New and better data generation, data marketplaces

Large compute network can compete with centralizing force of big tech

Alignment of human needs and AI models via incentives or slashing

Payment rails for AI agents / models

faster development via more contributors and battle testing

Independent auditing of models and data

Collaborative development and ownership of models, agents via novel payment, verification, and funding mechanisms | Story protocol

Grass

Bitcoin power consumption and hashing power

0G Ai alignment nodes

AI wayfinder

Linux

Morpheus

sentient |

Next-generation foundation model training

|

Serving layer (offering marketplace)

Storage and compute close by routed via network

Initial Training Offering (model training) |

Build: serving marketplace

Build: 0G storage into AI workflow

Partner: AI execution environments for finetuning

Build: data traceability, model ID, ML optimizations

Partner: algorithms (for now)

Partner: GPU clusters, and ITM funding platform

Build: on-chain to off-chain bidirectional payment rails |

| | Provenance (tracked data) | Better data Transparency & data authenticity to reduce corruption or misleading insights

IP management and attribution

Fingerprinting of deep fakes | Food supply chains, ai model training pipelines, mind attack prevention

Story protocol Open ledger

| SLMs that outperform LLMs

Proof of humanity, proof of what is real

Agent IDs to track and hold agents accountable | 0g storage utilized for this data

| Build: ML specific storage and database needs built on 0G storage/logging

Build: 0G storage access control built in, proof of humanity ID with traceability and privacy

Build: programmable economy rails; dAIOS already support agent-agent payments |

| | Coordinate underutilized hardware (storage, gpu, etc.) | up to 90% lower AI ops costs (inference, training, etc.)

more resources to run more AI operations

electricity arbitrage/ easier to source many small sources of power | Coreweave, TogetherAI, Foundry are all web2 companies providing cost-effective ML workloads, optimization and orchestration

AI takes 9% of US electricity by 2030 | Cost-competitive inference to further boost AI adoption

| Decentralized storage expertise

Proof of useful work | Build: router, delegating training/inference to cheapest electricity and compute

Build: OS acts as proving system

Partner: GPU providers, GPU hosting platforms, compute platforms, ranked by price/electricity cost/efficiency/reliability | | | Verification | Know exactly which model or agent produced which result that can be reproduced | Mistral and European governments endorsing a semi-open model with on-prem service | Smart Cities

Societal use cases (transportation, logistics, manufacturing, humanoid automation) | Agent accountability framework | Partner/Incubate: Public knows which model is in use and track accountability end-to-end from 0G storage, training, inference to payment settlement | | | Privacy-preservation

| FHE, ZK, etc. technologies preserve privacy including bringing your private data for public consumption

True ownership | Nvidia Flare for hospitals to train one model across proprietary data

Users retain ownership of data | Federated training

Higher quality models trained from proprietary data sets

| 0G storage private data to 0G serving for private training | Build: from data to training end-to-end blockchain flow

Partner: ZK, TEE, FHE providers | | | Censorship resistance | No single points of failure / resilience

Community determines how a model evolves and avoids OpenAI self censorship

Overcome proprietary data vaults | Trust in OpenAI lower than any other big tech aws failing > outage of system

Twitter sabotaging users’ legal right to export data | Uptime SLA guarantee

| Decentralized rail ensuring 100% uptime

ITO lets community have a say in models

BYOD - bring your own data | Build: router with fallback mechanism to ensure no single point of failure

Build: model listing, cluster listing, and data listing mechanisms, just mechanisms

Partner: model providers, cluster providers, data providers | | | Security | Transparency and consistent monitoring for tempering like data poisoning (esp. via incentives)

Monitor for malicious training data and bias

AI models auditing and monitoring smart contracts | Inference observability companies like Arthur.ai now need to offer training observability

AI has been used in cyber-security for decades |

Determining community standards to define bias (it’s all relative to the community in subject)

AI for smart contracts | 0G rails are decentralized, global, and community-driven in nature | Build: 0G’s ITO and foundation governance gives communities a voice

Partner: provide decentralized models to smart contract auditing companies | | | UX Automation (AI supporting Web3) | Abstracting away poor UX like bridges and wallets

Bots and NPCs for onchain gaming

trading automation

efficient search of onchain data

automating coding & smart contracts based on intent

Creating analytics (dashboards and reports)

AI agents managing DAO, or economy, or staking rates | Coframe: self-improving internet, self-improving UI

Overworld

(not bullish)

Kaito

Slate

theoriq

autonolas | | | Partner: agent builders and model builders that run on 0G infrastructure

Incubate: teams that deploy agents for particular use cases on 0G rails

Build/Partner: one-click agent deploy to priority web3 use cases, such as intents, analytics, auditing, DAOs | |

Visioning

1. What is Our Winning Aspiration?

• Make AI a public good
• Where do we see the future going?
  • World of abundance > work is a choice for fulfillment
  • Automation of most processes with AI (androids,

2. Where Will We Play? (Markets and Geographies: Deciding which regions, countries, or segments to focus on. Product or Service Categories: Determining which lines or niches align best with the company’s strengths. Customer Segments: Identifying which types of customers to target.

• Markets and Geographies:
  • Retail: Korea, China, USA, Turkey, LATAM, CIS
  • Devs: Vietnam, USA, CIS, Africa?
• Product or Service Categories:
  • Integrated trading experience copying binance
• Customer Segments:
  • Web3 degens

3. How Will We Win? (Value prop: How will your product or service stand out? Will it be through innovation, cost leadership, customer experience, or some other unique feature? Competitive advantage sustainability: What factors will ensure that your advantage is hard for competitors to replicate over time?)

• Fastest infrastructure
• Modular and Open Architecture: 0G’s decentralized operating system is built to be modular, allowing users and developers to customize and innovate rapidly. This openness creates a network effect where contributions continuously enhance the platform.
• Community-Driven Innovation: Leveraging a global community of developers and AI experts, the platform encourages collaborative breakthroughs that are both rapid and resilient.
• Replace choke holds on industry with decentralized offerings (e.g., exchanges)

4. What Capabilities Must Be in Place?

• Core Competencies: What unique strengths or processes will support the strategy?
• Talent and Skills: Which expertise and organizational capabilities are essential?
• Infrastructure: What technological or operational investments are necessary?