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 (TAO 0.55%) 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.
But does this coin actually have what it takes to be a good investment, or is it just a flash in the pan?
Image source: Getty Images.
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.
Today’s Change
(-0.55%) $-1.71
Current Price
$305.61
Key Data Points
Market Cap
$2.9B
Day’s Range
$301.60 – $314.90
52wk Range
$144.32 – $535.12
Volume
305M
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.
