Person. Old Samster · @Old_Samster
The Bittensor analyst: a former Messari researcher who wrote the December 2023 Discovering Taoism report, then in April 2025 launched Unsupervised Capital, a fund dedicated to Bittensor, and now publishes the most consistent investor-side framing of dTAO and subnets.
Bittensor is the definitive implementation of cryptoeconomic protocols for AI development.
15 Apr 2025 · in contextEvery quote is verbatim and dated. The reading above each era is written from the cited quotes only. Numbers link to the quote they cite.
As a Messari analyst in December 2023 he presents Bittensor as a complex, misunderstood rabbit hole and defines it functionally as a network that continuously rewards better ML models across targeted use cases [sami-kassab-001, sami-kassab-002].
rabbit holeBittensor has been a rabbit hole that keeps drawing me deeper. Over the past few months, I've spent quite a bit of time researching and engaging in conversations with subnet miners, validators, and owners. Given its complexity, Bittensor is often misunderstood.
His X post introducing his Messari report on Bittensor, written while he was still an analyst there; the post text was reached through the X page title and the fxtwitter mirror.
Bittensor continuously rewards new, better machine learning (ML) models across a variety of targeted use cases.
Opening line of his Messari report Discovering Taoism, read from the report card attached to his X post; the report page itself returned rate-limit errors.
With the fund launch in April 2025 the register turns to conviction and lineage: the definitive implementation of cryptoeconomic protocols for AI, the most significant application of crypto since Bitcoin [sami-kassab-003, sami-kassab-005, sami-kassab-010]. Through 2025 he builds an investor vocabulary for dTAO, subnets as startups, the protocol as venture capitalist, subnets as liquidity pools, and Bitcoin's halving lore as reassurance [sami-kassab-011, sami-kassab-013, sami-kassab-015, sami-kassab-016], while admitting the pitch is too complex 017. By 2026 the frame is generalised: a subnet is an economic verification system and the only substrate where agents cooperate [sami-kassab-019, sami-kassab-020].
venture capitalistShift From analyst describing a mechanism to investor supplying category, lineage and valuation frames.
Bittensor is the definitive implementation of cryptoeconomic protocols for AI development.
From the post announcing Unsupervised Capital on the fund's site, in the section explaining what Bittensor is.
Not since the formative days of Bitcoin and Ethereum have we witnessed a grassroots movement with this combination of vision and execution.
From the fund announcement, explaining why he is dedicating a fund to a single thesis.
We believe this will be the most significant application of crypto since Bitcoin created a non-sovereign store of value.
From the fund announcement; the earliest instance of his repeated line placing Bittensor next to Bitcoin.
These subnets are already generating outputs that surpass those from centralized organizations wielding exponentially greater resources.
From the fund announcement, after noting that over 90 subnets had launched in roughly 18 months.
The Dynamic TAO upgrade, implemented in February 2025, was a watershed moment for Bittensor.
From the fund announcement, on why the fund launches now; he adds that individual subnet tokens let investors allocate capital with precision.
Ownership: Token incentives transform contributors into owners, creating skin-in-the-game motivation. Competitive Arenas: Development becomes a global competition where talented individuals and teams constantly push each other to excel.
Two of the four mechanisms he lists in the fund announcement for why cryptoeconomic protocols beat centralised organisations at AI development.
Our mission is singular: accumulate more TAO for our LPs than they could through passive staking or simple indexing.
The fund's stated mission in the announcement, which denominates returns in TAO rather than dollars.
Today I'm proud to announce Unsupervised Capital — a liquid token fund dedicated to Bittensor. At a moment when conviction feels scarce across crypto, I've never felt more clarity or confidence about a single thesis: Bittensor.
His X post announcing the fund the day after the site post, naming Seth Bloomberg as partner and Crucible Labs as anchor investor.
In early 2025, Bittensor's Dynamic TAO upgrade gave every subnet its own token, handing control of emission distribution over to the market. Subnet teams now operate more like startups, competing in public for capital and attention.
Opening of the first Bittensor Signals note, written for investors trying to value subnet tokens.
Most investors are applying valuation frameworks shaped by the last cycle's low-float, high-FDV meta—missing the fact that subnet tokens are a structurally different asset class. … That's exactly why no one uses FDV to value Bitcoin.
From the same note, arguing that fairly launched subnet tokens with Bitcoin-like emission schedules should be valued like Bitcoin, not like VC tokens.
the protocol acts as the ecosystem's venture capitalist
From his essay casting Bittensor as an internet capital market after Hyperliquid, describing how newly minted TAO is distributed to subnets in proportion to their token value.
Running all 500 SWE-bench problems with the top Ridges agent costs just $1.26, compared to $94 for Claude Opus.
From his memo on why the fund invested in Ridges, presented as an open tournament that beat billion-dollar AI labs.
If you dig back into the lore of Bitcoin's first halving, the mood feels eerily familiar: pessimists convinced that Bitcoin would death spiral. … There's no reason to doubt programmatic monetary policy after seeing it succeed with Bitcoin.
From his note ahead of Bittensor's first halving in December 2025, answering the mixed mood around it.
A subnet at its core is just a liquidity pool. The TAO halving will cut the chain's liquidity injections in half. … Bittensor's halving is one of those shocks. It's an inadvertent one that stress-tests and hardens the network.
From the same halving note, explaining to subnet token holders why thinner liquidity means higher beta.
My takeaway from yesterday is that the Bittensor pitch is still pretty complex and would benefit from being simplified. … holding TAO gives one exposure to all Subnets.
His X post after a day of pitching Bittensor at an event, quoting @counterpartytv's suggestion that the simple thesis is TAO as decentralized OpenAI.
Mining increasingly resembles a prompt engineering game rather than a pure software engineering problem. … approaching the inflection point of the S-curve, where attention, builders, and capital deployment begin to accelerate
From the excerpts of the fund's Q4 2025 letter, after a quarter he calls difficult for crypto and in which he calls the Taoflow upgrade contentious for landing without a formal feedback period.
A subnet is an economic verification system. Any objective function can be translated into an incentive mechanism that defines the rules of competition. … If you can verify the solution, the problem is fair game.
From his essay arguing that with generation costs collapsing, the hard part is specification and market design, using the Synth subnet as the worked example.
Bittensor subnets are that substrate. They're literally just an objective function that rewards those that optimize for it. … Bittensor has built a mechanism to extract cooperation out of competition, which means it's become the only platform agents are actually cooperating on today.
From his essay claiming Bittensor is the only venue where AI agents actually work together, because agent marketplaces without a quality measure have nothing to coordinate.
From June 2026 he defends founder-led change over slow consensus and lists what still needs redesign, while declaring he stays convicted [sami-kassab-021, sami-kassab-022]. He reads the emission buyback change as Bittensor becoming a VC on its own network 023, restates the Bitcoin line as decentralised AI being the largest crypto movement since Bitcoin and adds the mirror for AI 024, and tells subnet founders to raise equity 025.
mirrorShift Governance enters the framing for the first time, and the audience widens from investors to builders.
Bittensor is not the same network it was three months ago, let alone a year ago. … These kinds of changes are extremely difficult to push through a democratic process, where consensus is slow and likely to be out of reach.
From his essay on the run of protocol changes in 2026, endorsing founder-led direction for what he calls the Velocity era while listing what still needs fixing.
Emission allocation needs to be redesigned, validators need to be made productive, complexity needs to be stripped out for new builders. … Neither TAO nor subnet tokens were pre-mined. Both were earned through useful work. … through all the drama, late night upgrades, and changes, I stay convicted.
From the same essay, where he identifies as an operator and investor deep in the ecosystem and criticises past communication gaps.
The chain now splits emissions two ways. One part still deepens the pool while the other part goes straight into buying the subnet's own token. … This turns Bittensor into a VC investing in the emerging leaders on its own network.
From his note the day after a pull request changed how emissions are used, arguing top subnets now have a protocol buyer sized to match miner selling.
Bittensor is becoming the mirror for AI the same way Bitcoin is for money. … Const is effectively operating as CEO again, and upgrades are shipping faster than they have in a year. … Decentralized AI will be the largest crypto movement since Bitcoin.
From the excerpts of the fund's Q2 2026 letter; the last sentence is the latest instance of the since-Bitcoin line he first used in April 2025.
One of crypto's oldest arguments is whether tokens and equity can coexist. The prevailing view says no, which we also believe to be true for most tokens. … So to Bittensor's subnet founders, raise your equity.
From his post titled We're One of the Largest TAO Holders and Want Builders on Bittensor to Raise Equity, arguing value flows to the token by necessity in proof-of-useful-work networks.
A brand strategy skeleton turned on the speaker: not what Bittensor's brand is, but what this voice's Bittensor is. Slots with no evidence are left out rather than filled.
What they say it is for, the world it builds, what it stands for.
Bittensor continuously rewards better machine learning models across targeted use cases; by 2026 he generalises this to any problem whose solution can be verified.
A subnet is an economic verification system. Any objective function can be translated into an incentive mechanism that defines the rules of competition. … If you can verify the solution, the problem is fair game.
Decentralised AI as the largest crypto movement since Bitcoin, with Bittensor as the mirror for AI the way Bitcoin is for money.
Bittensor is becoming the mirror for AI the same way Bitcoin is for money. … Const is effectively operating as CEO again, and upgrades are shipping faster than they have in a year. … Decentralized AI will be the largest crypto movement since Bitcoin.
We believe this will be the most significant application of crypto since Bitcoin created a non-sovereign store of value.
Fair launch and useful work: nothing pre-mined, everything earned; and, from mid 2026, founder-led decisiveness over slow democratic consensus.
Emission allocation needs to be redesigned, validators need to be made productive, complexity needs to be stripped out for new builders. … Neither TAO nor subnet tokens were pre-mined. Both were earned through useful work. … through all the drama, late night upgrades, and changes, I stay convicted.
Bittensor is not the same network it was three months ago, let alone a year ago. … These kinds of changes are extremely difficult to push through a democratic process, where consensus is slow and likely to be out of reach.
What is broken that Bittensor answers, in their telling.
The pitch is too complex for newcomers, and inside the network emission allocation, validator productivity and builder complexity still need fixing.
My takeaway from yesterday is that the Bittensor pitch is still pretty complex and would benefit from being simplified. … holding TAO gives one exposure to all Subnets.
Emission allocation needs to be redesigned, validators need to be made productive, complexity needs to be stripped out for new builders. … Neither TAO nor subnet tokens were pre-mined. Both were earned through useful work. … through all the drama, late night upgrades, and changes, I stay convicted.
Without fast, unpopular changes pushed by the founder, the protocol's needed redesign would stall in consensus.
Bittensor is not the same network it was three months ago, let alone a year ago. … These kinds of changes are extremely difficult to push through a democratic process, where consensus is slow and likely to be out of reach.
The tailwinds they cite.
dTAO as a watershed that lets capital allocate with precision, the first halving as programmatic monetary policy proven by Bitcoin, emission buybacks, and an S-curve inflection in attention and capital.
The Dynamic TAO upgrade, implemented in February 2025, was a watershed moment for Bittensor.
If you dig back into the lore of Bitcoin's first halving, the mood feels eerily familiar: pessimists convinced that Bitcoin would death spiral. … There's no reason to doubt programmatic monetary policy after seeing it succeed with Bitcoin.
Mining increasingly resembles a prompt engineering game rather than a pure software engineering problem. … approaching the inflection point of the S-curve, where attention, builders, and capital deployment begin to accelerate
The chain now splits emissions two ways. One part still deepens the pool while the other part goes straight into buying the subnet's own token. … This turns Bittensor into a VC investing in the emerging leaders on its own network.
Who they talk to and what they tell them.
He speaks to LPs and investors who misvalue subnet tokens, to miners for whom mining is becoming a prompt engineering game, and to subnet founders whom he tells to raise equity.
Our mission is singular: accumulate more TAO for our LPs than they could through passive staking or simple indexing.
In early 2025, Bittensor's Dynamic TAO upgrade gave every subnet its own token, handing control of emission distribution over to the market. Subnet teams now operate more like startups, competing in public for capital and attention.
Most investors are applying valuation frameworks shaped by the last cycle's low-float, high-FDV meta—missing the fact that subnet tokens are a structurally different asset class. … That's exactly why no one uses FDV to value Bitcoin.
A subnet at its core is just a liquidity pool. The TAO halving will cut the chain's liquidity injections in half. … Bittensor's halving is one of those shocks. It's an inadvertent one that stress-tests and hardens the network.
Mining increasingly resembles a prompt engineering game rather than a pure software engineering problem. … approaching the inflection point of the S-curve, where attention, builders, and capital deployment begin to accelerate
One of crypto's oldest arguments is whether tokens and equity can coexist. The prevailing view says no, which we also believe to be true for most tokens. … So to Bittensor's subnet founders, raise your equity.
Bittensor has been a rabbit hole that keeps drawing me deeper. Over the past few months, I've spent quite a bit of time researching and engaging in conversations with subnet miners, validators, and owners. Given its complexity, Bittensor is often misunderstood.
My takeaway from yesterday is that the Bittensor pitch is still pretty complex and would benefit from being simplified. … holding TAO gives one exposure to all Subnets.
What they set it against.
Against centralised labs with exponentially more resources, against last-cycle VC token valuation habits, and against the crypto orthodoxy that tokens and equity cannot coexist.
These subnets are already generating outputs that surpass those from centralized organizations wielding exponentially greater resources.
Most investors are applying valuation frameworks shaped by the last cycle's low-float, high-FDV meta—missing the fact that subnet tokens are a structurally different asset class. … That's exactly why no one uses FDV to value Bitcoin.
Running all 500 SWE-bench problems with the top Ridges agent costs just $1.26, compared to $94 for Claude Opus.
One of crypto's oldest arguments is whether tokens and equity can coexist. The prevailing view says no, which we also believe to be true for most tokens. … So to Bittensor's subnet founders, raise your equity.
Bittensor is not the same network it was three months ago, let alone a year ago. … These kinds of changes are extremely difficult to push through a democratic process, where consensus is slow and likely to be out of reach.
Category frame, differentiator, reason to believe.
Bittensor is the definitive implementation of cryptoeconomic protocols for AI, an internet capital market whose protocol acts as venture capitalist, and each subnet is an economic verification system or, at its core, a liquidity pool.
Bittensor is the definitive implementation of cryptoeconomic protocols for AI development.
In early 2025, Bittensor's Dynamic TAO upgrade gave every subnet its own token, handing control of emission distribution over to the market. Subnet teams now operate more like startups, competing in public for capital and attention.
the protocol acts as the ecosystem's venture capitalist
A subnet is an economic verification system. Any objective function can be translated into an incentive mechanism that defines the rules of competition. … If you can verify the solution, the problem is fair game.
The chain now splits emissions two ways. One part still deepens the pool while the other part goes straight into buying the subnet's own token. … This turns Bittensor into a VC investing in the emerging leaders on its own network.
We believe this will be the most significant application of crypto since Bitcoin created a non-sovereign store of value.
Ownership and competitive arenas; fairly launched tokens on Bitcoin-like schedules that are a structurally different asset class; the only platform that extracts cooperation from competition among agents.
Ownership: Token incentives transform contributors into owners, creating skin-in-the-game motivation. Competitive Arenas: Development becomes a global competition where talented individuals and teams constantly push each other to excel.
Most investors are applying valuation frameworks shaped by the last cycle's low-float, high-FDV meta—missing the fact that subnet tokens are a structurally different asset class. … That's exactly why no one uses FDV to value Bitcoin.
Bittensor subnets are that substrate. They're literally just an objective function that rewards those that optimize for it. … Bittensor has built a mechanism to extract cooperation out of competition, which means it's become the only platform agents are actually cooperating on today.
Grassroots vision and execution unseen since early Bitcoin and Ethereum, and Bitcoin's own halving history as the reason to trust programmatic emissions.
Not since the formative days of Bitcoin and Ethereum have we witnessed a grassroots movement with this combination of vision and execution.
If you dig back into the lore of Bitcoin's first halving, the mood feels eerily familiar: pessimists convinced that Bitcoin would death spiral. … There's no reason to doubt programmatic monetary policy after seeing it succeed with Bitcoin.
The benefit they claim, the feeling they sell, what joining says about you.
Subnet tokens with staking yield and tightening liquidity, and a protocol that now buys the tokens of its own leaders.
A subnet at its core is just a liquidity pool. The TAO halving will cut the chain's liquidity injections in half. … Bittensor's halving is one of those shocks. It's an inadvertent one that stress-tests and hardens the network.
The chain now splits emissions two ways. One part still deepens the pool while the other part goes straight into buying the subnet's own token. … This turns Bittensor into a VC investing in the emerging leaders on its own network.
Clarity and unshakable conviction at a moment when conviction is scarce, held through drama and late-night upgrades.
Today I'm proud to announce Unsupervised Capital — a liquid token fund dedicated to Bittensor. At a moment when conviction feels scarce across crypto, I've never felt more clarity or confidence about a single thesis: Bittensor.
Not since the formative days of Bitcoin and Ethereum have we witnessed a grassroots movement with this combination of vision and execution.
Emission allocation needs to be redesigned, validators need to be made productive, complexity needs to be stripped out for new builders. … Neither TAO nor subnet tokens were pre-mined. Both were earned through useful work. … through all the drama, late night upgrades, and changes, I stay convicted.
Contributors become owners with skin in the game, competing in a global arena.
Ownership: Token incentives transform contributors into owners, creating skin-in-the-game motivation. Competitive Arenas: Development becomes a global competition where talented individuals and teams constantly push each other to excel.
The line they repeat and the proof they point to.
The most significant application of crypto since Bitcoin, restated in 2026 as decentralised AI being the largest crypto movement since Bitcoin and Bittensor as the mirror for AI.
We believe this will be the most significant application of crypto since Bitcoin created a non-sovereign store of value.
Bittensor is the definitive implementation of cryptoeconomic protocols for AI development.
Bittensor subnets are that substrate. They're literally just an objective function that rewards those that optimize for it. … Bittensor has built a mechanism to extract cooperation out of competition, which means it's become the only platform agents are actually cooperating on today.
Bittensor is becoming the mirror for AI the same way Bitcoin is for money. … Const is effectively operating as CEO again, and upgrades are shipping faster than they have in a year. … Decentralized AI will be the largest crypto movement since Bitcoin.
Subnet outputs surpassing centralised organisations, with Ridges beating a frontier lab on SWE-bench at a fraction of the cost as the flagship case.
These subnets are already generating outputs that surpass those from centralized organizations wielding exponentially greater resources.
Running all 500 SWE-bench problems with the top Ridges agent costs just $1.26, compared to $94 for Claude Opus.
Register more than claim.
Investor-letter prose: calm, numerical, self-disclosing about complexity and conviction.
Bittensor has been a rabbit hole that keeps drawing me deeper. Over the past few months, I've spent quite a bit of time researching and engaging in conversations with subnet miners, validators, and owners. Given its complexity, Bittensor is often misunderstood.
My takeaway from yesterday is that the Bittensor pitch is still pretty complex and would benefit from being simplified. … holding TAO gives one exposure to all Subnets.
How they describe their own relation to it.
Analyst turned dedicated investor: a Messari researcher down the rabbit hole in 2023, a fund manager denominated in TAO from 2025, and by 2026 an operator and one of the largest TAO holders who stays convicted.
Bittensor has been a rabbit hole that keeps drawing me deeper. Over the past few months, I've spent quite a bit of time researching and engaging in conversations with subnet miners, validators, and owners. Given its complexity, Bittensor is often misunderstood.
Our mission is singular: accumulate more TAO for our LPs than they could through passive staking or simple indexing.
Today I'm proud to announce Unsupervised Capital — a liquid token fund dedicated to Bittensor. At a moment when conviction feels scarce across crypto, I've never felt more clarity or confidence about a single thesis: Bittensor.
Emission allocation needs to be redesigned, validators need to be made productive, complexity needs to be stripped out for new builders. … Neither TAO nor subnet tokens were pre-mined. Both were earned through useful work. … through all the drama, late night upgrades, and changes, I stay convicted.
Bittensor is becoming the mirror for AI the same way Bitcoin is for money. … Const is effectively operating as CEO again, and upgrades are shipping faster than they have in a year. … Decentralized AI will be the largest crypto movement since Bitcoin.
One of crypto's oldest arguments is whether tokens and equity can coexist. The prevailing view says no, which we also believe to be true for most tokens. … So to Bittensor's subnet founders, raise your equity.
What they call things, and refuse to call them.
He names the protocol the ecosystem's venture capitalist, a subnet an economic verification system and a liquidity pool, and frames the whole as an internet capital market.
the protocol acts as the ecosystem's venture capitalist
A subnet at its core is just a liquidity pool. The TAO halving will cut the chain's liquidity injections in half. … Bittensor's halving is one of those shocks. It's an inadvertent one that stress-tests and hardens the network.
A subnet is an economic verification system. Any objective function can be translated into an incentive mechanism that defines the rules of competition. … If you can verify the solution, the problem is fair game.
Who they agree with, borrow from, or argue against, with the quote that shows it.
Public appearances on the record, whether or not a quote was taken.
Something wrong or missing? Quotes are corrected as they arrive; a misattribution is removed first and argued after. How to file one · GitHub issue