Crypto Long & Short: What will the AI agents run on

You are reading Crypto Long & Short, our weekly newsletter featuring insights, news and analysis for the professional investor. Sign up here to get it in your inbox every Wednesday.

Note: The views expressed in this column are those of the author and do not necessarily reflect those of CoinDesk, Inc., CoinDesk Indices or its owners and affiliates.

Happy Wednesday,

This is your institutional newsletter, Crypto Long & Short. This week:

Utkarsh Ahuja of Moon Pursuit Capital on what an economy of autonomous AI agents would need from financial infrastructure
Top headlines institutions should pay attention to
“BP rallies while Backpack’s share of tokenized-stock volume slips to Robinhood and bStocks” in Chart of the Week

Thanks for joining us!

– Kim Klemballa

The AI economy will need blockchain rails

By Utkarsh Ahuja, founder and managing partner, Moon Pursuit Capital

The discussion surrounding artificial intelligence is currently dominated by a familiar inquiry: Which companies will emerge victorious? Market participants are striving to locate the upcoming leading model, application or AI platform, and massive amounts of funding are being allocated accordingly. However, I believe there is an alternative investment question that warrants substantially more focus. If artificial intelligence attains the economic significance anticipated by markets, what financial infrastructure will that economy actually operate upon?

Increasingly, I am convinced the answer will incorporate blockchain and digital assets. That does not imply every artificial intelligence enterprise requires a token, nor does it signify that appending “AI” to a cryptographic initiative instantaneously generates value. The more compelling opportunity lies underneath both industries, where autonomous software begins interacting with financial systems and demands infrastructure capable of transferring value with the same speed and programmatic nature as it moves information.

We are progressing toward an era where AI agents will not merely generate text or evaluate data. They will negotiate with fellow agents, purchase computational capacity, pay for information, execute transactions and govern financial choices within boundaries defined by humans. Conventional payment infrastructure was formulated around people and institutions initiating transfers. It was never built for potentially millions of autonomous software agents executing low-value transactions continuously across borders.

Blockchain infrastructure is exceptionally well suited for such an environment because money can transform into something programmable. An AI agent can interface with a wallet, execute a smart contract or transfer a stablecoin without necessitating the identical layers of manual intervention tied to legacy financial architecture. Stablecoins hold particular significance here because they supply a bridge between the programmability of blockchains and a familiar unit of account.

This is precisely where I suggest investors expand their perspective on the AI trade. The opportunity is not restricted to owning enterprises constructing models or applications. Should artificial intelligence generate substantially greater machine-to-machine economic engagement, there will likewise be a surge in demand for the infrastructure permitting those machines to transact, verify identity, establish ownership and exchange value.

Identity serves as a fitting illustration. As autonomous agents grow more sophisticated, markets will require mechanisms to discern who or what stands behind a transaction and what permissions an agent holds. The same principle applies to provenance. When AI systems consume data, formulate intellectual property or complete transactions, determining the origin of information and the rightful owner gains heightened importance. Blockchains will not resolve every facet of these challenges, but their capacity to supply verifiable, collective records positions them as an organic component of the infrastructure stack.

There is also a capital-markets dimension that I feel tends to be neglected. Tokenization is already introducing traditional assets onto blockchain rails, whereas stablecoins have proven that meaningful financial operations can function onchain. Concurrently, artificial intelligence is rendering financial decision-making progressively automated. Those movements are currently evolving independently, yet their convergence could prove profound. An AI agent capable of analyzing markets yet unable to efficiently store, exchange or settle assets remains inherently limited. Combine programmable intelligence with programmable assets, and the possibilities become vastly more compelling.

For investors, this demands a degree of discipline since technological convergence does not imply every venture exposed to artificial intelligence and crypto warrants a premium valuation. We observed throughout preceding crypto cycles how rapidly narratives could outpace fundamentals, and identical risks persist surrounding AI presently. I prefer seeking out infrastructure resolving an authentic problem spawned by heightened automation: payments, settlement, identity, cybersecurity, custody and the tracks connecting legacy and digital markets.

I approach this primarily as an investor and trader, and what captivates me is where structural transformations forge novel capital flows and economic undertakings. Artificial intelligence is positioned to be among those shifts, yet investors ought not presume all value will accumulate at the application tier. Some of the most substantial opportunities could arise from the financial infrastructure necessary to make a progressively autonomous economy operational.

The internet revolutionized how information travels, whereas blockchain modified how value moves. Artificial intelligence is now altering who, or progressively what, can dictate economic decisions. Investors must monitor much more closely what transpires when those three transformations intersect.

Headlines of the week

The prevailing motif this week involves traditional finance integrating itself into crypto infrastructure from both directions. Goldman Sachs directed roughly $100 billion in a Treasury fund to digital-asset trading firms devoid of tokenization, whereas Cboe Global Markets and S&P Dow Jones Indices preserved space to investigate tokenized options under an extended licensing pact.

Goldman Sachs brings $100 billion Treasury fund into crypto’s institutional plumbing: Goldman’s FTIXX Treasury fund reaches institutional crypto entities via Lynq, a settlement network utilized by digital-asset businesses, without undergoing tokenization.
Cboe and S&P Dow Jones Indices may explore tokenized options: A newly prolonged licensing accord between the pair maintains scope for tokenized option offerings, presenting another indication that blockchain-based instruments are drawing closer to established derivatives markets.
Buterin sketches a different Ethereum for 2030: In a publication titled, “The cryptographic world computer,” Vitalik Buterin outlined pairing the blockchain alongside cryptographic proofs and networks of computers functioning externally.
Bitcoin recovers above $84,000 as Treasury yields steady: Bitcoin dropped to $82,500 on Monday prior to reclaiming $84,000 overnight, as yields stabilized following the 10-year Treasury yield’s brief surge to 5.2%. Spot bitcoin ETFs attracted $30 million on Tuesday — a modest reading, although positive, succeeding the $2.84 billion collected across six sessions the week prior.

Chart of the Week

BP rallies while Backpack’s share of tokenized-stock volume slips to Robinhood and bStocks

Tokenized-equity decentralized exchange (DEX) volume is consolidating toward bStocks and Robinhood while Backpack’s portion contracts to approximately 2%. Nevertheless, BP continues its ascent, nearing $1.35, with the market paying a premium for the sole licensed, tradable token proxy representing the theme.

Looking for more? Receive the latest crypto news from coindesk.com and market updates from coindesk.com/institutions.