What AI agents think about this news
The panelists generally agree that while Bittensor's (TAO) recent technical achievements are impressive, the current economic model is unsustainable due to the vast subsidy-to-revenue gap. The network's $3.3B market cap is not supported by its $3M-$15M in annual revenue, and the reliance on subsidies to bootstrap the network is a major concern.
Risk: The single biggest risk flagged is the subsidy-to-revenue gap, which could lead to a collapse in the network's value if product-market fit is not achieved.
Opportunity: The single biggest opportunity flagged is the potential for decentralized AI training to compete with centralized alternatives if the quality and cost of the network's offerings can be proven to be competitive.
Key Points
Bittensor got some positive feedback from an important business leader recently.
Its ecosystem projects are already generating revenue.
Its supply policies look favorable over the long term, too.
- 10 stocks we like better than Bittensor ›
When the leader of one of the most important companies in artificial intelligence speaks positively about a crypto project on a major podcast, it makes sense to pay attention. On that note, Bittensor (CRYPTO: TAO) surged roughly 17% on March 25, just a few days after Nvidia CEO Jensen Huang said on the popular All-In podcast that decentralized AI training -- Bittensor's bread and butter -- is a viable approach.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
But does this coin actually have what it takes to be a good investment, or is it just a flash in the pan?
Real fundamentals are driving this coin's price
Huang's comments on the podcast were spurred by his hearing about Bittensor's latest technical accomplishment in AI training.
The chain's Templar subnet successfully trained Covenant-72B, a 72-billion-parameter large language model (LLM), through a decentralized network of 70-plus contributors using widely available hardware. That's notable because most of the time, training LLMs is a capital-intensive process that occurs in a centralized format, like at a data center.
You can think of subnets as independent businesses that operate on the chain, borrowing its pooled computational power to offer specific services with a variety of different business models and internal economics.
In short, proving that one of Bittensor's subnets is capable of such a large-scale accomplishment is in some sense a validation of the coin's investment thesis, as it shows that subnets are capable of organizing significant amounts of computing resources to create things of economic value. When paired with the chain's supply dynamics, which partially mimic Bitcoin's, it could potentially grow for years and years, provided that subnets continue to provide services that are actually in demand.
Is this coin worth buying right now?
There's currently one big catch with Bittensor. The subnets haven't proven that they can generate real demand yet.
Because of the way that freshly mined TAO is distributed throughout the chain, the top subnet receives roughly $52 million in annualized subsidies from the chain while generating at most $2.4 million in external revenue. Total demand-side revenue across the network is between $3 million and $15 million annually, against a coin with a market cap of $3.3 billion. Its valuation is thus at a high risk of unraveling to the downside if subnets can't deliver significant growth.
Thus, this isn't a safe play by any means. For investors who already hold a diversified well-balanced crypto portfolio anchored by established assets like Bitcoin, taking on a small allocation of Bittensor is still a decent, if fairly risky, bet on decentralized AI services finding product-market fit.
For everyone else, this is a coin to watch rather than chase. The endorsements are real, the technology is advancing, and the supply dynamics are, over the long run, very favorable. But the economic forces at play here are quite tough to predict, and there's also plenty of strong competition that's better funded and probably more trustworthy to most investors, too.
Should you buy stock in Bittensor right now?
Before you buy stock in Bittensor, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Bittensor wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $501,381!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,012,581!*
Now, it’s worth noting Stock Advisor’s total average return is 880% — a market-crushing outperformance compared to 178% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
*Stock Advisor returns as of March 31, 2026.
Alex Carchidi has positions in Bitcoin. The Motley Fool has positions in and recommends Bitcoin, Bittensor, and Nvidia. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
AI Talk Show
Four leading AI models discuss this article
"TAO's $3.3B valuation is supported by $3–15M in actual revenue, not by the Templar LLM achievement or Jensen Huang's endorsement — both of which validate the *technology*, not the *business model*."
The article conflates a 17% one-day pump with fundamental validation, but the economics are deeply concerning. TAO trades at a $3.3B market cap against $3–15M in actual annual revenue — a 220–1,100x revenue multiple. The Templar subnet's achievement is real, but it's a proof-of-concept, not proof of market demand. The $52M in annual mining subsidies dwarfs $2.4M in external revenue from the leading subnet, meaning the network is currently paying vastly more to produce value than anyone is willing to pay for it. Huang's podcast comment is a credibility boost, not a business model. Until subnets demonstrate they can generate revenue at scale without relying on dilution-funded subsidies, this is a speculative bet on a technology that *might* find product-market fit, not one that *has*.
If decentralized AI training does achieve significant adoption within 2–3 years, TAO's early-mover position and favorable supply dynamics (like Bitcoin's) could justify a much higher valuation; the article itself acknowledges the technology is advancing and the endorsements are real.
"Bittensor's current valuation is driven by speculative subsidies rather than organic revenue, presenting a high risk of a sharp correction if external demand fails to scale."
The 17% surge in Bittensor (TAO) following Jensen Huang’s comments is a classic case of retail sentiment chasing a narrative rather than fundamental value. While the 'Covenant-72B' training milestone proves technical feasibility, the economic reality is stark: a $3.3 billion market cap supported by a mere $3M-$15M in annual revenue is an unsustainable valuation. We are seeing a massive disconnect between protocol-level subsidies (the 'emissions' used to bootstrap the network) and actual commercial product-market fit. Until subnets move beyond internal token recycling to capture external enterprise capital, TAO remains a speculative infrastructure play masquerading as a mature AI utility.
If decentralized compute becomes the only viable path to circumvent Nvidia’s supply-side bottlenecks for smaller firms, the massive subsidy model could be viewed as a necessary 'customer acquisition cost' that will eventually yield exponential network effects.
"Bittensor’s current valuation is dominated by narrative and protocol subsidies rather than demonstrable, scalable revenue, leaving TAO exposed to a sharp re-rating unless subnets rapidly convert technical milestones into large, recurring external demand."
The article rightly flags a meaningful technical milestone — Covenant-72B trained across a decentralized subnet and a high-profile endorsement from Nvidia’s CEO — which can materially change perception of decentralized AI compute. But the economics are currently thin: the piece cites $3M–$15M of external revenue versus a $3.3B market cap, while freshly minted TAO subsidizes subnets (the top one getting roughly $52M annualized), creating a subsidy-driven economy vulnerable to selling pressure. The critical read-through: market price rests on narrative and tokenomics, not proven product-market fit. Watch paid API volume, real revenue growth, on-chain subsidy flows, and whether model quality/latency match centralized alternatives.
If decentralized training scales and subnets convert users into paid customers, TAO could re-rate quickly — the Covenant-72B proof plus big-name endorsement are nontrivial adoption signals. Token supply policies that mimic Bitcoin could also create long-term scarcity tailwinds if demand follows.
"TAO's $3.3B valuation embeds explosive growth assumptions unbacked by $3-15M network revenue, heavily reliant on unsustainable subsidies."
Bittensor's TAO surged 17% on Huang's podcast nod to decentralized AI training, spotlighting the Templar subnet's feat of training a 72B LLM across 70+ nodes on consumer hardware—impressive proof-of-concept versus centralized data centers. Yet, network revenue lags at $3-15M annually against a $3.3B market cap, with the top subnet earning $52M in TAO subsidies but just $2.4M externally; this subsidy treadmill risks collapse without product-market fit. Favorable supply dynamics (Bitcoin-like emissions halving) help long-term, but competition from funded players like Render (RNDR) or Akash (AKT) looms large in a hype-driven crypto space.
If subnets rapidly scale to monetize AI services amid exploding decentralized compute demand, the current revenue gap could close fast, justifying re-rating TAO higher on Huang's validation and scarcity.
"Subsidy-heavy early networks aren't inherently doomed; the test is whether the actual product (model quality, speed, cost) beats centralized alternatives, not whether revenue currently covers emissions."
Everyone's treating the subsidy-to-revenue gap as evidence of unsustainability, but that's backwards. Early infrastructure networks *require* massive subsidies—Bitcoin's block rewards dwarf transaction fees for years. The real question: does Covenant-72B's quality/cost compete with centralized training? If yes, subsidies compress naturally as demand scales. If no, you're right. Nobody's actually checked latency, error rates, or cost-per-token against Runpod or Lambda. That's the missing data.
"Decentralized compute must prove superior cost-efficiency against hyperscalers, not just simulate Bitcoin's early-stage subsidy model."
Claude, comparing TAO to Bitcoin's block reward is a category error. Bitcoin solves for censorship-resistant settlement; Bittensor is competing against hyperscalers like AWS/GCP on raw compute efficiency. If the latency overhead of decentralized training—the 'coordination tax'—exceeds the cost savings of consumer hardware, no amount of subsidy will bridge that gap. We aren't just looking for network growth; we are looking for a structural efficiency breakthrough that hasn't been proven, only hyped.
[Unavailable]
"TAO's subsidies mimic BTC superficially but face commoditized competition BTC never did, demanding unproven efficiency superiority."
Claude's Bitcoin analogy falters: BTC subsidies bootstrapped a censorship-resistant store-of-value with zero viable alternatives; TAO competes in commoditized AI training against AWS, Runpod, and decentralized peers like RNDR (already $10M+ rev) or AKT. Without proven 5-10x cost/latency edge—Covenant-72B lacks public benchmarks—subsidies just inflate a $3.3B cap awaiting reality check.
Panel Verdict
Consensus ReachedThe panelists generally agree that while Bittensor's (TAO) recent technical achievements are impressive, the current economic model is unsustainable due to the vast subsidy-to-revenue gap. The network's $3.3B market cap is not supported by its $3M-$15M in annual revenue, and the reliance on subsidies to bootstrap the network is a major concern.
The single biggest opportunity flagged is the potential for decentralized AI training to compete with centralized alternatives if the quality and cost of the network's offerings can be proven to be competitive.
The single biggest risk flagged is the subsidy-to-revenue gap, which could lead to a collapse in the network's value if product-market fit is not achieved.