Bitwise's NEAR ETF draws $50M in first three days as hosts tout AI-crypto convergence thesis

Bitwise launches NRR after years of behind-the-scenes NEAR involvement

Hunter Horsley of Bitwise joined the show alongside Sal Ternullo of Sovereign to discuss Bitwise's newly launched NEAR ETF, ticker NRR. Horsley said Bitwise's relationship with NEAR predates the ETF by over a year, starting with institutional research and valuation coverage, a European ETP, and direct network participation as what he said is currently the fourth-largest validator on NEAR, as well as work on NEAR Intents through an MPC network.

Horsley framed Bitwise's hands-on approach as aligned with client interests: "Bitwise manages something like $16 billion... it uh is long bias meaning um at the highest level the best thing for Bitwise for our clients is that Near or other crypto assets and ecosystems do well." Ternullo praised Bitwise's early research coverage, noting it came when NEAR was trading around $130, well below recent levels he says touched $55 before settling near $5 just before the recording.

Early ETF demand looks strong relative to other 2026 launches

Horsley said NRR's first three days brought over $10 million in daily inflows and volumes ranging from $10 million to $20 million a day, which he characterized as extraordinarily strong compared to other single-asset crypto ETFs launched this year, including unnamed ones with larger market caps that saw less volume and fewer inflows. He noted he was still awaiting Friday's results at time of recording.

He attributed the strength partly to timing — citing the Hyperliquid ETF's successful launch even in a bear market as precedent — and partly to growing investor interest in AI narratives connecting to NEAR, plus Bitwise's efforts to raise awareness.

NEAR Intents volume and buybacks, not burns

Ternullo described NEAR Intents as a framework where users express a desired outcome and a market of solvers fulfills it, now expanded beyond simple digital-asset swaps to include tokenized equities via Ouno and roughly 400 tokenized equity products. He said the system has processed more than $32 billion in volume, driven largely by human interactions through near.com and B2B wallet partners like Ledger and Zashi.

He clarified that net revenue from Intents funds buybacks of NEAR tokens held in a "permanent capital vehicle," not burns, cautioning that Bitwise's prior report calling this "deflationary" needs correction since the tokens aren't removed via burning. Separately, he mentioned a governance forum proposal to cut inflation from 2.5% to 1.6% over a two-year linear curve, and raised — without committing to — the idea of researching a fixed-supply NEAR.

NRR's staking feature and the value-capture pitch to mainstream investors

Horsley said NRR handles staking on behalf of investors, removing a decision point. He argued investors in 2026 have "moved on from the coinmarketcap era" and no longer rank assets by ranking alone; instead they respond to a value-capture story. His pitch to mainstream investors: NEAR was created by "the co-inventor of LLM" and lets both individuals and AI agents move assets across a blockchain while scaling rapidly.

He compared the sector's maturation to early-2000s internet firms shifting from "page views and eyeballs" to metrics like ARPU and retention, saying NEAR shows strength across usage, revenue, and value-capture dimensions.

Betting on fragmentation across AI models and blockchains

Both guests argued the future holds many models and many chains rather than one dominant system. Horsley said: "we're probably going to wind up in a world with uh multiple uh L1s um multiple different stable coins," and argued NEAR is positioned as "a simplifying force" letting agents or humans traverse different blockchains and models. Ternullo tied this to Ilia's original vision, predating NEAR's mainnet, that centralized AI intelligence would not produce a "utopian outcome," motivating his decision to found Sovereign.

The host drew a parallel to a conversation with Venice's John, framing a split between "Frontier Labs" chasing one dominant model versus a "biodiversity of models." Ternullo agreed, describing enterprise adoption of tiered model structures — frontier intelligence for some tasks, open-source confidential models (like NEAR AI's private inference) for sensitive data — as a trend he expects to become a "fundamental requirement" within two to three years.

Privacy as a growing investor and consumer theme

Horsley questioned the old Silicon Valley assumption that consumers claim to value privacy but don't act on it, suggesting recent anxieties about data feeding into LLMs may be shifting real preferences. Ternullo said there is "empirical data on this now," citing Google search trends showing privacy-related searches at a multiple of past peaks, which he compared to the shape of NEAR Intents' volume growth. Horsley linked this to Bitcoin's original sovereignty thesis, calling privacy "the exact same concept just applied to your information." Neither speaker cited specific figures for the search-trend claim beyond the general comparison.