Why Normal People Rarely Get the Big Upside
There is something strange about the way ordinary people are expected to invest in technological progress.
We are constantly told that artificial intelligence may become one of the most important economic developments of our lifetime. We are told that new AI companies may transform healthcare, software, finance, media and almost every other industry. But when a genuinely promising company is still young—precisely when the uncertainty is greatest and the potential upside is largest—most normal people cannot invest in it.
They can follow the founders on X. They can listen to the podcasts. They can use the product and perhaps even provide the data that helps improve it. What they usually cannot do is buy a meaningful piece of the company.
The earliest investment rounds tend to be reserved for venture funds, wealthy individuals, employees and people already connected to the right networks. By the time a successful startup finally lists on a public exchange, several private rounds may already have taken place at increasingly higher valuations.
Public investors are eventually invited in.
They are simply not invited first.
I have always found this one of the more frustrating parts of the normal investment world. Society celebrates entrepreneurship and broad participation, but the legal and financial structure often divides people into two groups: those allowed to invest while the future is still uncertain, and those allowed to buy after professional investors have already selected the winners.
Bittensor does something structurally different.
It allows almost anyone with TAO and a compatible wallet to allocate capital toward young subnet markets from a very early stage. There is no venture fund deciding whether your bank account is large enough. There is no invitation-only funding round. You do not need to know the founder or live in San Francisco.
That does not make subnet tokens equivalent to startup shares. They are not.
But it does open a door that is normally closed.
Why early-stage investing is so restricted
There are understandable reasons why private investment markets developed this way.
Young companies are risky. Information is incomplete. Many startups fail. Their shares are difficult to value and often impossible to sell for years. Regulators therefore try to protect inexperienced investors from fraud, unsuitable investments and losses they may not be able to absorb.
The practical result, however, is rather peculiar: people who already possess substantial wealth are given access to some of the highest-risk, highest-upside investments, while people with less wealth are protected by excluding them from the opportunity entirely.
In the United States, for example, many common private placements are mainly available to “accredited investors.” An individual can qualify through criteria including more than $1 million in net worth, excluding the primary residence, or annual income above $200,000 individually. Other routes exist through professional qualifications, but wealth and income remain central access tests. The precise rules differ between countries, yet the broader division between private capital and ordinary public-market access is familiar across much of the investment world.
The intention is so-called protection.
The consequence is that wealth often becomes the admission ticket to early-stage risk.
That distinction matters because risk and upside normally travel together. Venture investors accept that many companies will fail because a small number of exceptional successes can compensate for the losses. Ordinary investors are frequently prevented from constructing the same kind of portfolio directly. They may gain indirect access through listed investment companies, pension funds or specialised retail vehicles, but they rarely receive the same choice over individual early-stage companies.
Even when a company eventually launches an initial public offering, the public is entering a company that has already spent years developing privately. An IPO (think of SpaceX IPO…) is the first registered sale of shares to the broader public, not the beginning of the company. It creates public access and liquidity, which is valuable, but it arrives after founders, employees and private investors have already owned the business through its earlier stages.
This does not mean public investors can no longer earn excellent returns. A great company can continue growing for decades after listing. The unfairness is subtler than that.
Normal people can participate in the later journey.
They are usually denied the choice to participate in the earliest one.
AI makes this exclusion particularly visible
The structure becomes especially frustrating in artificial intelligence because so many people can see the transformation happening in real time.
Millions use products built by private AI companies. Developers build businesses on their models. Workers alter their careers because of them. Governments debate their influence. Yet direct ownership of many important AI startups remains unavailable to the same public living through the disruption.
A normal investor may buy shares in Microsoft, Alphabet, Amazon, NVIDIA or another listed company with substantial AI exposure. That is useful, but it is not the same as choosing an early AI startup because you believe in its particular product or technical approach.
The individual is left with a strange selection.
He can invest in large public corporations that may benefit from AI. He can invest in small listed companies that have attached AI to their investor presentation, occasionally with admirable enthusiasm. Or he can wait and hope that the private company he actually finds interesting eventually launches an IPO.
Meanwhile, venture investors can compare private model companies, agent platforms, robotics startups, compute providers and data businesses years before those opportunities become public.
Again, this is not a secret conspiracy.
It is simply how the structure works.
And once a structure has existed for long enough, people begin to treat it as though no alternative could exist.
Bittensor opens the markets much earlier
Bittensor contains many separate subnets, each trying to produce a specific form of digital value.
Chutes focuses on open-source AI inference. Targon works on confidential compute. Hippius provides decentralized storage. Other subnets develop search, cybersecurity, forecasting, identity systems, computer vision and numerous forms of specialised intelligence.
These projects are often extremely young. Their products may be unfinished, their business models uncertain and their token markets highly volatile. In other words, they resemble early-stage experiments.
The difference is that the public does not need to wait for an IPO.
Every normal subnet has an alpha token. At the protocol level, someone with TAO can exchange it for the alpha token of a chosen subnet and stake that alpha with a validator. The position can later be exchanged back into TAO through the subnet’s on-chain market. There is no accredited-investor test built into the Bittensor protocol and no subnet investment committee that decides whether the participant is sufficiently wealthy or well connected. Local law, exchange access and tax treatment still differ by country, but the network itself is designed for open participation.
This is one of the coolest parts of Bittensor to me.
Someone can discover a small subnet, study its product and decide that its approach deserves support. That person can allocate ten TAO, one TAO or a fraction of one TAO. A venture capitalist may allocate considerably more, of course, but both enter the same public market.
The small investor is not guaranteed the same information, influence or outcome.
But at least the door is open.
Alpha tokens are not startup shares
This is where the comparison with venture capital needs to become much more precise.
When you buy shares in a startup, you own equity in a legal company. Depending on the share class and agreements, that equity may give you economic ownership, voting rights, a claim on part of the company if it is sold and potential participation in future dividends or distributions.
An alpha token normally gives you none of those things.
Holding Chutes alpha does not mean you own shares in the company behind Chutes. You do not automatically own part of its servers, intellectual property, customer contracts or bank account. You do not receive a legal percentage of its revenue. If the company is acquired for an enormous amount of money, alpha holders do not necessarily receive anything from that sale.
Alpha is the staking and market token of a Bittensor subnet. It represents an economic position inside the subnet’s incentive system, not corporate ownership. When TAO is staked into a subnet, it enters that subnet’s market and is exchanged for alpha. The alpha is then held as stake with a validator and can earn additional alpha rewards.
That distinction is not a small legal disclaimer hidden at the bottom of the page.
It changes the investment thesis completely.
You may believe strongly in the team behind a subnet, but the team’s success and the alpha token’s success are not automatically the same thing. A subnet company could build a valuable product, earn substantial revenue and keep most of that value inside the private company. Alpha holders would not necessarily share directly in the business success.
The reverse is also possible. An alpha token can increase rapidly because investors become excited, even while the underlying product remains small or unfinished.
Anyone considering subnet tokens therefore needs to ask two separate questions:
Does the start-up need to have a bittensor subnet?
And:
Does success of the product create value for the alpha token?
The second question is often harder.
What does buying alpha actually support?
Buying alpha does not work like wiring money into a startup’s bank account.
When you exchange TAO for alpha, the TAO moves into the subnet’s on-chain liquidity pool. It does not simply arrive as cash that the subnet founder can spend on salaries and office chairs.
The effect is more indirect.
Buying alpha raises the subnet token’s market price. Under Bittensor’s current design, sustained market support can influence how much newly created TAO the network directs toward that subnet. Those network rewards are then distributed among the subnet’s participants, including miners, validators, stakers and the subnet owner. A stronger token market can therefore help the subnet attract providers and increase the economic resources supporting its operation.
This is why staking can be compared with allocating research and development capital, although the comparison should not be taken too literally.
A subnet trying to build decentralized storage may use network rewards to attract people who contribute storage. A compute subnet can use them to attract GPU providers. A model-evaluation subnet can reward people who produce or judge better models.
By choosing where to stake, TAO holders influence which subnet economies receive the strongest market support and protocol rewards.
They are not purchasing shares in a research company.
They are helping direct an open incentive budget toward a particular field of work.
I actually find that more interesting in some ways. Traditional venture capital funds a company, which then hires employees and owns the resulting product. Bittensor can reward a broader network of independent participants who compete to produce the desired resource.
Whether that works better depends entirely on the subnet. Some problems fit open competition. Others may still be organised more effectively inside a conventional company.
Bittensor gives people the opportunity to make that judgement themselves.
A liquid form of early-stage allocation
The second major difference is liquidity.
A venture investment can remain locked for many years. The investor may have little ability to sell, and there may be no meaningful market price between funding rounds. An ordinary investor accessing venture capital through a fund has even less control over individual decisions. The fund manager chooses the companies, the timing and the eventual exits.
Subnet tokens trade continuously through on-chain markets.
Someone can enter a young subnet, follow its progress and later reduce or exit the position. Capital can move from one subnet to another as new products launch and new information emerges. Normal subnet staking does not require a multi-year holding period.
This flexibility is remarkable, but it should not be confused with guaranteed liquidity.
When someone leaves a subnet, the alpha is exchanged back into TAO through the subnet’s automated market maker. A large sale can push the price downward during the transaction, especially when the subnet is small. This is known as slippage. If many holders try to leave at once, the displayed price may prove considerably more optimistic than the price they actually receive.
So yes, you can change your mind.
The market may charge you quite enthusiastically for doing so.
Still, the difference from conventional venture investing remains substantial. A subnet position can usually be adjusted without waiting for the team to arrange another funding round, find a buyer for private shares or spend ten years reaching an exit.
This creates something that resembles liquid venture allocation: very early exposure combined with an ability to reallocate capital in real time.
It is not quite venture capital.
It is not quite public-equity investing.
It is something new and, naturally, rather messy.
Open access also opens the door to bad investments
The closed investment system has obvious unfairness, but opening access does not remove the reasons why early-stage investing is difficult.
It transfers more of the responsibility to the individual.
A private investor may receive detailed financial statements, legal agreements, access to management and contractual rights. Alpha holders may have little more than public documentation, GitHub activity, product announcements, on-chain data and whatever the team says on social media.
Subnet teams are not necessarily required to provide the level of reporting expected from listed companies. Token holders may have no legal protection when a roadmap changes, a team disappears or the private company makes a decision that benefits its shareholders rather than alpha holders. It’s a bit more wild west style in Bittensor then in traditional equity investment…
The market is also vulnerable to speculation. A subnet may attract capital because its product is genuinely improving. It may also attract capital because its yield is high, its community is loud or its token rose last week.
Open participation allows normal people to enter early.
It also allows them to make early-stage mistakes without professional supervision.
I do not think this weakens the argument for open access. Adults should generally be permitted to take risks with their own money, provided the system is honest about what they are buying.
But Bittensor should not copy the worst part of crypto culture, where removing the gatekeeper is treated as though it also removed the need for research.
The gatekeeper has disappeared.
The risk has not.
The dilution story is more complicated
One of the most interesting aspects of Bittensor is that the max supply of both the TAO and subnet tokens are known: 21 million. This is in contrast to equity shares which can be printed like dollars. Many people are not aware of this but companies can file new shares if they need to raise capital. They literally dilute the existing shares, and hence your ownership percentage!
Subnet tokens are continually issued as rewards to miners, validators, owners and stakers. The total supply may have a defined maximum, but the circulating supply can still grow substantially over time. Someone holding alpha can therefore be diluted by ongoing emissions even when the protocol never changes the final cap. At this moment (July 2026, the oldest subnets have a little bit more then 5 million subnet tokens emitted so there are still more then 15 million tokens to come!).
This dilution is more transparent than a private company unexpectedly issuing a new funding round on terms an ordinary shareholder cannot influence. The rules are visible on-chain, and holders can observe how quickly supply is expanding.
But transparent dilution is still dilution.
A subnet must generate enough demand to absorb the new alpha entering circulation. If miners and stakers continuously receive tokens and sell them, while very few users or investors want to buy, the price can decline even when the subnet’s product remains operational.
This is another reason why real revenue matters.
A subnet that earns money from customers and uses some of it to purchase alpha creates a source of demand outside the staking economy. The product begins supporting the token, rather than the token merely financing the product.
Without that connection, alpha holders may be funding experimentation through inflation while the private company retains most of the eventual commercial value.
That is not automatically wrong.
It simply needs to be understood.
The opportunity is different from the traditional one
The cleanest way to describe subnet investing is not that Bittensor has placed private startup equity on a blockchain.
It has not.
Bittensor has created open markets around early-stage digital economies. These markets may be closely associated with startup teams, but the token represents participation in the subnet’s incentive system rather than ownership of the company.
This means the opportunity is both broader and weaker than equity.
It is broader because almost anyone can participate, positions can be small, markets are continuously available and capital can be reallocated quickly.
It is weaker because token holders may lack ownership rights, revenue claims, legal protections and direct access to company information.
The upside may be substantial when a subnet’s product succeeds and that success creates sustained demand for alpha.
The disappointment can be equally substantial when the company succeeds but the token captures little of the value—or when neither succeeds.
This is why I would never tell someone to buy a subnet token merely because the subnet team looks talented. The important question is whether the token is genuinely connected to the value being created.
That connection differs from subnet to subnet.
Why I still find this revolutionary
After all these qualifications, I remain extremely enthusiastic about the underlying idea.
For most of modern finance, access to early-stage innovation has depended heavily on wealth, geography and personal networks. A small group of funds decides which founders receive capital. The investors in those funds are themselves usually wealthy individuals or large institutions. Ordinary people encounter the successful companies later, once the early ownership structure has already been established.
Bittensor allows a person sitting at home to study a young AI network and take a position in its development.
That person might discover a compute subnet before it has many customers. He might find a privacy project that nobody outside Bittensor understands yet. He might allocate a small amount of TAO, follow the project for years and increase or reduce the position as the evidence changes.
No one asks for his résumé.
No one checks whether he owns a million-dollar portfolio.
No venture partner needs to invite him into the round.
He can simply participate.
Perhaps he will be right. Perhaps he will make the same terrible investment decisions that wealthy venture capitalists regularly make, only with a smaller office and fewer catered lunches. Open access does not create wisdom. It does create possibility.
A fairer system does not guarantee equal outcomes
Bittensor will not make investing equal.
Large holders can allocate more capital. Insiders may still possess better information. Founders and early miners can accumulate tokens before outsiders understand the project. Technically sophisticated participants may identify opportunities sooner. Wealth will continue to create advantages, as it tends to do with impressive persistence.
But there is a fundamental difference between unequal outcomes inside an open market and a market that forbids most people from entering at all.
Bittensor lowers the gate.
It does not level the entire landscape.
I think that is still a meaningful improvement.
The ordinary participant gains the right to evaluate an early project, accept the risk and allocate his own capital. He no longer has to wait until a regulator, investment bank and group of private shareholders decide the company is ready for him.
With that freedom comes the responsibility to understand what he is buying.
An alpha token is not a share.
A high staking return is not business revenue.
Liquidity is not guaranteed.
A good product does not automatically produce a valuable token.
But when a subnet builds something useful, connects real revenue to its alpha economy and continues attracting capable miners and validators, ordinary participants can gain exposure while the network is still young.
That is extremely rare in the normal investment world.
What Bittensor is really experimenting with
Bittensor is usually described as an experiment in decentralized artificial intelligence.
I think it is also an experiment in access to technological upside.
Can early digital infrastructure be funded through public markets from the beginning? Can small investors allocate alongside large investors instead of arriving years later? Can capital move continuously between competing research networks rather than being locked inside private funds? Can open markets identify valuable AI projects without a venture committee deciding what deserves to exist?
I do not know whether the answer to all of these questions will be yes.
Open token markets may fund brilliant infrastructure. They may also fund spectacular nonsense (as we often see in the general crypto market…) Most subnets will probably not become enduring businesses, just as most startups do not. I know many subnets with ridiculously high token prices (in my opinion) and I have no clue why. Bittensor is still young I like to think…
But at least the experiment is happening in public.
That is the part I care about.
Bittensor does not promise normal people that they will capture the next great technological fortune. It gives them something the traditional system often withholds much earlier than any guarantee of success:
the opportunity to participate.
For once, the public does not necessarily have to wait outside until the insiders have finished arranging the cap table.
The door is already open.