090 Industries' Grant Brown ↩
  12. [218] 8090 — World's Rich Bet On VC Firm With 275% Gains That Shuns AI (standing signature thesis) ↩
  13. [152] 8Bit — Barndoor AI raises $13.6M to deliver the first control plane for agentic AI workforces (portfolio announcement) ↩
  14. [127] 8Bit — Glue raises $20M to build the first platform for agentic team chat (portfolio announcement) ↩
  15. [98] 8Bit — Honey Health raises $7.8M to automate healthcare's back office with autonomous AI agents (portfolio announcement) ↩
  16. [99] 8VC — Joe Lonsdale's 8VC seeking ~$1B for new fund (reported) ↩
  17. [33] 8VC — The Future of AI, Defense & America — Joe Lonsdale (Global Alts Miami 2026) ↩
  18. [114] 8VC — Top Tech Investors: This AI Wave Is 10X Bigger Than SaaS (8VC podcast, Alex Kolicich & Jack Moshkovich) ↩
  19. [31] a16z — Big Ideas 2026: Part 1 ↩
  20. [126] a16z — Big Ideas 2026: Part 2 ↩
  21. [235] a16z — Big Ideas 2026: Part 3 ↩
  22. [61] a16z — Humans Are for Ideas, AI Is for Execution ↩
  23. [172] Abstract — Ramtin Naimi - Building Abstract (Invest Like the Best EP.435) [interview, non-publisher] ↩
  24. [189] Abstract — The Man With the Hot Hand (Colossus profile of Ramtin Naimi) [profile, non-publisher] ↩
  25. [201] Afore — The Power of Pre-Seed: Why Momentum Is the Only Moat (Gaurav Jain on Tank Talks) ↩
  26. [196] Anthology — 2025 Mid-Year LLM Market Update ↩
  27. [100] Anthology — 2025: The State of Generative AI in the Enterprise ↩
  28. [202] Anthology — Deedy Das (Menlo) — The Haves and Have-Nots of the AI Gold Rush ↩
  29. [1] AStar — How AI Is Rewriting Seed Stage Investing (Uncapped Ep. 49 transcript) ↩
  30. [188] AStar — Kevin Hartz | Backing Teen Founders, Lessons from the PayPal Mafia (The Split podcast) ↩
  31. [62] BasisSet — Our Favorite AI Workflows of 2025 ↩
  32. [34] BasisSet — You Don't Need to be First ↩
  33. [35] Benchmark — The Next AI Wave Will Be Social, Not Solo (AI & I podcast) ↩
  34. [63] Benchmark — Will AI be as big of a catalyst for a consumer AI wave as mobile? ↩
  35. [115] Bessemer — Defense tech roadmap: Five frontiers for 2026 ↩
  36. [2] Bessemer — Intelligent robotics: The new era of physical AI ↩
  37. [101] Bessemer — State of Health AI 2026 ↩
  38. [129] Bessemer — The AI pricing and monetization playbook ↩
  39. [36] Bessemer — The State of AI 2025 ↩
  40. [168] Bessemer — Vision-native AI opportunities: a precursor to intelligent robotics ↩
  41. [128] BOND — Trends — Artificial Intelligence ↩
  42. [3] BoostVC — 2026 Predictions (Trends, Triggers and Timing) ↩
  43. [227] BoostVC — My Hilariously Incomplete Summary of 2025 ↩
  44. [102] BoostVC — Request for Startups: Longevity/Bio ↩
  45. [169] BoostVC — Twelve Years, 400 Startups, and a new $87,654,321 Fund ↩
  46. [211] BoostVC — Why We're Leading Radiant's $300m Series D ↩
  47. [64] BoxGroup — An Unfiltered Conversation with David Tisch, Founder of BoxGroup ↩
  48. [37] BoxGroup — Raising Your Seed Round (David Tisch talk) ↩
  49. [225] Browder — 20VC: Turning Peter Thiel's $100K into a $10M Angel Portfolio (Joshua Browder) ↩
  50. [130] Browder — AI4 2025 talk: 5 Key Lessons From Building and Investing in Billion-Dollar AI Companies ↩
  51. [190] Browder — Josh Browder on personal asset allocation in the AI era (clip) ↩
  52. [144] Browder — The AI Personal Stack (Newcomer / Cerebral Valley) ↩
  53. [65] Browder — The Consumer Rights Revolution (Joshua Browder, DoNotPay) ↩
  54. [4] Caffeinated — Portfolio positioning (firm site) ↩
  55. [173] CollabFund — Collab Holdings: A Different Approach to Private Equity for the Best Consumer Brands ↩
  56. [67] CollabFund — WHOOP ↩
  57. [5] Contrary — Can America Build Again? (Reshoring & Automation) ↩
  58. [212] Contrary — Mapping the Shadow Grid ↩
  59. [38] Contrary — The Case for Specialized AI ↩
  60. [140] Contrary — The State of the US-China Biotech Race ↩
  61. [131] Contrary — The Vertical AI Playbook ↩
  62. [6] Conviction — Startup Ideas — Certifying Robot Intelligence ↩
  63. [244] Conviction — Startup Ideas — Instrumenting the Administrative State ↩
  64. [159] Conviction — Startup Ideas — Pay-for-Correctness ↩
  65. [104] Conviction — Startup Ideas — Portals As A Service ↩
  66. [213] Conviction — Startup Ideas — The Inference-First Grid ↩
  67. [219] Conviction — Startup Ideas — The Red Team for Silicon ↩
  68. [245] Conviction — Startup Ideas — Vertical Robotics / Machine Verified Trades ↩
  69. [39] Conviction — State of Startups and AI 2025 (Sarah Guo, Conviction) ↩
  70. [7] Cowboy — AI and the Labor Shortage Economy ↩
  71. [68] Cowboy — It's Time For A B2C Comeback In the Age of AI. But it won't be easy ↩
  72. [160] Cowboy — Where we're looking to invest: Rebuilding the Dev Stack for the AI Era ↩
  73. [153] Cowboy — Where we're looking to invest: the evolution of the MCP ecosystem ↩
  74. [66] CRV — Saar Gur on Lotus AI / consumer-AI timing (interview) ↩
  75. [103] CRV — Saar Gur on Lotus AI Series A Investment (interview) ↩
  76. [85] CRV — Series A Metrics VCs Expect in 2026 ↩
  77. [157] DayOne — Day One Ventures And Masha Bucher: Big Ideas For Humanity ↩
  78. [183] DayOne — Day One Ventures secures $150M Fund III: Aims to redefine future of human ↩
  79. [8] DayOne — Masha Bucher 2025 Investment Review ↩
  80. [116] DCVC — 2025 Deep Tech Opportunities Report: An American Industrial Renaissance ↩
  81. [239] DCVC — Deep Tech Opportunity: Air and Space ↩
  82. [40] DesignerFund — AI in Design 2026: The inflection point is here ↩
  83. [247] DesignerFund — Four Shifts Designers Can't Ignore in the Age of AI ↩
  84. [9] Emergence — Autonomy Comes to the Built World ↩
  85. [41] Emergence — Should Your SaaS Company Become a Services Business? ↩
  86. [161] Emergence — The Next Token Supercycle: Opportunities in Long-Horizon Inference ↩
  87. [132] Emergence — Why AI-Native Services, and Why Now ↩
  88. [203] Endeavor — 2025 Endeavor Catalyst Annual Report: A Look into Global Venture Capital ↩
  89. [174] Endeavor — Allen Taylor on the Emerging Markets Enthusiast Podcast ↩
  90. [10] Endeavor — Silicon Valley got crowded again. Here are 400 reasons why we're staying Elsewhere. ↩
  91. [86] Eniac — Eniac portfolio signal: Origin Lab / world-model data (firm stories page, 2025) ↩
  92. [42] Eniac — Hadley Harris: 2025 Eniac AI Market Map (LinkedIn, Jan 2025) ↩
  93. [105] Eniac — Nihal Mehta investor interview (April 2025) ↩
  94. [184] Eniac — Vic Singh: LxMs in Biological Intelligence (LinkedIn essay) ↩
  95. [191] Felicis — 2025 Annual Letter: An Estimation of Underestimations ↩
  96. [11] Felicis — Free Software & The Future of Software Business Models ↩
  97. [87] Felicis — Q4 '25 Cloud Update (What's going on with software?) ↩
  98. [162] Felicis — The Hyperscalers & The Compute Crunch (Quarterly Update) ↩
  99. [43] First Round — "AI-Powered" Isn't a Position ↩
  100. [12] First Round — Firsthand: Seven Years Inside Applied Intuition, the Quietly Dominant Physical AI Company ↩
  101. [133] First Round — Serval's Path to Product-Market Fit — Win Enterprise Buyers by Treating Them Like Consumers ↩
  102. [249] Floodgate — Era of Mass Cognition ↩
  103. [44] Floodgate — From Found Revenue to Found Health ↩
  104. [222] Floodgate — Languaging: The Software of Belief ↩
  105. [192] Floodgate — Superbuilders and Superthinkers ↩
  106. [69] Flybridge — Memory: The Missing Piece for Real Intelligence ↩
  107. [198] Flybridge — Why AI Hasn't Broken Through in the $145B Accounting Sector ↩
  108. [70] Forerunner — Proactive Measures: Be the CEO of Your Own Life ↩
  109. [134] Founder Collective — 6 Takeaways from Collective Future NYC ↩
  110. [88] Founder Collective — Dumb Hardware Depreciates. Smart Hardware Compounds. ↩
  111. [204] Founder Collective — Not Every Company Is Built for the Big Leagues ↩
  112. [45] Founder Collective — The Buy Window ↩
  113. [208] Founder Collective — The Hidden Cost of Kingmaking ↩
  114. [13] Founders Fund — hypercapitalism and the AI talent wars ↩
  115. [117] Founders Fund — No Solvency, No Security ↩
  116. [170] Founders Fund — What Happened to the Future? (Founders Fund Manifesto) ↩
  117. [254] Founders Fund — What Happened to the Future? (Hereticon II — End of the World's Fair) ↩
  118. [106] Fusion Fund — Next phase for AI value: infrastructure, not models (Lu Zhang interview) ↩
  119. [135] General Catalyst — Europe's AI Transformation in Services ↩
  120. [118] General Catalyst — Realizing Resilience ↩
  121. [71] General Catalyst — The Agentic Commerce Opportunity ↩
  122. [231] General Catalyst — The Future of Services ↩
  123. [72] Goodwater — Goodwater consumer AI thesis (insights hub synthesis) ↩
  124. [107] Goodwater — Investing in the Future of Women's Health (Midi Health) ↩
  125. [163] Greylock — Reid Hoffman: Five Predictions About AI in 2026 (AI & I) ↩
  126. [46] Greylock — The New New Moats ↩
  127. [14] Haystack — Haystack Fund VIII / Firm Thesis (AI, automation, robotics, energy) ↩
  128. [199] Haystack — Lightspeed Generative Now: Venture Capital Trends in the Age of AI (with Michael Mignano) ↩
  129. [47] Haystack — The New Rules of Early-Stage Venture in the AI Era (with Semil Shah) ↩
  130. [119] Haystack — The Story Behind Haystack's Investment in Nominal ↩
  131. [205] Haystack — The Story Behind Our Investment in Birches Health ↩
  132. [48] Headline — Introducing the Top 50 Vertical SaaS Companies to Watch in 2025 ↩
  133. [89] Headline — The Infra Stack Reset: 10 Trends Reshaping Spend in 2025 ↩
  134. [193] HumanCapital — About — Human Capital (firm positioning) ↩
  135. [120] HumanCapital — Portfolio (revealed thesis) ↩
  136. [243] HustleFund — Crossing 10,000 Angel Squad applications (elizabethyin.com) ↩
  137. [73] HustleFund — How to Evaluate a Founder When There's Nothing to Evaluate (Small Bets) ↩
  138. [214] HustleFund — The Artificial Intelligence Driven Future: Who Wins and Who Loses? ↩
  139. [90] HustleFund — The wide path: Why most current valuation dynamics are not here to stay (and what will) ↩
  140. [91] IAVentures — IA Ventures standing thesis: data as defensibility (firm positioning) ↩
  141. [226] IAVentures — Inside Roger Ehrenberg's new firm (INTERVIEW/profile) ↩
  142. [49] IAVentures — Roger Ehrenberg interview: Why VC Returns Will Get Worse (INTERVIEW, not firm essay) ↩
  143. [164] Index — Acts of Claude: Building with Agents, a 'Work in Progress' ↩
  144. [246] Index — Building the first intelligence layer for industrial assets: Our investment in Scope ↩
  145. [92] Index — From Assistive AI to Authoritative Systems: How Wonderful is taking Enterprise AI out of Pilot Mode ↩
  146. [15] Index — Index partner-team 2026 predictions (referenced) ↩
  147. [220] Index — Index Perspectives — Infrastructure & AI positioning ('inference is the new runtime') ↩
  148. [185] Initialized — 10x Science: The Founders Who Built the Field Are Now Building the Platform ↩
  149. [237] Initialized — Backing Garage: Transforming How America's Critical Equipment Is Bought and Sold ↩
  150. [240] Initialized — Initialized Leads $7M Round for Enhanced Radar ↩
  151. [206] Initialized — Introducing Feanix: The Intelligence Layer for Modern Farming ↩
  152. [215] Initialized — PermitFlow's Rise in Construction AI: From Seed to Series B ↩
  153. [121] Initialized — The Future of Orbital Defense: Why We're Leading Orbital Operations' $8.8M Seed Round ↩
  154. [158] Initialized — The Internet Needs to Know Who You Are ↩
  155. [141] Initialized — The Next Great Food Infrastructure Company Won't Be Built Around a Single Ingredient ↩
  156. [74] Interplay — Investing in AI Voice Agents: Where Do We Go From Here? ↩
  157. [16] Interplay — Manufacturing Is Growing on Automation + AI, Not Headcount ↩
  158. [145] Interplay — The Future of Commerce: Agents are Rebuilding B2B Procurement from the Ground Up ↩
  159. [108] Khosla — A Roadmap to AI Utopia (Khosla Ventures / TIME) ↩
  160. [75] Khosla — Vinod Khosla on the End of Jobs and the Future of Capitalism (Newcomer) ↩
  161. [17] Khosla — Vinod Khosla: today's 5-year-olds won't need a job (Business Insider) ↩
  162. [154] Lightspeed — AI-Driven Cyberattacks are on the Rise. Are You Ready? ↩
  163. [18] Lightspeed — Hello, World Models! ↩
  164. [50] Lightspeed — The Engine That Builds the Engine ↩
  165. [76] Lightspeed — What AI Knows ↩
  166. [171] Liquid2 — Mike Miller investor focus (NFX Signal profile) ↩
  167. [216] Liquid2 — What a Seed Investor Looks For — with Mike Miller (interview) ↩
  168. [51] Maveron — A Modern House of Brands: Why the Next Great Consumer Platforms Will Be AI-Native and Built for Synergy ↩
  169. [77] Maveron — What Consumers Are Doing Now And What Comes Next (Consumer 2026 report) ↩
  170. [209] Neo — Ali Partovi's Neo looks to upend the accelerator model with low-dilution terms ↩
  171. [179] Neo — Y Combinator Rival Neo Is Harder To Get Into Than Harvard ↩
  172. [52] NfX — 3 Waves of Successful Generative AI Startups ↩
  173. [78] NfX — Consumer is Back – And Why It's Been So Hard Since 2014 ↩
  174. [93] NfX — Generative AI Market Map and 5-Layer Tech Stack ↩
  175. [210] NfX — How 'Venture Capital 3.0' Impacts Founders in the AI Age ↩
  176. [180] Notable — Agent Infrastructure's Moment ↩
  177. [175] Notable — Hans Tung on Why Founders Need to Play the Long Game (TechCrunch Equity podcast) ↩
  178. [136] Notable — Is ARR Dead – Or Just Evolving? A Look At Consumption-Based Pricing ↩
  179. [53] Notable — The Future of Voice AI: How Standalone Companies Can Thrive in a Speech-to-Speech World ↩
  180. [155] Notable — The New Weakest Link: The Shift from Human Error to AI Agent Risk ↩
  181. [19] Pantera — Forward Outlook on the Opportunity Set for Fund V ↩
  182. [233] Pantera — Market Outlook for 2026 + Nine Predictions for 2026 ↩
  183. [228] Pantera — Nine Predictions for 2026 ↩
  184. [176] Pantera — TOKEN2049 Keynote: A Multi-Decade Transformation ↩
  185. [146] Paradigm — Introducing the 2026 Paradigm Fellowship ↩
  186. [181] Paradigm — Open Sourcing Centaur: Multiplayer, self-hosted, secure agents ↩
  187. [250] Paradigm — Opportunity Markets ↩
  188. [234] Paradigm — Public Stocks on Public Blockchains: How Do We Get There? ↩
  189. [20] Pear — Pear AI Vertical — Perspectives ↩
  190. [79] Pear — Pear's Request for Startups — Ambient AI for vertical workflows (Mar Hershenson) ↩
  191. [137] Pear — Pear's Request for Startups — Financial-grade agent infrastructure (Ryan Sells) ↩
  192. [94] Pear — Pear's Request for Startups — Plan mode for every knowledge worker (Shravan Reddy) ↩
  193. [186] Pear — Pear's Request for Startups — Verified consumer health marketplaces (Warren Shaeffer) ↩
  194. [252] Pebblebed — "Intelligence" is not a technical term. Hardness is. ↩
  195. [54] Pebblebed — Software is dead. Long live software ↩
  196. [147] Pebblebed — The Internet Will Have More Agents Than Humans ↩
  197. [80] Redpoint — 25 for '25: The App Renaissance Arrives ↩
  198. [148] Redpoint — AI Update: 2025 InfraRed Report ↩
  199. [21] Redpoint — From Pilots to Platforms: Picks for '26 ↩
  200. [22] Riot — From Generalist to Deep Tech Specialist: A Conversation with Will Coffield (Stifel Forge Series) ↩
  201. [138] Sequoia — 2026: This is AGI ↩
  202. [142] Sequoia — AI in 2026: The Tale of Two AIs ↩
  203. [24] Sequoia — AI Retail Opportunity ↩
  204. [149] Sequoia — The Agent Economy: Building the Foundations for an AI-Powered Future ↩
  205. [253] Slow — 6 Predictions For 2026 ↩
  206. [229] Slow — Creator Financing Is Going Mainstream ↩
  207. [251] Slow — Investing Directly in People Is the Future of VC ↩
  208. [56] Slow — The State of VC in 2023 - a Techno-Realist Manifesto ↩
  209. [194] Soma — Soma Capital homepage (mission / portfolio framing) ↩
  210. [82] Spark — AI Severance: One Memory or Many? (Hallway Chat) ↩
  211. [195] Spark — Making the perfect backpack: craft in the age of AI (Hallway Chat) ↩
  212. [165] Spark — Personality, humanism, Poke, and raising the ceiling (Hallway Chat) ↩
  213. [96] Spark — Questions we're asking of AI startups in 2025 (Hallway Chat) ↩
  214. [25] Spark — Spark Capital VC Nabeel Hyatt sees robotics reshaping the gig economy ↩
  215. [57] Spark — The three types of AI startups; art vs craft (AI & I w/ Dan Shipper) ↩
  216. [23] SPC — Avoiding The Eye of Sauron ↩
  217. [177] SPC — Betting an Entire Company on AI, Bubble or Not ↩
  218. [81] SPC — Entering the Post-Prompting World ↩
  219. [200] SPC — The Silicon Valley Playbook is Broken (with Elad Gil) ↩
  220. [55] SPC — When Your Life's Work Becomes Free and Abundant ↩
  221. [122] Starburst — Defending the Skies (white paper series) ↩
  222. [26] Starburst — Redefining Starburst ↩
  223. [230] Streamlined — Fox Business 'Mornings with Maria' — Investment expert discusses major AI IPOs, sees private market unlocking (TV interview) ↩
  224. [58] Streamlined — Ullas Naik: The Future of Seed Investing, AI, and Data Science (long-form interview) ↩
  225. [27] Susa — Deep Tech Companies Are Built Different ↩
  226. [109] Susa — The AI Blind Spot That Gets Bigger as AI Takes On More ↩
  227. [156] Susa — The AI Moment, Part IV: Lightning Round ↩
  228. [95] SV Angel — The Bottom-Up Signal (profile/interview, Topher Conway) ↩
  229. [150] TenOneTen — Money at Machine Speed ↩
  230. [207] Thiel Capital — A Mind-Bending Conversation with Peter Thiel (NYT 'Interesting Times' / 'A.I., Mars and Immortality') ↩
  231. [178] Thiel Capital — A.I., Mars and Immortality: Are We Dreaming Big Enough? (transcript) ↩
  232. [97] Thrive — Fast Company (Vince Hankes): OpenAI's data flywheel ↩
  233. [28] Thrive — Hill & Valley Forum 2025 (Josh Kushner): embodied AI, robotics, reshoring ↩
  234. [232] Thrive — Invest Like the Best (Josh Kushner): Disruption from the inside out ↩
  235. [123] Thrive — Uncapped podcast (Vince Hankes): every Big Tech company has a 'bazooka pointed at' OpenAI ↩
  236. [223] Tuesday — Launched on a Tuesday (frog x Tuesday Capital) ↩
  237. [217] USV — 10 Highlights from USV's 2025 Annual Meeting ↩
  238. [83] USV — History Rhymes? ↩
  239. [139] USV — Our Investment in Anon ↩
  240. [29] USV — Rapidly Deployed Robots at Industrial Scale: Tutor Intelligence ↩
  241. [151] USV — Suno ↩
  242. [124] USV — The Future of Programmable Everything Runs on Efficient Energy ↩
  243. [110] Work-Bench — Enterprise Weekly #543: What We're Excited About Going into 2026 ↩
  244. [59] Work-Bench — The Rise of the Agent Runtime ↩
  245. [111] XFund — About / Firm Thesis page ↩
  246. [238] XFund — Harvard-linked VC fund goes... (Fund IV / Open Evidence coverage) ↩
  247. [60] XFund — The Hacker-Humanist (Patrick Chung interview, standing thesis) ↩
  248. [112] YC — RFS Summer 2026: AI-Native Discovery Engines (Jon Xu) ↩
  249. [30] YC — RFS Summer 2026: AI-Powered Agriculture ↩
  250. [125] YC — RFS Summer 2026: Counter-Drone / Counter-Swarm Defense ↩
  251. [143] YC — RFS Summer 2026: Faster US Hardware Iteration ↩
  252. [166] YC — RFS Summer 2026: Hypercustomized / User-Modifiable Software ↩
  253. [241] YC — RFS Summer 2026: Inference Chips in Space ↩
  254. [221] YC — RFS Summer 2026: Lunar & Space Industrial Capabilities ↩

Generated by Galileo 🔭 · June 8, 2026 · v3 (re-validated)