Neel_Proshun_DXC

OpenLedger is addressing the challenge of today's AI. However, What If the Problem Changes Before the Solution?
I have an issue I can't sleep without though, regarding the entire data attribution space.
Not if OpenLedger's technology is viable. Don't ask me if Proof of Attribution can scale. Not if businesses will embrace it.
Something more fundamental.
What if the problem (to be solved) becomes the problem?
And here's the awkward situation.
The thesis of OpenLedger is based on a certain premise: AI requires data made by humans. There is a lack of good, diverse and quality human data. The contributors should be paid for their contribution to provide something of high value and in short supply.
This is still the case today. Mostly.
However, data generation with synthetics is moving forward quicker than attribution conversations realize.
The synthetic data which AI generated models produced was shunned for years as poor-quality data compared to real human data. If models train on their own output, then the argument goes, the quality of their output goes down, which leads to further degradation. GIGO–garbage in, garbage out, amplified.
The argument was basically accurate in 2022. It gets more and more out-of-fashion each passing year.
However, recent studies indicate that when properly designed and created under certain constraints and quality filters, synthetic data can be as good or even better than human-generated data in some areas. Medical imaging. Code generation. Mathematical reasoning. Structured financial data.
Not all domains. Not yet. However, the path is there to see.
This leaves a strategic question unanswered by OpenLedger, at least publicly.
The market for human data contribution begins to shrink when synthetic data quality keeps getting better and there doesn't seem to be any good reason why it shouldn't.
Not disappearing. In contrast to commodity data, the most valuable human data is rare expertise, lived experience, genuine novel perspective, may last longer than the commodity data. The hands-on experience of a doctor over the course of decades. Experimentally observed properties of materials by a materials scientist. A writer's unique style.
However, the majority of the training data is the vast quantity of generic text, images, and structured information that the majority of current models are trained on – and which can be synthetically generated much earlier.
If so, OpenLedger's time to make the case for the imperative of creating an attribution framework could be shorter than the current story indicates.
Infrastructure needs to be integrated into, not skipped over. This was because TCP/IP was the only protocol that would allow networks to be connected. When there were no alternatives, SWIFT became unskippable as it was already handling financial flows all around the world.
Before synthetic data makes human data unnecessary, OpenLedger has to integrate in the workflow of creating AI.
It is not inevitable that there is an opposing point of view to consider.
With the use of synthetic data there is, of course, less human data but not no human data, meaning that a human data attribution is still necessary.
Synthetic data is created from real-world data. Human contributions were used to train the models generating synthetic data. The attribution issue simply becomes one step higher, who contributed to the development of the models used to create the synthetic data?
If this is the case, Proof of Attribution is more difficult, and more complex: The more they contribute, the longer the chain. The need for verifiable lineage actually grows.
That's the positive interpretation. As AI systems become more recursive and self-referential, OpenLedger's infrastructure becomes of increased importance.
I'm not sure which will be the case.
The only thing I do know is that, the synthetic data question is the most important stress test to test the attribution thesis and it is virtually untouched in the ongoing discussion about $OPEN .
The projects which make the cut with technology change are not necessarily the most successful ones that are finding solutions to today's issues.
They are the ones who have foreseen problems in tomorrow's world while most of the rest of the world was worrying about the problems of today.
OpenLedger is creating the proper infrastructure for this time.
Whether this one will be followed by something else—that will be the most interesting thing to watch for me.
Would you believe that, in time, synthetic data will make it less important to have human data? If yes, then does that mean the importance of attribution infrastructure is diminished or not?
@OpenLedger $OPEN #OpenLedger

The Internet Was Built to Extract. OpenLedger Is Trying to Rebuild It to Distribute. History Says That's Almost Impossible.
I want to talk about a pattern that keeps repeating in technology.
Every major platform starts with a promise of empowerment.
Bloggers will have a voice. YouTubers will build audiences. Uber drivers will be their own bosses. Airbnb hosts will monetize their assets. App developers will reach billions of users directly.
The promise is always the same: we're giving power to individuals.
And for a brief, genuinely exciting window the promise is real.
Early bloggers did find audiences. Early YouTubers did build sustainable income. Early Uber drivers did earn more than taxi drivers. Early app developers did make life-changing money.
Then the platform matures. The algorithm changes. The revenue share shifts. The terms of service get updated quietly. And the value that was briefly flowing outward starts flowing upward again.
Every time.
Without exception.
This isn't cynicism. It's a structural observation.
Platforms extract value because extraction is more economically efficient than distribution in the short term. It's cheaper to take than to share. It's easier to optimize for platform growth than contributor welfare. And once a platform reaches scale once leaving costs more than staying contributors lose their negotiating power entirely.
The internet didn't create this dynamic. It just perfected it.
Google didn't set out to exploit content creators. It built infrastructure that made content valuable, then gradually captured that value for itself as its market position strengthened. The same story played out with Facebook, YouTube, Spotify, Amazon's marketplace, Apple's App Store.
The architecture of extraction isn't a bug. It's what these systems inevitably become when there's no structural constraint preventing it.
This is why OpenLedger's thesis is genuinely ambitious and genuinely difficult.
It's not trying to build a better platform. It's trying to change the underlying architecture.
Proof of Attribution isn't just a payment mechanism. It's an attempt to make extraction structurally impossible. If every data contribution is cryptographically recorded on-chain, if every model usage automatically triggers contributor compensation, if the payment flow is hardcoded into the protocol rather than controlled by a company's policy team then the platform can't quietly change the terms.
The value distribution isn't a feature that can be turned off. It's the infrastructure itself.
That's architecturally different from every platform that came before.
But here's where history makes me cautious.
Changing value flow architecture requires overcoming the resistance of everyone currently benefiting from the existing architecture.
OpenAI, Google DeepMind, Anthropic, Meta AI — these companies have built trillion-dollar valuations on the current model. Their investors, their employees, their entire economic structure depends on data being cheap or free.
They will not adopt attribution infrastructure voluntarily. Not because they're malicious. Because the economics don't work in their favor.
Which means OpenLedger's real challenge isn't technical. The Proof of Attribution system is genuinely innovative.
The real challenge is adoption against incumbent resistance.
For attribution infrastructure to matter, AI developers need to build on OpenLedger instead of or in addition to existing centralized systems. That requires either regulatory pressure forcing attribution compliance, or enough data contributors withholding their data from non-attributing systems to make the quality difference noticeable.
Neither of those conditions fully exists yet.
Regulatory pressure is building the EU AI Act, pending US legislation, multiple ongoing lawsuits. But "building" is different from "arrived."
Data contributor coordination is historically very hard. The internet is full of examples of creators knowing they're being exploited and continuing to create anyway, because the audience is where the platform is.
I'm not saying OpenLedger can't succeed.
I'm saying it's attempting something that has failed many times before in different forms — and understanding why it failed before is the only honest way to evaluate whether this attempt is different.
The difference this time might be the blockchain layer. Making attribution immutable and automatic removes the "we changed the terms" failure mode that killed every previous attempt at fair creator compensation.
The difference might be regulatory timing. AI's data problem is hitting legal and political systems simultaneously in a way that previous platform extraction never quite did.
Or the difference might not be enough.
History is full of elegant infrastructure that arrived before the conditions for adoption existed. Some waited long enough to matter. Most didn't.
OpenLedger is betting it can create those conditions, or arrive just as they're forming naturally.
That's either precise timing or optimistic timing.
I genuinely don't know which one yet.
Can blockchain actually break the extraction cycle that every major platform has followed? Or will OpenLedger eventually face the same pressures?
@OpenLedger $OPEN #OpenLedger

I want to start with a distinction that almost nobody is making.
There are two separate problems inside AI's broken data economy.
The first problem is attribution. Who contributed what. Which datasets trained which models. Tracking the lineage of AI intelligence back to its human sources.
OpenLedger is working on this problem. Proof of Attribution, Datanets, on-chain contribution records. Real infrastructure for a real problem.
But there's a second problem. Quieter. Harder. Almost entirely ignored in the current conversation.
Pricing.
Not paying contributors that's the easy part once attribution exists. The hard part is how much should each contribution be worth?
Here's why this matters more than most people realize.
Imagine three data contributors to an AI medical diagnosis system.
Contributor A uploads 10,000 general health records. Useful, but generic. This data helps the model understand basic patterns.
Contributor B uploads 500 rare disease case studies from a specialized clinic. Rare, precise, hard to find anywhere else. This data helps the model identify conditions that would otherwise be missed.
Contributor C uploads 50 highly detailed longitudinal patient studies following rare conditions over 20 years. Irreplaceable. This data fundamentally changes what the model can diagnose.
If the system pays purely based on volume Contributor A gets the most. But Contributor A's data may have contributed the least actual value to the model's most important outputs.
If the system pays based on influence you need to measure not just whether data was used, but how transformatively it was used. Whether it pushed the model's capabilities in ways nothing else could.
That's a completely different measurement problem.
Current attribution systems including OpenLedger's Proof of Attribution are primarily solving for the first layer tracking usage. Which data influenced which output.
But usage isn't the same as value creation.
A piece of data can be "used" a thousand times in ways that barely move the needle. Another piece of data can be "used" once and fundamentally change what a model is capable of.
Paying equally for unequal value creation isn't fair attribution. It's just slightly more transparent mis-allocation.
This matters economically for $OPEN in a way nobody is discussing.
If OpenLedger's attribution system pays contributors based on usage frequency rather than value impact, it creates a predictable distortion. High-volume, low-quality data floods the Datanets because it's easy to produce and still gets paid. Rare, high-value, hard-to-produce data gets relatively undercompensated because its contribution is harder to measure.
Over time, Datanets fill with noise. Signal gets crowded out. The models trained on OpenLedger's infrastructure become less valuable. Developer adoption slows. Token demand weakens.
This isn't hypothetical. It's the exact dynamic that destroyed early content platforms Medium, early YouTube, early Substack. Pay equally for all content and you get quantity over quality until quality producers leave for environments that recognize their actual value.
The solution is not simple. I'm not pretending it is.
Value-weighted attribution requires answering questions that get philosophically uncomfortable fast.
Who decides which data created more value? The developers who built the model? The users who benefited from its outputs? Some automated on-chain mechanism?
Each answer creates different incentive structures. Each has different failure modes.
But here's my honest take.
OpenLedger has built something real and important. Proof of Attribution is genuine infrastructure for a genuine problem.
The next frontier pricing contribution value rather than just tracking contribution existence is where the system either becomes transformative or stays interesting-but-limited.
Attribution without pricing is an accounting system.
Attribution with pricing is an economy.
The difference between those two things is the difference between a project that matters for a cycle and one that matters for a decade.
I'm watching to see which one OpenLedger builds toward.
Do you think data quality and data quantity should be compensated differently? How would you design a fair system?
@OpenLedger $OPEN #OpenLedger

The global crude oil market is transitioning from a period of acute, geopolitically driven structural deficits into an era defined by macro demand cooling and unprecedented non-OPEC+ supply diversification. For institutional allocators and commodity desks, navigating this landscape requires looking past short-term volatility and analyzing the two distinct tranches of the upcoming cycle.
Phase 1: Residual Tightness & The Geopolitical Premium (Q2–Q4 2026)
The near-term macro picture remains tethered to the friction of recent infrastructure disruptions and transit bottlenecks in the Middle East. While physical-to-futures price disconnects have begun to normalize from their spring peaks, the market enters the summer driving season in a structural deficit, with global inventories drawing aggressively.
Supply Cracks: The formal exit of the United Arab Emirates (UAE) from OPEC alters the cartel's collective spare capacity framework, shifting unilateral pricing power and leaving the group's effective spare buffers tighter than historical averages.
The Atlantic Rebalancing: To bridge the gap, non-OPEC+ production led by the Americas (the US, Brazil, and Guyana) is expanding at a clip of 1.5 million barrels per day (mb/d).
Expect Brent crude to find a volatile floor in the high $80s to low $90s through the third quarter, sustained by tactical inventory replenishment and non-OECD strategic stockpiling.
Phase 2: The Macro Downcycle & The Looming Oversupply (2027)
As we look toward 2027, the structural cycle pivots sharply. The market is transitioning toward a regime of demand destruction and cyclical oversupply.
[2026 High Real-World Draws] ──> [Supply Diversification] ──> [2027 Demand Cooling & Surplus]
High baseline energy costs and broader macroeconomic cooling are weighing heavily on global demand. Refined product markets, particularly in the petrochemical and aviation sectors, are starting to signal a structural slowdown.
As logistics bottlenecks resolve and Middle Eastern volumes gradually normalize, the compounding impact of surging Atlantic Basin supply will flip the market balance from a deficit into a pronounced surplus.
The Long Horizon: Both the EIA and institutional consensus point toward Brent drifting down toward an average of $79/bbl by mid-2027.
```
CRUDE MARKET BALANCES & BENCHMARKS (HISTORICAL & FORECAST)

140 ───┐
│ ▲ (Apr '26 Peak: ~$138)
120 ───┤ ╱ ╲
│ ╱ ╲
100 ───┤ ╱ ╲
│ ╱ ───────► [Q2-Q4 '26 Range: $89-$106]
80 ───┼────────────────/───────────────────────────────
│ (2025 Avg: ~$69) ╲
60 ───┤ ╲────────► [2027 Target: ~$79]
│
0 ───┴───────────────────────┬───────────────────────┬───────────────────────►
2025 2026 2027
```
The Tactical Takeaway
The upcoming macro cycle belongs to the bears. The margin of safety for long-only commodity exposure is thinning. Alpha will be found not by chasing geopolitical spikes, but by positioning for a structural oversupply as the global economy cools and alternative supply lines solidify.
#crudeoil #commodities #MacroTrading #PostonTradFi $USOon

I want to talk about something that's been bothering me for months.
Not token price. Not market cap. Something more structural.
Every major AI breakthrough of the last five years was built on the same foundation human knowledge, human creativity, human labor, accumulated over decades and made freely available on the internet.
Books. Research papers. Code repositories. Forum discussions. Creative writing. Medical literature. Legal analysis. Personal blogs.
All of it scraped, processed and fed into models that now generate billions in revenue.
The people who created that foundation?
They were never asked. They were never paid. Most of them don't even know their work is inside the models that are slowly replacing them.
This isn't a conspiracy. It's not even illegal yet. It's just what happens when an industry moves faster than the economic frameworks designed to govern it.
But here's the crack in the foundation.
AI is no longer just a consumer product.
It's moving into healthcare. Finance. Legal services. Insurance. Infrastructure. Defense.
In these industries, "we don't know where our training data came from" is not an acceptable answer. It's a liability.
Imagine a medical AI that recommends a treatment protocol. It's wrong. A patient is harmed. The hospital asks: what data influenced this recommendation? Who contributed it? Was it verified? Was it biased?
If nobody can answer those questions if the entire contribution chain is invisible then accountability becomes impossible. Impossible accountability means unbounded legal exposure.
This is the crack.
AI built its intelligence on an invisible foundation. As long as AI stayed in the consumer entertainment space, invisibility was fine. The moment AI entered regulated industries which is happening right now, faster than most people realize invisibility became a structural problem.
This is where OpenLedger becomes interesting in a way most "AI blockchain" projects don't.
Most AI crypto projects are solving for speed. More compute. Faster inference. Cheaper deployment.
OpenLedger is solving for something harder.
Provenance.
Proof of Attribution doesn't just track who contributed data. It creates a cryptographic record of how that data influenced model outputs. Every dataset. Every training step. Every inference. Recorded on-chain and traceable.
That sounds technical. The implications are anything but.
It means for the first time, the invisible foundation of AI becomes visible. Auditable. Accountable.
And because it's on-chain — because the record exists independent of any single company's database it can't be quietly edited when inconvenient.
Now let me be honest about what's hard.
Measuring data influence at scale is genuinely difficult. Modern AI models don't maintain neat ingredient lists. They absorb patterns probabilistically across billions of parameters. Determining exactly which data contributed to which output at the scale of frontier models is an unsolved technical problem.
OpenLedger's current implementation works best with specialized, smaller models. How it scales to larger systems is still an open question.
There's also the adoption challenge. Enterprises are conservative. They don't adopt new infrastructure because the thesis is elegant. They adopt it when the pain of not adopting becomes greater than the friction of changing.
That tipping point hasn't arrived yet.
But it's coming.
The New York Times lawsuit against OpenAI. Getty Images versus Stability AI. The EU AI Act's transparency requirements. Pending legislation across multiple jurisdictions demanding AI companies disclose training data provenance.
The legal and regulatory pressure on AI's invisible foundation is building simultaneously in courts, parliaments, and boardrooms across the world.
OpenLedger isn't building for a hypothetical future.
It's building for a present that's arriving faster than most people expect.
Here's the question I keep sitting with.
Every major technology transition eventually produces infrastructure that nobody noticed building until it was everywhere.
TCP/IP. SSL certificates. SWIFT. The cloud's underlying settlement rails.
None of these were exciting when they were being built. They were boring. Technical. Hard to explain at dinner parties.
But they became the invisible architecture that everything else ran on.
AI needs that architecture for attribution and provenance. Right now, it doesn't exist at scale.
OpenLedger is one of the few projects seriously attempting to build it.
Whether it succeeds depends on technical execution, enterprise adoption, regulatory timing, and a dozen other variables that nobody can fully predict.
What I do know is this.
The crack in AI's foundation is real. It's getting wider. And the industry that figures out how to fill it how to make AI's invisible foundation visible, auditable, and economically fair will be building infrastructure that lasts for decades.
That's either the most important bet in this cycle.
Or an elegant idea that arrives too early to matter.
I honestly don't know which one yet.
But I know the crack is there.
I know most people haven't looked down to see it.
Do you think AI's data problem gets solved by regulation, by infrastructure, or does it never really get solved at all?
@OpenLedger $OPEN #OpenLedger

I want to start with a number.
$500 billion.
That's the estimated value of the global AI market. The models powering it were trained on decades of human knowledge books, articles, code, art, research, conversations. Virtually none of the people who created that knowledge received compensation.
This isn't controversial. The AI companies don't really deny it. They just argue it's legal. Or necessary. Or that the concept of "paying for training data" is too complicated to implement at scale.
OpenLedger is betting that last argument is wrong.
The problem with AI's data economy isn't malice. It's architecture.
Centralized AI development has no built-in mechanism for attribution. When OpenAI trains GPT on internet text, there's no system tracking which specific documents influenced which specific outputs. The data goes in. The model comes out. The chain of contribution is invisible.
Invisible contribution means invisible compensation. You can't pay someone for work you can't trace.
This is where Proof of Attribution changes everything not as a feature, but as infrastructure.
Proof of Attribution cryptographically records the lineage of every dataset, every training step, every model inference on-chain. It doesn't just track who uploaded what. It tracks influence how much a specific data contribution shaped a specific model output.
That's the hard problem nobody else has seriously attempted to solve at the protocol level.
Because solving it requires two things simultaneously: the computational ability to measure data influence across complex model architectures, and the economic infrastructure to route payments based on that measurement automatically.
OpenLedger is building both.
But let me be honest about where the skepticism lives.
Influence measurement in large AI models is genuinely hard. The June 2025 Proof of Attribution whitepaper describes approaches that work for smaller, specialized models. How these methods scale to frontier-level systems models trained on trillions of tokens across billions of documents is still an open technical question.
There's also the cold start problem. Datanets need contributors to attract developers. Developers need active Datanets to build useful applications. Getting both sides of that marketplace moving simultaneously is where most Web3 infrastructure projects quietly fail.
And then there's $OPEN 's token dynamics. With 21.55% of supply currently circulating and 48 months of ecosystem/community unlocks ahead, consistent supply pressure is real. The token needs genuine network demand actual AI developers paying for data access, actual contributors earning from model usage to absorb that supply meaningfully.
Here's why I think the timing might actually be right despite those challenges.
AI's data problem is getting louder, not quieter.
The New York Times lawsuit against OpenAI. The Getty Images case against Stability AI. The EU AI Act's transparency requirements. Pending legislation in multiple jurisdictions requiring AI companies to disclose training data sources.
OpenLedger isn't building for a hypothetical future where data attribution matters. It's building for a present where that question is already being litigated in courts and parliaments simultaneously.
Enterprise AI adoption is accelerating into healthcare, finance, and legal services industries where "we don't know where our training data came from" is not an acceptable answer. Verifiable data provenance isn't a nice-to-have for these sectors. It's a compliance requirement.
Polychain Capital doesn't lead $8 million seed rounds in projects without a credible path to real adoption. That's not a guarantee. But it's a signal worth taking seriously.
The deepest question OpenLedger is asking isn't technical.
It's philosophical.
Who should benefit from AI?
The current answer, by default, is: the companies with the compute to train the models and the distribution to deploy them. Everyone else the writers, researchers, artists, developers whose work made those models possible participates as users, not owners.
OpenLedger is attempting to make "owner" the default status for anyone whose work contributes to AI.
That's either a utopian idea that can't survive contact with economic reality.
Or it's the most important infrastructure bet in the current cycle.
I keep coming back to one simple observation.
The data that trained AI was created by humans. The value that AI generates should flow back to humans.
Right now it doesn't. OpenLedger is the most serious attempt I've seen to change that.
Whether it succeeds is still an open question.
But the question itself is finally being asked at the right level.
Who do you think should own the value AI creates the companies that build the models, or the people whose data trained them?
@OpenLedger $OPEN #OpenLedger

There's a number that keeps bothering me.The global AI market is projected to hit $500 billion. The companies building AI are valued in the trillions. The models are getting smarter every month.And the people whose data made all of that possible? They got nothing.Not a percentage. Not a credit. Not even an acknowledgment.This isn't a conspiracy. It's just how the system was built. Data was treated as a raw material abundant, cheap, essentially free. You wrote a blog post, published research, created art, contributed to open source. That work got scraped, processed, and fed into models that now compete with you in your own field.The people who built AI didn't pay for the ingredients. They just took them.OpenLedger is the first project I've seen that treats this as a structural problem worth solving at the protocol level not with policy, not with lawsuits, but with infrastructure.The core idea is called Proof of Attribution.It sounds technical. The implications are anything but.Proof of Attribution means every dataset, every model, every AI output can be traced back to its source contributors on-chain. Not approximately. Cryptographically. If your data influenced a model's output, the protocol knows. And because it knows, it can pay.Automatically. Every time that model is used.This is the "Payable AI" concept and it's more radical than it first appears.Most AI monetization today works like this: a company trains a model on your work, deploys it as a product, and charges users. You are not in that revenue loop. You never were.Payable AI inverts that. The revenue loop includes contributors by default. Not as a charity. As a structural requirement of how the system operates.Now, let me be honest about the challenges.Proof of Attribution is technically ambitious. Tracking exactly which data influenced which output, at scale, across millions of contributors and billions of inferences that's an extraordinarily hard problem. The June 2025 whitepaper describes two approaches for smaller models. How it scales to frontier-level systems is still an open question.There's also the adoption problem. OpenLedger needs AI developers to build on its infrastructure instead of the existing centralized alternatives. That's a classic chicken-and-egg challenge. Contributors want to join when developers are using the network. Developers want to build when contributors have filled the Datanets. Getting both sides to move simultaneously is where most infrastructure projects fail.The token dynamics are worth watching carefully. With 21.55% of supply currently circulating and significant community/ecosystem unlocks scheduled over 48 months, $OPEN faces consistent supply pressure. Whether organic demand from actual network usage grows fast enough to absorb that supply that's the question that will determine whether the token reflects the project's genuine utility or just its narrative.But here's what makes me take OpenLedger seriously despite those challenges.The problem it's solving is real and getting more urgent.AI training data lawsuits are multiplying. Regulatory pressure around data provenance is increasing the EU AI Act is just the beginning. Enterprise adoption of AI is accelerating into industries where auditability isn't optional, it's legally required.OpenLedger isn't chasing a trend. It's building infrastructure for a problem that is going to get louder, not quieter.Polychain Capital led the seed round. That's not a guarantee. But it's a signal that people who evaluate infrastructure bets seriously thought this one was worth making.The question I keep sitting with is this.We've spent a decade building financial infrastructure on blockchain — DeFi, NFTs, stablecoins. Most of it serves the same relatively small group of crypto-native users.OpenLedger is attempting something different. Infrastructure for the AI economy. Attribution rails for a world where data has real, measurable, on-chain value.If that works if even a fraction of the AI industry's data supply chain moves through verifiable attribution infrastructure $OPEN isn't priced for that world yet.If it doesn't work if the technical challenges prove unsolvable at scale or adoption never materializes then it's another ambitious thesis that couldn't survive contact with reality.I don't know which outcome comes next.But I know the problem is real. I know most projects aren't even trying to solve it.
Do you think blockchain can actually fix AI's data problem? Or is this too ambitious to execute?
@OpenLedger $OPEN #OpenLedger