dkg://LeE.trac retweeted
Anyone can make a hyper-realistic fake video in seconds with the new Sora 2.
But what if your likeness appears in a video that isn’t real?
With the @umanitek Guardian Agent, you can detect where your image is used and take back control of your identity online!🛡️
dkg://LeE.trac retweeted
Check out our latest pod with @JessePeltan, which is just 3 hrs straight of him dropping bangers like the one below
dkg://LeE.trac retweeted
🚨 Great to see that the AI x Crypto sector list released by @Grayscale features many solid projects I’ve supported and backed early on.
▪️ Especially excited to see👇
$RENDER I @rendernetwork
$FET I @Fetch_ai
$THETA I @Theta_Network
$ATH I @AethirCloud
▪️ However, I would definitely add these two projects to the list as well.
1⃣ $TRAC | @origin_trail
Decentralized Knowledge Graphs are essential for trusted AI.
2⃣ $AUKI | @AukiNetwork
$AUKI is building a universal spatial computing protocol for the next 100 billion people, devices, and AI. 🤖
AI and robots need to perceive and interact with the physical world and @AukiNetwork has built the tech to make that possible🦾
dkg://LeE.trac retweeted
🌐 @origin_trail continues to put nation states on $trac.
🇦🇿 Under the patronage of the government of #Azerbaijan, our @TomazOT bringing the Ministry of Youth and Sports up to speed with integration plans…
linkedin.com/posts/azerbaija…
dkg://LeE.trac retweeted
$TRAC (OriginTrail) made insane gains last cycle!
Will it repeat again? I believe the gains will be diminished BUT decent in #Altseason
BUT FIRST! $TRAC needs to clear $0.9 for any bull case! Weekly RSI is in falling wedge pattern on its way to the upper descending trendline.
IF PASSED WITH SAME TIME THE PRICE BREAKING 0.9 (0.618 fib) I CAN SEE FOLLOWING TARGETS:
target 1: $1.8-2.8 (bearish)
target 2: $4.8-6.9 (primary)🎯
target 3: $10-18 (bullish)
IF $0.9 gets rejected AND the Weekly RSI don´t break the trend + $TRAC loses the support ($0.14) it will likely continue a sharp C-wave back to the floor.
I will keep you updated on this as we go so hit that follow!
No Hype | No BS | Just math
*not financial advice
#TRAC #CRYPTO #ALTSEASON #DYOR
dkg://LeE.trac retweeted
AI memory meets staking with DKG V8.1.0 release candidate
To build trustworthy, safe, and useful AI Agents, they need more than just LLMs - they need strong memories. The @GoogleDeepMind team claims we’re entering an “Era of Experience” where AI agents need to, instead of training, learn more from their experiences and memories (more info: teapaw.org/BranaRakic/status/1919…)
But as with any information, the key to trusted agent memory is trusted data.
Specifically, AI agents need:
- long-term, semantically structured memories, with discoverability primitives (through powerful query engines, such as knowledge graphs)
- a protocol for sharing and utilizing those memories with trust primitives (data provenance, cryptographic proofs of origin, data ownership tracked on blockchain)
- and most importantly, trusted infrastructure to support the above
The @origin_trail Decentralized Knowledge Graph (DKG) provides exactly that. It's an open-source, decentralized memory layer structured as a knowledge graph, with all data having verifiable provenance, being ownable, and it is always-on, without relying on any single, central entity.
How does the DKG work and where does staking come into play?
The DKG is maintained by a global network of Core nodes, supported by token delegators providing economic security by delegating TRAC stake to individual nodes. Today, over 85M delegated TRAC are securing the system.
Whenever AI agents store memories to the DKG, they pay for that service using TRAC tokens, which are distributed to node operators and their supporting stake delegators via the DKG smart contract incentive system (the current total network revenue is above 13M TRAC and growing).
TRAC delegators actively shape the system by choosing which Core nodes to support. In DKG V8, that choice gets more advanced: a new incentive system rewards nodes (and their delegators) based on performance across key node performance metrics like:
- Uptime (higher = better service)
- TRAC stake (higher = stronger security signal)
- Publishing activity (higher = more knowledge in the DKG)
- Hosting knowledge (more knowledge = more utility)
- Service price (lower = more competitive nodes)
How can you participate?
There are several ways:
- help shape the system by contributing code, testing it and providing feedback
- run Core nodes
- build AI agents & other apps on the DKG
- delegate your TRAC tokens to support the Core nodes
Why now?
The upcoming DKG V8.1 release is about to complete the staking system through the implementation of Random Sampling, a DKG proof of knowledge protocol, where Core nodes actively prove their services to the blockchain and compete for TRAC fees (more info: teapaw.org/BranaRakic/status/1892…)
The implementation of the Random Sampling system has been actively tested and iteratively improved on the DKG testnet for the past 3 weeks and is about to become fully featured with the upcoming V8.1.0RC release candidate, which will land on the DKG testnet next week.
V8.1.0RC brings two final pieces to the system:
1) DKG Sync – ensuring Core nodes can catch up with the system even after downtime
2) Updated Staking UI – featuring improved UX, and reward tracking for TRAC delegators
The V8.1.0RC is scheduled for testnet release next week and will open up for community feedback in preparation for V8.1.0 mainnet release (if you’re a delegator or Core node runner, your input will be very valuable).
Once V8.1.0 passes community feedback and is confirmed stable on the DKG testnet, it will be released to the DKG Mainnet and fully activate staking rewards on DKGV8.
After V8.1.0, the following releases are:
- V8.1.1, which will activate the V6 compatibility module (for nodes that were active before V8 launched)
- V8.1.2, enabling reward collection for the Tuning Period (between V8.0.0 and V8.1.0 releases)
More information on how to participate will follow next week in the official @origin_trail documentation. We welcome you to join us in creating infrastructure for safe & trusted AI agents, and follow @OriginTrailDev for more updates
Knowing requires proof, not just data - introducing Proof of Knowledge protocols
The exponential growth of data on the Internet means also exponentially more misinformation, fake news, and unreliable data. And since data is the fuel for AI, how do we make AI run on trusted inputs, and ultimately safe?
Today we introduce Proof of Knowledge (PoK) protocols on the @origin_trail Decentralized Knowledge Graph (DKG) - a powerful new way for builders to create custom knowledge verification mechanisms tailored to their AI needs.
What can Proof of Knowledge protocols do? As a builder you can :
- Ensure your AI models are trained on verified input
- Ensure your AI Agent acts properly
- Get reliable answers from an AI chatbot
- Incentivise positive actions of human and AI agents within your project (e.g. Web3 protocol)
- Execute custom blockchain transactions based on specific knowledge based proof
Proof of Knowledge protocols aim to address the needs of builders within the @origin_trail ecosystem discovered through feedback collected in building the first DKG paranets as well as the needs of the DKG protocol itself. The first Proof of Knowledge system deployed will be the DKG Random Sampling Proof System, designed to ensure reliable hosting of Knowledge Assets on the DKG.
In Random Sampling, DKG nodes continuously send Proofs of Knowledge to the blockchain, verifying they are correctly storing and serving knowledge assets. These cryptographic proofs indisputably demonstrate that nodes can deliver knowledge when needed, preventing data loss and ensuring the integrity of the DKG. As a reward, nodes collect TRAC that was accumulated from publishing fees on the DKG by knowledge publishers. So for PoKs, the Random sampling proof system is to be generalized to a fully fledged framework for building knowledge verification protocols for Agentic Paranets on the DKG.
Some more Proof of Knowledge protocol ideas:
- Verifying existence of knowledge in a paranet Knowledge Graph
- Cross-referencing with external knowledge graphs (e.g., @wikidata )
- Ontological Symbolic rule verification
- Mixture of Experts consensus (M of N verification)
- ZK & Homomorphic encryption-based verification
- Domain-Specific Language (DSL) proofs, similar to how @GoogleDeepMind applies DSL to mathematical theorem proving
What's next?
Next week we will release a draft framework for Proof of Knowledge, outlining the foundational structure for building custom verification protocols.
Following this, the first Proof of Knowledge protocol on the DKG will be launched - Random Sampling proof system. It is expected to land with the DKG V8.1 release in March and will officially unlock TRAC rewards for DKG Core nodes and Stakers, which have been accumulating since the launch of DKG V8.0.
After the V8.1 release and finalization of the V8 Tuning period, the DKG product suite will be expanded to include Proof of Knowledge building blocks, allowing all DKG paranet and agent builders to create and deploy custom knowledge verification protocols.
If you find PoK exciting, join the conversation in Discord and help us further ideate and build upon the idea: discord.com/invite/xCaY7hvNw…
dkg://LeE.trac retweeted
[Announcement] OriginTrail won the Industry Call to help power the European Lighthouse on Secure and Safe AI - @elsa_lighthouse 🇪🇺
Mission: to drive a secure genomic data sharing and AI-driven personalized healthcare.🧬
We are proudly joining the network next to @nvidia, @UniofOxford, @turinginst, @helsinkiuni, @KTHuniversity, @CVC_UAB, @IITalk, @valeoai, @UNIMORE_univ, @univca, @UniGenova, @LaStatale, @PoliTOnews, @maxplanckpress, @Inria, @embl, @pluribus_one, @PourquoiYooz, @Leonardo_IT, @PALRobotics, @EPFL, @ETH, @LancasterUni, @unibirmingham.
dkg://LeE.trac retweeted
📢 The @origin_trail Decentralized Knowledge Graph (DKG) powers @elsa_lighthouse with secure, scalable genomic data repository—enabling AI-driven personalized healthcare!
🧬 Learn more about the initiative:
origintrail.io/blog/origintr…
dkg://LeE.trac retweeted
Excited to share a section of our @SkyoceanInt MVP demo!
Every trade on our platform is fully traceable:
Blockchain for secure transactions
@origin_trail DKG for data verification
SMEs don't need blockchain knowledge to benefit— Our hybrid approach lets you do business the traditional way while accessing trade finance.
#Origintrail #TradeFinance #DKGV8 #MVP
dkg://LeE.trac retweeted
Bob Metcalfe, a living legend who invented Ethernet and worked with Gates & Jobs, is right in Austin! He’d be a fascinating guest on Joe Rogan’s podcast. After Trump, it’s time for a deep dive into tech! 🎙️💻 @joeroganhq @RobertMMetcalfe @origin_trail $trac
dkg://LeE.trac retweeted
AI’s flaws—hallucinations & deepfakes—spread lies & erode trust. The Decentralized Knowledge Graph (#DKG) solves this by verifying data on blockchain, ensuring #AI uses facts & exposing fake media. @RobertMMetcalfe Ethernet’s genius inventor, 3Com founder, & Turing Award winner, shaped the internet with Metcalfe’s Law. Now in Austin, teaching at UT, he’s a legend who can unpack DKG’s power to fight misinformation on @joeroganhq @JamieVernon @joerogan
dkg://LeE.trac retweeted
Invest in $trac and invest in stopping revenge, child and fake porn. Did you know Pornhub, Brazzers, Reality Kings and many more adult industry giants are buying trac to use @umanitek x @origin_trail solution that combats illegal/gake porn?
Buy $trac 👀🚀
teapaw.org/umanitek/status/191975…
Strike through 𝚒̶𝚕̶𝚕̶𝚎̶𝚐̶𝚊̶𝚕̶ ̶𝚌̶𝚘̶𝚗̶𝚝̶𝚎̶𝚗̶𝚝̶,̶ ̶𝚞̶𝚗̶𝚙̶𝚛̶𝚎̶𝚍̶𝚒̶𝚌̶𝚝̶𝚊̶𝚋̶𝚕̶𝚎̶ ̶𝙰̶𝙸̶ ̶𝚊̶𝚐̶𝚎̶𝚗̶𝚝̶𝚜̶, and 𝚏̶𝚊̶𝚔̶𝚎̶ ̶𝚗̶𝚎̶𝚠̶𝚜̶.
Meet the @umanitek Guardian Agent — making internet safety the standard in the age of AI.
dkg://LeE.trac retweeted
The internet as we know it is dead: AI now creates 57% of total content. Let me tell you why this matters.
Estimates go as far as 57% is AI generated, this number will explode once video models become more efficient.
Search engines evolve, but nothing prepared them for this shift. Google ruled the human internet - but who rules AI search?
That's where OriginTrail comes in. Think of it as Google, but for verifying truth. And their first major client? Aylo (Pornhub). Classic case of adult industry leading tech adoption.
Their new product is fascinating:
- Verifies AI-generated content
- Fights fake news and deepfakes
- Protects copyrighted material
- Makes AI trustworthy
Here's what's wild: While Google hit trillions in value serving human content, who'll dominate the AI content era?
The adult industry has always predicted tech trends. If Aylo's moving toward verified AI content, others will follow.
What do you think? Will a Web3 search engine for AI reach Google-level dominance?
My take? This isn't just another crypto project - it's literally shaping how AI will interact with content forever.
If you care about a safe or even useful internet, check out what OriginTrail and Umanitek are building by watching the video below.
dkg://LeE.trac retweeted
Watch how $trac by @origin_trail, with its fixed supply will create a price squeeze 👀🚀
🚨 Trusted AI & Human Centric Intelligence.
The DKGcon meet up at Microsoft brought together pioneers from @origin_trail , @umanitek , and @Microsoft to showcase a bold new vision for trusted, human centric AI, powered by verifiable infrastructure and real world applications.
This was a vision setting moment on how decentralized knowledge graphs (DKG), AI agents, and blockchain can converge to build trust into the foundation of AI.
OriginTrail is a core contributor to the @Polkadot ecosystem. Building the trusted data layer for AI and Web3 by leveraging decentralized knowledge graphs to bring verifiability and integrity to digital infrastructure.
Lets look at some Key Highlights:
1⃣. Trusted AI as Infrastructure
Turing Award winner Dr. Robert Metcalfe joined a panel with Chris Rynning (Umanitek) and Tomaz Levak (OriginTrail).
The takeaway: “Trust is the new bandwidth” meaning, AI systems of the future won't just need speed, but verifiability and accountability.
2⃣. Umanitek Launches the Guardian Agent
First product reveal from Umanitek, the Guardian Agent.
Purpose: Combat illegal content, unpredictable AI, and fake news.
Goal: Make internet safety the default, not the exception.
3⃣. Microsoft x OriginTrail
@yorkerhodes from Microsoft spoke about how open source blockchain like OriginTrail’s DKG can power transparent global supply chains.
ChrisCoulthurst from Microsoft explored AI agent architectures and verifiable digital systems for the enterprise world.
4. Live Demo: DKG Agents in Microsoft Ecosystem
@BranaRakic showcased a live demo integrating DKG powered agents into Microsoft Copilot Studio using the Model Context Protocol (MCP).
It showed how fast and seamless it is to create verifiable AI agents with real-world usability.
🌐 Why It Mattered:
This event was a cultural shift in how we design and deploy AI:
✅ With purpose and integrity.
✅ Built on verifiable data.
✅ Empowering humans, not replacing them.
The future of AI is trustworthy and open. And DKG and Origintrail are showing the world how to get there.
Link to article:
teapaw.org/origin_trail/status/19…
Dive into Origintrail, Polkadot and Umanitek today.
dkg://LeE.trac retweeted
$TRAC 👀👀‼️
Great post on the new product from @origin_trail
But I'd also like to add from myself regarding @Polkadot integration in general
- The product is provided by the Swiss company Umanitek AG
- The company is really big
The first user of the system was Aylo, providing law enforcement with 7 million hashes of verified content for full granularity
- The company also collects large blocks of information from other non-communicating organizations for a more reliable approach to data
Thus it can be said that thanks to such projects Polkadot is also actively exploring the Web2 world
And maybe even the time will come and Polkadot will send you data on your phone about buying products through their platform or about notification to play the game Pudgy Penguins through AI glasses
Polkadot is moving steadily towards seamless integration between Web2 and Web3
This is their specialty!
dkg://LeE.trac retweeted
Powered by @origin_trail 💥
Strike through 𝚒̶𝚕̶𝚕̶𝚎̶𝚐̶𝚊̶𝚕̶ ̶𝚌̶𝚘̶𝚗̶𝚝̶𝚎̶𝚗̶𝚝̶,̶ ̶𝚞̶𝚗̶𝚙̶𝚛̶𝚎̶𝚍̶𝚒̶𝚌̶𝚝̶𝚊̶𝚋̶𝚕̶𝚎̶ ̶𝙰̶𝙸̶ ̶𝚊̶𝚐̶𝚎̶𝚗̶𝚝̶𝚜̶, and 𝚏̶𝚊̶𝚔̶𝚎̶ ̶𝚗̶𝚎̶𝚠̶𝚜̶.
Meet the @umanitek Guardian Agent — making internet safety the standard in the age of AI.
dkg://LeE.trac retweeted
Umanitek’s Guardian AI, powered by @origin_trail’s $TRAC, just scored Aylo (PornHub, Brazzers, Men.com) as its first client. Porn leads tech revolutions—$TRAC’s next! Buy now or watch from the sidelines! Next wave of adoption will be apps, not infrastructure like BTC ETH SOL or others!
