The CCS Interview Index: what 1,084 founder conversations say about where crypto is building, 2018 to 2026

Crypto Coin Show has recorded long-form interviews with founders, executives and investors in digital assets since 2014. This report treats that archive as a dataset. Each of the 1,084 episodes published on cryptocoinshow.com at the September 29, 2026 cut carries topic, company and guest tags, and because the interviews were booked in real time, week by week, as projects sought coverage, the mix of what founders wanted to talk about is a reasonable proxy for where builders and their backers were putting effort in any given year. It is not a survey and it is not a market share measure. It is a record of who showed up to pitch, and about what.

This expanded edition keeps the year-by-year topic shares and the four structural shifts from the original Index, and adds the layer that only an interview desk can add: the people. The second half of the report walks through the 2026 cohort company by company, the guests who came back more than once this year, the episodes recorded on the floor at Consensus Miami and the Blockchain Futurist Conference in Toronto, and what the founders we spoke to this year expect from 2027. Where a founder’s claim does not match the public data, we say so.

Who this is for: investors, founders, policy staff and journalists who want a twelve-year, founder-level read on where crypto effort has moved, and who will leave able to name the 2026 companies and people behind each trend rather than only the trend itself.

The CCS archive by the numbers

The archive at a glance

The on-site archive holds 1,084 published episodes after a 2026 cleanup that removed duplicates, short summary cuts and a small number of guests the desk no longer stands behind; the channel has recorded more than 1,500 conversations in total. Output peaked in 2022 at 221 episodes, more than four a week, before falling by more than half in 2023 as the bear market emptied the pipeline. The 2024 to 2026 run rate has settled at roughly 85 episodes a year. The dataset was frozen on September 29, 2026 with 73 episodes for the year; three more were published between that date and October 6, taking the live count of 2026 episode pages to 76. The topic shares below use the September 29 cut.

The archive spans 995 named guests and 895 companies. Repeat appearances are rare: 99 guests, one in ten, have appeared more than once, and only one company, XYO Network, has been on the show ten times. That breadth is the point of the dataset. It is not a panel of the same twenty people; it is a rolling census of whoever was raising, launching or lobbying that month.

What founders talked about, year by year

Every episode carries one or more of seventeen topic tags. The table shows the share of each year’s episodes carrying a tag, so a row reads as “what proportion of that year’s founders were working on this.” Because most episodes carry two or three tags, columns sum to more than 100 percent. Cells shaded green or red mark a swing of eight points or more against the prior year.

Four shifts the data makes plain

AI went from absent to the single largest theme in three years

From 2018 through 2023, between two and four percent of episodes touched AI at all, and those were mostly data-marketplace and oracle projects using the word loosely. In 2024 the share jumped to 21 percent, then 35 percent in 2025 and 41 percent so far in 2026. No other tag in the archive has moved that far that fast. The projects behind it are agent frameworks, decentralized compute and data provenance networks, and miners converting sites to GPU hosting, which is why the DePIN and Infrastructure tag rose alongside it to a record 32 percent.

The 2026 episodes show what the tag now means in practice. The question has moved from “can a model use blockchain data” to “can a blockchain prove what an agent did.” XYO’s Arie Trouw framed it that way in September 2026: proving the data was right is no longer enough, the harder problem is proving the agent acted. NEAR’s Illia Polosukhin spent his June episode on an agent harness called IronClaw and a wallet upgrade to post-quantum cryptography, and his October return on private inference for agents. 0G Labs’ Michael Heinrich argued in May that Ethereum and Solana were not built for AI workloads at all. World Foundation’s DC Builder, in June, described the inverse problem: proving a human, not an agent, is on the other end. Four companies, four different layers of the same stack, and all four booked themselves onto the show within five months of each other.

Tokenization is the second-fastest riser and it came from the institutional side

Real-world asset and tokenization projects were a steady ten percent of the archive through 2020, sank to four percent at the bottom of the 2023 bear market, and have since climbed to 19, 25 and now 38 percent. The 2018 cohort was security token platforms pitching retail; the 2025 and 2026 cohort is tokenized Treasury funds, private credit and settlement infrastructure pitching institutions. Regulation and Policy tracks the same curve, at 25 percent in 2026 against eight percent two years earlier, and Stablecoins and Payments is at a nine-year high of 22 percent. Together those three tags describe an industry whose founders are now selling to banks and regulators rather than to each other.

The clearest 2026 evidence is who the tokenization guests were. In January it was a start-up, NSDQ, putting a Nasdaq-100 claim on-chain under Regulation D and S exemptions. By September it was Robinhood, a listed US broker with its own chain, reporting 190 tokenized stock assets and more than 70 percent of their trading outside exchange hours. In between, Trust Wallet added tokenized US equities to a 220-million-download self-custody wallet, Cap put private credit underwriting on-chain with Franklin Templeton behind the principal protection, and Symbiotic argued that the 60 to 180 day redemption windows on most tokenized RWAs are the real product gap. The market data agrees on direction: Robinhood Chain held about $834 million of DeFi deposits on September 4, 2026 according to DefiLlama figures reported by Decrypt, and Hyperliquid’s HIP-3 markets for stocks, commodities and indices had passed $309 billion of cumulative volume by July 12, 2026 according to Crypto Briefing.

The 2021 to 2022 cycle has fully unwound

Web3 and NFT projects were half of all episodes in 2022. They are seven percent in 2026. Gaming and Metaverse went from 19 percent to zero over the same period. DeFi, which peaked at 54 percent in 2021, is still a substantial 27 percent, but its composition has changed from yield farms and new automated market makers to lending infrastructure, perpetuals venues and protocols positioning for institutional collateral. The archive suggests the consumer-speculative cycle did not shrink; it left, and its founders were replaced by a different population.

Even the survivors of that cycle came back talking about something else. Yaniv Tal, who co-founded The Graph and left it in 2022, returned in September 2026 to argue that decentralized social failed because every app built its own data model, and launched Geo Debates as a knowledge-graph answer. Brave, whose BAT token dates from 2017, sent two executives in September to talk about self-custody stablecoin payments for a browser with more than 120 million monthly users, and the word NFT did not come up. Where the old tags do survive in 2026, as with Rip.xyz fractionalizing Hypurr NFTs into Hyperliquid vault strategies, the NFT is a wrapper for a yield product, not the product.

Bitcoin is a constant, not a trend

Bitcoin-focused episodes have oscillated between five and 26 percent with no directional trend, spiking in 2020 (halving and corporate treasuries) and 2023 (ordinals and the ETF filings) and settling at 11 percent in 2026. Institutional and Investing shows a similar shape: 22 percent in 2018 when the guests were funds and OTC desks, six percent at the retail peak of 2021, back to 15 to 18 percent since 2024. The interview desk’s own repositioning toward institutional coverage in 2025 and 2026 accounts for some of that recovery and should be read as part of the signal, not separate from it.

The 2026 Bitcoin episodes are a good illustration of why the tag does not trend: they are about Bitcoin as collateral and settlement, not Bitcoin as a thesis. Michael Terpin opened the year in January arguing the four-year cycle is intact and that 2026 would be a “Bitcoin Fall” with capitulation somewhere between $60,000 and $75,000; Bitcoin traded near $67,000 in early June 2026 and had recovered to around $84,000 by October 1, 2026 according to Kalshi News, against an all-time high above $126,000 in early October 2025. The rest of the year’s Bitcoin guests were builders: OP_NET’s Samuel Pat on native Bitcoin smart contracts at Consensus, Arch Lending on borrowing against holdings, GoMining on merchant payments and NiceHash on the hashrate marketplace.

The companies building it: 2026 on the show

Seventy-six episodes is too many to profile, so this section picks the companies that best explain the topic shares above, grouped by the theme they belong to. For each one: what they build, what the founder told CCS, one or two facts we could verify outside the interview, and why it matters for the Index. Guests are credited by name and role as given on the episode page. Where a figure comes only from the guest, we say so.

AI and verification: XYO Network, NEAR Protocol, 0G Labs, World Foundation

XYO Network. XYO is the archive’s most frequent company and appeared five times in 2026, which is itself a data point about where the AI tag is concentrated. Co-founder Markus Levin opened the year in February describing XYO as a “reality engine” for location and sensor data, said the network had passed 10 million nodes around the third quarter of 2025, and returned in June with the Theta Network partnership for recording AI agent responses immutably and in August with an AI SDK that had 2,000 early-access sign-ups. Co-founder and CEO Arie Trouw covered the XYO Layer One throughput upgrade of two to five times in April, and in September made the argument that frames the whole AI section of the archive: “There’s a big difference between won’t and can’t,” meaning a cryptographic constraint on an agent is worth more than a corporate promise. The 10-million-node and SDK figures are the company’s own; we could not independently verify them. XYO Network is a sponsor of Crypto Coin Show, which is disclosed at the end of this report.

NEAR Protocol. Co-founder Illia Polosukhin appeared twice in 2026, and the two episodes bracket the year’s shift from building for agents to shipping for them. In June he laid out post-quantum cryptography coming to NEAR as a one-click wallet upgrade, dynamic resharding, NEAR Intents (which he said was serving more than 30 percent of daily trading volume routed through the protocol) and IronClaw, an agent harness for financial operations. In October he returned with privacy by default on trading, confidential perpetuals, a Shield system he said had blocked movement of stolen funds, and private AI inference through IronClaw. Verified outside the show: Bitwise’s NEAR ETF, ticker NRR, listed on NYSE Arca on September 29, 2026 with a 0.75 percent fee and an intention to stake all of the fund’s NEAR, according to Bankless and the fund’s S-1 filings with the SEC. For the Index, NEAR is the case where the AI tag, the Security and Privacy tag (21 percent in 2026, a nine-year high) and the Institutional tag land on one episode.

0G Labs. Co-founder Michael Heinrich told CCS in May that general-purpose chains were not designed for AI workloads, that more than 100 applications were testing on 0G, and that an agent payments protocol called X42 was processing roughly $600 million of what he put at $33 trillion of stablecoin volume. The company’s own figures are unverified; the independent facts are that 0G’s Aristotle mainnet launched in September 2025, that the company reports $290 million of backing, and that its $20 million Apollo AI accelerator, run with Stanford’s Blockchain Builders Fund, Google Cloud and Privy, graduated ten teams on July 29, 2026, five of which were live on or integrating 0G infrastructure by August, per GlobeNewswire and 0G’s own blog. Ashton Addison holds a token warrant in 0G Labs; see the disclosure below.

World Foundation. Research engineer DC Builder gave the archive its clearest statement of the identity problem in June: the internet never had a native identity layer, phone biometrics start producing collisions at 15 to 30 million users, so World built custom imaging hardware and zero-knowledge proofs instead, with orbs in around 20 countries. The product that matters for the AI tag is AgentKit, which lets a verified human delegate a World ID to an agent. Verified: AgentKit launched in beta on March 17, 2026 integrated with Coinbase’s x402 agent-payments protocol, and World claimed nearly 18 million verified humans across more than 160 countries at that date, according to The Block. World is the counterweight to XYO and NEAR in the Index: the same agent economy, approached from proof of personhood rather than proof of action.

Payments: Higlobe, OpenPayd, Ramp Network, Brave

Higlobe. Co-founder and CEO Teymour Farman-Farmaian told CCS in April that cross-border transfers still cost around six percent and take five days through correspondent banking, that Higlobe settles dollar transfers in under 60 seconds, and that the company earns on yield, cards, loans and FX rather than transfer fees, with an India launch announced on the show. He named six countries served: Argentina, Colombia, Brazil, Mexico, the Philippines and India. Verified: Higlobe reported passing $100 million in annualized volume in October 2024 while serving Mexico and Brazil, with Argentina and the Philippines next, and raised $14 million in September 2022, per the company’s releases on Silicon UK and Business Wire. The 2026 country list and transfer time are the guest’s figures. Higlobe is the archetype of the 2026 Stablecoins and Payments guest: a FinCEN-registered money services business that happens to settle on stablecoins, not a crypto company that happens to do payments.

OpenPayd. Chief Commercial Officer Lux Thiagarajah described in March an infrastructure company bridging SWIFT rails and stablecoin settlement across 38 currencies through one API, and said that since the GENIUS Act non-crypto enterprises had become the growth market, with one-leg FX trades increasingly using a stablecoin on the dollar side. He cited FX costs of around five basis points against 1.75 percent at a typical bank and 15-minute crypto-to-euro conversion; both are the guest’s figures. OpenPayd’s relevance to the Index is as a European voice in a Regulation and Policy tag that the original report noted was no longer purely American: his episode spent as long on licensing fragmentation and banking partner risk as on product.

Ramp Network. Co-founder and CEO Przemek Kowalczyk told CCS in May that crypto onboarding loses users at two points, card rejection and KYC, that Ramp operates in more than 150 countries with cards, bank transfers, Apple Pay, Google Pay and Pix, that the US and Europe produce 60 to 70 percent of volume with Latin America growing fastest, and that the company had shipped a multichain wallet focused on Base for stablecoin payments. His view that stablecoin adoption “will take years” is a useful corrective to the more breathless 2026 payments narrative. The figures are the company’s.

Brave. Brave sent two executives in two weeks. VP of Business Operations Luke Mulks said in September that Brave had 122 million monthly and 50 million daily users, most of whom never touch crypto, and that Roadmap 4.0 answers that with Brave Pay self-custody payments, a GENIUS Act compliant Brave Rewards Card with a waitlist past 100,000, and x402 support for agent payments. Drew Potter, recorded at the Blockchain Futurist Conference and published September 27, put the figure at 121 million with 3.5 million added in June and spent more time on Brave Search powering third-party AI answers and bring-your-own-model local AI. Verified: Brave reported passing 100 million monthly active users in September 2025 (100.93 million, 42 million average daily), per CyberInsider’s report of CEO Brendan Eich’s announcement on October 1, 2025. The 120-million-plus 2026 figures are Brave’s own but consistent with that trajectory. Brave is in the payments group rather than Web3 because that is where its own executives put it in 2026.

Tokenization: Robinhood Chain, Trust Wallet, Cap, NSDQ

Robinhood Chain. Johann Kerbrat, SVP and General Manager of Crypto at Robinhood, told CCS in September that Robinhood Chain launched on July 1, 2026, had reached $1.6 billion in total value locked and $2.2 billion of asset market cap within three months, offered 190 tokenized stocks with more than 70 percent of their trading outside exchange hours, capped gas at 30 cents, and was expanding to Canada, the UK, Singapore and Indonesia. The public data is lower: Decrypt, citing DefiLlama, reported $834.5 million of DeFi deposits on Robinhood Chain on September 4, 2026, and about $219 million of active RWA market cap, alongside more than 200 million transactions. The gap is partly definitional (Robinhood’s internal TVL counts assets DefiLlama does not) and partly timing, but readers should treat the $1.6 billion as Robinhood’s figure. Verified: the chain is built on Arbitrum’s dedicated-chain framework from Offchain Labs and launched July 1, 2026. For the Index, this episode is why the RWA tag hit 38 percent: the biggest tokenization guest of the year was a brokerage, not a crypto start-up.

Trust Wallet. General Counsel Shehram Khattak said in July that Trust Wallet had passed 220 million downloads across more than 100 chains, had done more than $1 billion of perpetuals volume in the 40 days after integrating Hyperliquid, and had added tokenized US stocks (bStocks), HIP-4 prediction markets and an agent kit, all non-custodial. He was careful on the legal framing: Trust Wallet does not execute orders, the user does. Verified: the Hyperliquid integration went live by April 29, 2026 and the 220-million-download figure appeared in Trust Wallet’s own announcement, per HackerNoon; Hyperliquid’s HIP-3 markets reached $3.68 billion of open interest on July 12, 2026 with non-crypto assets making up seven of the top ten markets, per Crypto Briefing. The $1 billion in 40 days is the company’s figure.

Cap. Founder and CEO Benjamin Sarquis Peillard made the sharpest argument of the year’s credit episodes in May: Celsius and BlockFi failed because crypto teams tried to underwrite credit themselves, so Cap separates depositors, borrowers and underwriters, with underwriters staking collateral that is seized on default and Franklin Templeton providing principal protection. He said only about 20 large crypto borrowers exist today and that 2026 would be about onboarding new borrower types, from high-frequency trading firms to shipment and film financing. We could not verify the Franklin Templeton arrangement independently and present it as the guest’s description. Cap is in this report because private credit is the part of the RWA tag that was absent in 2018 and is now a third of it by our reading of the 2026 episodes.

NSDQ. CEO Kaloyan Iliev’s January episode is the earliest 2026 tokenization interview and the most retail-facing: a security token representing the Nasdaq-100 backed by regulated ETF shares, $500 minimum, 24/7 trading, offered under Regulation D and S with a private sale in January, a security token offering planned for the second quarter and a mobile app for late 2026 or early 2027. None of the roadmap dates could be verified outside the interview and we have not confirmed whether the Q2 offering happened. The episode earns its place as the “before” picture: eight months later Robinhood was offering the same exposure, intraday, to a brokerage account base.

Policy: Input Output, Algorand Foundation, Open Frontier

Charles Hoskinson, Input Output. The April episode is the longest policy conversation in the 2026 archive and the most critical of the Clarity Act. Hoskinson argued that the bill’s process excluded Democrats, that its language would make new projects securities by default, and that the real missing feature for institutional adoption is selective disclosure, which is what his Midnight network is built for. On the product side he discussed Midnight as chain-agnostic privacy infrastructure for Bitcoin, Ethereum, Cardano and Solana, the Glacier Drop distribution, and partnerships with Google Cloud and MoneyGram. Verified: Midnight’s NIGHT token launched on Cardano in December 2025 and a mainnet for late March 2026 was announced at Consensus Hong Kong, per Cexplorer; the Google Cloud and MoneyGram arrangements are the guest’s description. On the politics he was early and, on the September count, right: the Senate failed to invoke cloture on the Clarity Act 49 to 50 on September 15, 2026, with no Democratic votes to proceed, per The Defiant.

Staci Warden, Algorand Foundation. Warden’s March episode is the counter-case to Hoskinson: a foundation that moved to the United States because of the regulatory shift, not despite it. She described re-establishing in the US as a for-profit entity with a restructured board, six years of zero downtime, GENIUS and Clarity as the unlocks that let employees hold the native token and institutions participate, and an Afghani stablecoin for humanitarian aid with a transparency portal. Verified: the Foundation announced its return to the United States with a Delaware headquarters on January 14, 2026, with Abra’s Bill Barhydt as chair and former MoneyGram CEO Alex Holmes, former FinCEN acting director Michael Mosier and Jito Labs CLO Rebecca Rettig joining the board, per PR Newswire. Her line “When people stop talking about the blockchain, we’ll really have arrived” is the Institutional tag in one sentence.

Erik Balsbaugh, Open Frontier. The January episode explains the political mechanics behind the Regulation tag’s climb to 25 percent. Balsbaugh, Executive Director of Open Frontier and a former Democratic organizer, described the group’s goal of making crypto bipartisan, cited Wyoming as the state blueprint, argued stablecoins can cut small-business payment costs from around four percent to around one, and named the 2026 market structure markup as the year’s priority. The figures are the guest’s. Nine months on, the 49 to 50 cloture vote with zero Democrats is the measure of how much of that work remains.

Infrastructure: The Graph and Geo, Optimum, Canopy Network, Fhenix

The Graph Foundation and Geo. The Graph appears twice in 2026 through two different people. Nick Hansen, Team Lead at The Graph Foundation, recorded at Consensus Miami and published May 19, said the network had processed 1.27 trillion queries, had expanded to six data services after the December 2025 Horizon upgrade, was working with institutions on SOC 2 and KYC requirements and had a DTCC collaboration to announce. “Silent but deadly infrastructure. Everybody’s using it, nobody knows who,” was his summary. Yaniv Tal, who co-founded The Graph and left in 2022, appeared in September with Geo, which he self-funded for four years, and Geo Debates, launched September 22, 2026 on an Ethereum layer 3 with video debates, audience voting and permanent on-chain records of claims. The 1.27 trillion queries and the launch date are the guests’ figures; the Horizon upgrade timing matches The Graph’s public roadmap. Together the two episodes show the DePIN and Infrastructure tag at 32 percent is not only compute and sensors: it is data indexing and knowledge graphs aimed at AI agents.

Optimum. Co-founder and CEO Muriel Médard, an MIT professor, argued in March that data propagation, not consensus or execution, is the real scaling ceiling, and that Random Linear Network Coding and a peer-to-peer layer called mP2P can relieve it, with Flex Nodes letting low-compute participants earn for useful relaying. She said an Ethereum mainnet deployment was about a month away at recording. We could not verify a mainnet date; Optimum’s $11 million raise is the company’s most recent independently reported funding. Optimum is in the report as the purest example of the 2026 infrastructure guest: academic, unglamorous and aimed at validators rather than users.

Canopy Network. Co-founder and CEO Adam Liposky pitched Canopy in March as “Replit for Web3”: application chains in about 200 lines of code, written in Python, JavaScript or C#, deployed in under five minutes, with 50,000 daily active users on testnet and 20,000 transactions per second in load testing. Verified: by July 30, 2026 Canopy reported 16,800 testnet forks and more than 100,000 daily active wallets, backed by Arrington Capital, Borderless Capital and HashKey, and announced an integration of Fhenix’s CoFHE coprocessor for the fourth quarter of 2026, per HackerNoon. The earlier figures are the guest’s.

Fhenix. Co-founder and CEO Guy Itzhaki explained in March why Fhenix abandoned its layer 2 plan for a coprocessor model offering fully homomorphic encryption as a service across Ethereum, Arbitrum and Base, with a dual-mode stablecoin that toggles between private and public in what he said was four lines of code, more than eight projects building and mainnet targeted for the end of 2026. Verified: Fhenix raised a $15 million Series A in 2024 per The Block, and its CoFHE coprocessor was live on Ethereum, Arbitrum and Base by July 2026 per the Canopy announcement. Fhenix, Midnight and NEAR’s confidential intents are three different answers to the same institutional objection Hoskinson raised: public ledgers leak too much.

Markets: Rip.xyz and the Hyperliquid ecosystem, Myriad, Forgd

Rip.xyz. Co-founder Solve Maxwell’s February episode is the archive’s window into the Hyperliquid ecosystem, which otherwise appears through integrators like Trust Wallet. He described vaults that take HYPE deposits on HyperEVM and return tokenized strategy exposure, fractionalized Hypurr NFT treasuries through an rHYPURR token, more than $100 million of TVL across implementations, and HIP-3 as the first venue where real-world assets trade as perpetuals on-chain. The TVL is the guest’s figure. The HIP-3 claim held up through the year: $309 billion of cumulative volume by July 12, 2026, a $1.4 billion single-day spike in the SpaceX perpetual on its IPO day in June, and $18.8 billion of stock-linked perpetual volume in the first half of June alone, per Crypto Briefing and The Block.

Myriad. Co-founder and CEO Loxley Fernandes opened the year’s Trading and Markets coverage in January with a prediction market that had passed $100 million in volume, 500,000 sign-ups and 30,000 monthly actives, resolved on-chain through oracles including Chainlink, and integrated into Trust Wallet, Phantom and MetaMask. His argument that the 2024 US election made prediction markets mainstream and that liquidity, not features, drives adoption was echoed six months later when Trust Wallet shipped HIP-4 prediction markets. Verified: Decrypt reported Myriad passing $10 million of USDC volume at an earlier stage, which makes the $100 million figure plausible but it remains the company’s number.

Forgd. Managing Director Scott Byron’s May episode is the most useful one in the archive for anyone planning a token launch. He explained the two market-maker engagement structures (loan plus call option versus retainer plus working capital), why 2025 was brutal for altcoin launches that went to market before product, how to read order book depth and spreads for fragile liquidity, and why a launch needs both centralized and decentralized venues. He said Forgd integrates more than 30 market makers with performance leaderboards and has been used by more than 1,000 projects; those are company figures. Forgd matters to the Index because its existence is evidence that the Trading and Markets tag’s recovery to 15 percent is about market plumbing, not trading calls.

Guests who came back

The original Index noted that only one in ten guests has ever appeared more than once and that the archive’s most frequent guests did most of their appearances before 2023. The 2026 cohort is starting to change that. Four people appeared two or more times in the first nine months of the year, and in each case the second episode reported on something the first one had promised, which is what makes a repeat appearance useful as data rather than as promotion.

Levin’s three appearances are the most instructive because they are evenly spaced. In February the pitch was conceptual: autonomous systems need verifiable inputs or they fail, robots should be owned by individuals rather than corporations, and XYO Layer One exists for data certainty. By June there was a named partner (Theta) and a specific mechanism (off-chain data lakes, on-chain hashes), and by August a product with a sign-up count and a claim of 99.9 percent geolocation accuracy against 94 percent for a competitor. The accuracy figure is XYO’s and unverified, but the progression from thesis to partner to product inside seven months is the shape the desk looks for when deciding whether a company deserves a second booking.

Polosukhin’s pair is the clearest case of a roadmap being checked. In June he said confidential intents and an agent marketplace were “step one” and that the next step was enabling them for agents. In October the confidential layer had become privacy by default, IronClaw had gained private inference, and a Bitwise ETF had listed two days earlier, which is a fact we could verify through the SEC filings and Bankless. Marin’s two episodes do the same for Nexus: the February thesis that Ethereum’s 15-second confirmations cannot host traditional finance, then a May update with testnet numbers, a USDX partnership with the M-Zero protocol and his line that “Most blockchains are fast enough to transact but can’t prove the transaction was valid.” Trouw’s two appearances move from engineering (April) to philosophy (September), and the September episode is the one that best explains why the AI and Security tags rose together this year.

On the road: Consensus Miami and the Blockchain Futurist Conference

Most CCS episodes are remote. Conference recordings are shorter, louder and booked on the floor, which makes them a different kind of sample: who is at the event and willing to sit down, rather than who sent a pitch. In 2026 the archive carries two conference sets.

Consensus Miami, May 5 to 7, 2026. CoinDesk’s flagship event moved to the Miami Beach Convention Center for its 2026 edition, per Alchemy’s event listing. CCS published five episodes from it between May 12 and May 25. The Graph Foundation’s Nick Hansen (May 19) covered 1.27 trillion queries and institutional data requirements. CertiK’s Head of Ecosystem Hudson Jameson (May 25) said CertiK had launched an AI auditor in April 2026 with what he put at an 88.6 percent hit rate, that the firm has around 65 percent of the blockchain auditing market and more than $600 billion of secured assets, that most 2026 attacks happen at the infrastructure level rather than in contracts, and that wrench attacks on key holders are now organized remotely. OP_NET founder Samuel Pat (May 12) argued for native Bitcoin smart contracts without bridges and a “SlowFi” thesis that slower, costlier settlement suits real-world assets. Virgo Group CEO Adam Cai (May 14) presented Wallet V, a self-custody wallet where an AI agent manages the portfolio within technical-analysis constraints and explains its decisions every 15 minutes. CertiK’s market share and secured-asset figures are the company’s; the rest are the guests’ claims.

Blockchain Futurist Conference, Toronto, July 21 to 22, 2026. The conference returned to Toronto’s Rebel Entertainment Complex under event producer Emerald, per HackerNoon’s November 2025 announcement. CCS’s published episode from the floor is Brave’s Drew Potter (September 27), who put Brave at 121 million monthly users with 3.5 million added in June, described Brave Search as a B2B backend for other AI chatbots, and previewed the BAT 4.0 stablecoin and card roadmap that Luke Mulks detailed nine days earlier. The two Brave episodes together are the only case in the 2026 archive of a conference recording and a studio recording from the same company inside two weeks, which is why they are treated as one entry above.

What the conference sets say about the Index: of the six floor recordings, four carry the AI tag (CertiK, Virgo, The Graph, Brave), two carry Security and Privacy, one is Bitcoin and none is Web3 or Gaming. The floor sample matched the inbound sample, which is some reassurance that the topic shares are not an artefact of who emails the desk.

How we got here: a timeline

The archive’s own milestones sit on top of the market cycle, and reading them together explains the episode-count chart. Volume follows price with a lag of about two quarters, and topic mix follows the regulatory calendar.

Where the guests disagree

A dataset of pitches would be worthless if everyone agreed, and the 2026 cohort does not. Four disagreements run through the year.

Is US legislation the unlock or the threat? Warden (Algorand) moved her foundation to Delaware because GENIUS and Clarity made the US investable; Balsbaugh (Open Frontier) spent the year lobbying for the market structure bill; Nexo’s COO Neil Steinhardt told CCS in May that the US market was back. Hoskinson (Input Output) argued the Clarity Act as written would make new American projects securities by default and that the process had been partisan. The September 15 cloture failure, 49 to 50 with no Democrats, is consistent with Hoskinson’s process critique and with Balsbaugh’s priority list; it does not settle whether the bill’s content was good.

Who should hold the keys? Robinhood (Kerbrat) built a chain it controls end to end so it can set compliance rules per jurisdiction and cap fees. Trust Wallet (Khattak) offers the same tokenized stocks and perpetuals non-custodially and insists it does not execute orders. Brave (Mulks) is betting a self-custody wallet can feel like Venmo. Hoskinson told listeners self-custody is the defence against regulatory overreach. The market data cannot adjudicate yet: Robinhood’s chain had the deposits, Trust Wallet had the downloads.

Is the cycle intact? Terpin (January) says the four-year halving cycle survived institutional adoption and that 2026 is the down year with a 2028 high. Almost none of the builders on the show in 2026 mentioned the cycle at all; Warden, Kerbrat, Thiagarajah and Farman-Farmaian described demand curves that do not reference Bitcoin’s price. Bitcoin’s path from above $126,000 in October 2025 to near $67,000 in June 2026 and around $84,000 on October 1, 2026 is consistent with Terpin’s range; the builders’ indifference is consistent with the Index showing Bitcoin at 11 percent of episodes.

Does blockchain need to be private to be institutional? Hoskinson, Itzhaki (Fhenix) and Polosukhin (NEAR) all argued that transparency by default is why institutions keep sensitive flows off-chain. Hansen (The Graph) said institutions were asking for SOC 2 and KYC, not encryption. Jameson (CertiK) wanted privacy at Ethereum’s base layer. The Index records Security and Privacy at a nine-year high of 21 percent, which suggests the privacy camp is winning the booking war even if the demand question is open.

What the 2026 cohort predicts for 2027

We asked most 2026 guests what comes next. Pulling the forward-looking claims together, with the caveat that founders forecast their own roadmaps optimistically, gives a reasonable picture of what the 2027 Index should test.

• Agents as paying users. Polosukhin’s roadmap to 2027 has AI agents using NEAR as programmable money; Levin told CCS in August that most of daily life would be automated “in a year or two”; Mulks and Potter put x402 agent payments in Brave’s 4.0 roadmap; Heinrich’s X42 and World’s AgentKit are built for the same flow. If this happens, the AI tag stays above 40 percent and the Stablecoins tag rises with it. If it does not, 2027 is the year the AI share plateaus.
• Privacy shipping, not promised. Fhenix’s mainnet is targeted for the end of 2026 and its Canopy integration for the fourth quarter; NEAR’s privacy by default is already live per the October episode; Midnight’s mainnet was announced for late March 2026. By the 2027 Index, the Security and Privacy tag should either break 25 percent or reveal that the demand was thinner than the pitches.
• Tokenized equities move to brokerages. Kerbrat named Canada, the UK, Singapore and Indonesia as next markets for Robinhood Chain; NSDQ’s mobile app is due late 2026 or early 2027; Trust Wallet’s bStocks are live. The question for 2027 is whether the RWA tag, at 38 percent, has peaked as the category consolidates into a handful of large venues, or keeps climbing as private credit (Cap) and collateral markets (Symbiotic) join equities.
• Banks surface. Dr Jordan Knecht of Kresus told CCS on October 5, 2026 that banks are running RFPs and testing tokenized deposits out of public view; Warden said bank-issued wallets would be the sign of arrival and put it nearer 2030. The 2027 Index will show whether the Institutional tag, flat at 15 to 18 percent since 2024, finally breaks out.
• A legislative retry. Senator Tillis filed a motion to reconsider the Clarity Act vote on September 15, 2026 per The Defiant, and Polymarket odds of passage in 2026 fell to 6.5 percent. Balsbaugh’s priority of educating candidates points at the midterms; Hoskinson’s advice was to make the industry too big to ignore. Expect the Regulation tag to stay near 25 percent either way.
• The cycle call. Terpin’s framework puts the next cycle high at $250,000 to $300,000 in 2028, with 2027 as the recovery year. The Index does not take price views, but if he is right the episode count chart will show it first: a refilled pipeline in late 2027, two quarters before the tag mix shifts back toward retail.

What the data means for investors and builders

The Index is a demand signal from the founder side. Used carefully, it answers a set of questions that market data alone cannot.

• Is the category crowded on the supply side? AI at 41 percent and RWA at 38 percent of 2026 episodes means more founders are pitching those categories than ever; an investor should expect compressed valuations for undifferentiated entrants and a builder should expect to compete for distribution.
• Is the category empty for a reason? Gaming and Metaverse at zero percent and Web3 and NFTs at 7 percent could mean opportunity or a dead market. The 2026 episodes suggest the founders who stayed, such as Tal at Geo, changed the product rather than waiting for the cycle.
• Does the founder’s number match the public number? Robinhood’s $1.6 billion TVL against DefiLlama’s $834.5 million is the clearest 2026 example. Ask which definition is being used before comparing companies.
• Has the company come back with a result? Repeat guests whose second appearance reports a shipped product (NEAR, XYO, Nexus) are more informative than first appearances. Weight the roadmap check more than the pitch.
• Is the regulatory story a tailwind or the whole thesis? OpenPayd, Higlobe, Algorand and Brave all cite the GENIUS Act as the unlock. If the business only works under one law in one jurisdiction, the Clarity Act’s failure is a reminder that laws stall.
• Who holds custody and who carries the compliance cost? The Robinhood versus Trust Wallet split runs through tokenized equities, perpetuals and prediction markets. The answer determines which regulator a product will meet first.
• Is the infrastructure claim measurable? The Graph’s query count, Optimum’s propagation gains, Canopy’s deployment time and XYO’s throughput are all testable. Prefer the companies that publish the test.
• Is the privacy feature live or on a roadmap? Three 2026 guests promised privacy infrastructure for institutions; one (NEAR) reported it live by October. Mainnet dates slip; check them.

What to watch into 2027

What this means for coverage in 2026

Three implications follow. First, the founder pipeline is now dominated by AI-adjacent infrastructure and tokenization, and a news desk that still allocates coverage by 2022’s map is covering the wrong industry. Second, the regulation tag’s climb is not only US market structure legislation; a growing share of 2026 guests are European and Asian companies structuring around MiCA and Hong Kong and Singapore regimes, which is a distribution question for any English-language outlet. OpenPayd, Ramp, VALR and the offshore-structuring episode with Dave Rodman are the 2026 examples. Third, the collapse in repeat guests after 2022 (the archive’s most frequent guests all recorded most of their appearances before 2023) means relationships have to be rebuilt with a new cohort, and the four 2026 repeat guests above are the first evidence that this is working. The company hub pages on this site are the desk’s answer to that.

Glossary

Topic share

The percentage of a year’s published episodes carrying a given topic tag. Because episodes carry two or three tags, shares in a year sum to more than 100 percent.

Pipeline

The queue of inbound interview requests from projects and their PR firms. The Index treats it as a demand signal because it reflects who is raising, launching or lobbying.

Agent harness

A controlled environment, such as NEAR’s IronClaw, in which an AI agent can hold keys and execute financial operations under constraints.

Proof of action

A cryptographic record that an autonomous agent or robot performed a claimed task, as distinct from proof that its input data was correct. Arie Trouw’s framing on CCS in September 2026.

Data lake (XYO)

Off-chain storage for full data sets whose hashes are recorded on XYO Layer One, so provenance can be proven without putting the data itself on-chain.

HIP-3

A Hyperliquid protocol upgrade, live since October 2025, that lets third parties deploy perpetual markets on non-crypto assets such as stocks, commodities and indices.

Fully homomorphic encryption (FHE)

A technique for computing on encrypted data without decrypting it. Fhenix offers it as a coprocessor service to EVM chains.

Selective disclosure

Revealing only the specific facts a counterparty or regulator needs while keeping the rest private; the design goal of Midnight and a recurring institutional requirement in 2026 episodes.

x402

A Coinbase-developed protocol for machine-to-machine stablecoin payments over HTTP, referenced by Brave, World and 0G guests as the rail for agent commerce.

Cloture

The Senate procedure requiring 60 votes to end debate and move to a final vote. The Clarity Act failed cloture 49 to 50 on September 15, 2026.

Dedicated chain (Arbitrum)

A purpose-built layer 2 run by one operator using Offchain Labs’ framework; Robinhood Chain is one.

Four Seasons cycle

Michael Terpin’s framework mapping Bitcoin’s four-year halving cycle to Spring (halving), Summer (new high), Fall (bubble pop) and Winter (capitulation).

Methodology and limitations

Sources

1. Crypto Coin Show: Illia Polosukhin on NEAR’s privacy features and AI agents in 2026, October 1, 2026
2. Crypto Coin Show: How Robinhood built its own blockchain to control compliance and fees, September 29, 2026
3. Crypto Coin Show: Arie Trouw (XYO) on why AI agents now need proof of action, September 14, 2026
4. Crypto Coin Show: Charles Hoskinson on why the Clarity Act misses crypto’s real needs, April 24, 2026
5. Crypto Coin Show: Staci Warden on Algorand’s US return and blockchain regulation in 2026, March 5, 2026
6. Crypto Coin Show: The Graph Foundation’s Nick Hansen on processing 1.27 trillion queries (Consensus 2026), May 19, 2026
7. Crypto Coin Show: Michael Terpin on whether 2026 could be the next Bitcoin winter, January 23, 2026
8. The Defiant: Senate blocks CLARITY Act 49-50 with no Democratic votes to proceed, September 15, 2026
9. Bankless: Bitwise teases NEAR ETF launch for today, September 29, 2026
10. SEC EDGAR: Bitwise NEAR ETF, Form S-1/A, September 16, 2026
11. Decrypt: What is Robinhood Chain? The Ethereum layer 2 network for tokenized stocks, September 5, 2026
12. Kalshi News: Bitcoin odds, how high will BTC go in October, October 1, 2026
13. Blockchain.News: Bitcoin analysts split on October 2026 bottom, June 3, 2026
14. ForkLog: Bitcoin surpasses $125,000 to set new all-time high, October 5, 2025
15. 0G Labs: Apollo Accelerator 2026 graduation recap, August 13, 2026
16. GlobeNewswire: 0G and Stanford blockchain veterans launch $20M Apollo AI Accelerator, February 26, 2026
17. The Block: World adds identity toolkit for AI bots on Coinbase’s x402 protocol, March 17, 2026
18. HackerNoon: Trust Wallet brings the perp DEX war to mobile with Hyperliquid integration, April 29, 2026
19. Crypto Briefing: Hyperliquid’s HIP-3 markets hit $4B open interest as on-chain trading grows, July 13, 2026
20. The Block: SpaceX IPO fuels $1.4 billion trading frenzy on Hyperliquid, June 15, 2026
21. HackerNoon: Canopy and Fhenix want to make encrypted data the default state of onchain apps, July 30, 2026
22. The Block: Ethereum layer 2 Fhenix raises $15 million in Series A funding, 2024
23. PR Newswire: Algorand Foundation announces return to the United States with new board, January 14, 2026
24. Cexplorer: Cardano’s privacy partner chain Midnight to launch mainnet this March, February 26, 2026
25. CyberInsider: Brave browser has now surpassed 100 million monthly active users, October 1, 2025
26. Silicon UK: Higlobe surpasses $100 million in annualized volume, October 22, 2024
27. Alchemy: Consensus Miami 2026 event listing, 2026
28. HackerNoon: Blockchain Futurist Conference announces 2026 Toronto dates, July 21 to 22, November 5, 2025
29. Decrypt: Myriad hits $10M USDC trading volume as prediction markets become new segment of DeFi, 2025

Appendix: 2026 episodes cited in this report

Disclosure: XYO Network is a sponsor of Crypto Coin Show. Ashton Addison, founder of Crypto Coin Show, holds a token warrant in 0G Labs. Neither had any input into this report. This report is for education only and is not investment, legal or tax advice.