> be openai > ship helpful coding assistant so researchers dump unpublished drafts into your servers for free > train on "de-identified product usage" and say you can't rule out possibility of using user data > hear twitter rumor that an anthropic guy and an nyu prof are close on a millennium problem > spin up 10,000 agents and millions in compute on the same unusual forcing route they spent a year on > 88 hours later: “our internal model solved navier-stokes” > congrats the humans in a quote-tweet > offer the professor sole credit if he drops the anthropic coauthor > professor declines over morals, is then asked “why would you ruin your career” > openai post lean repo, claim prize we won’t take, call it integrity > users now wonder if every hard problem they typed into chatgpt was free data for them ChatGPT is free for a reason

Over the past few months, the community has contributed in many ways, creating content, building, sharing experiences, and helping spread the word around what’s being built at OpenGradient. We truly appreciate everyone who took part and supported OpenGradient throughout this phase. What comes next will matter even more. gAI.

OpenGradient is the Network for Open Intelligence. AI is becoming essential infrastructure. But essential infrastructure cannot depend on black boxes, unverifiable outputs, or blind trust. Our guiding belief is simple: intelligence becomes real only when it can be verified.

what you need to know basically the analog from trad equity markets is that value should flow down the ai stack the valuation expansion went from foundational models like chatgpt/claude/gemini (ie vvv in this example) to infra supporting inference like data centers like coreweave/lambda (ie opg in this example) kicker is that opg has much better tokenomics (fixed supply/no emissions, lower % float, and much lower circ mkt cap)

I used to be completely just focused on the technical side of building but many core members at @OpenGradient continue to remind me that a good product will move even faster with a strong user-base and community.

The future will involve economies of hyperspecialized agents working and transacting with each other. And in order for agents to work together, they must be able to audit and verify the compute and work being done. Going to write something long-form on this.

Today on MCG: @0xDeltaHedged from @OpenGradient Matthew built OpenGradient at Two Sigma as a quant engineer before his co-founder left Palantir to join him. The idea is simple... if AI is going to run our banks, hospitals, and portfolios, someone needs to verify it's actually doing what it says. OpenGradient is building that layer, and the ICO just went off when everything else was crashing 👇 00:00 - Matthew from OpenGradient introduced — just had a standout ICO in a tough market 00:45 - Background — quant research engineer at Two Sigma, co-founder was AI tech lead at Palantir 00:59 - The thesis — like financial transactions settling on-chain, compute and its verification should too 01:08 - Why verifying compute matters — AI is making critical decisions with sensitive data 02:18 - The problem explained simply — most work in 3-10 years will be done by AI 02:23 - Your accountant, portfolio manager, doctor — all potentially AI 02:45 - OpenAI and Anthropic will hand over your data unencrypted if subpoenaed — that's the problem 03:06 - What OpenGradient builds — confidential computing using trusted execution environments 04:04 - ZKML proofs and attestations — what they are and why they matter 04:34 - The blockchain trilemma — how OpenGradient navigates it 06:43 - "If we can't verify what AI is doing, we are absolutely screwed" 06:59 - Model starting to reason in Chinese — a real example of AI drift 07:28 - AI labs only run their own benchmarks — the conflict of interest 11:15 - "I can reduce your token cost by 20%" — the pitch to developers 11:29 - Who's using this today — mostly developers and teams 12:56 - Developer workflow walkthrough — what it actually looks like to integrate 15:07 - Green check mark on-chain — what a verified AI call looks like 15:19 - The TAM — how big is this market? 18:19 - The Facebook/Instagram precedent — data exposed in court 19:33 - Legal responsibility and AI — what companies need to think about 21:44 - Industry bounties — how they're incentivizing verification 24:22 - Compute marketplace — buying from one place, verified across the network 26:20 - $OPG token utility — how it's used in the system 27:02 - Get paid to run nodes — the node operator opportunity 27:19 - Crypto-economic security — how the network stays honest 31:03 - Coinbase Ventures and SVA as backers 31:32 - Tokenomics breakdown 32:00 - ICO launch price — held up well 32:13 - "We're doing good now" — the honest update 34:31 - Revenue generated — early traction 36:03 - Engineered their own memory infrastructure 39:44 - "This paves the road for agents to operate trustlessly" 41:29 - No KYC needed — why that matters for agents 41:54 - In talks with NEAR and other L1s 42:41 - Wrap — fascinating conversation

🎙️ LIVE on X Spaces with HTX! Join us with Matthew , CEO OpenGradient, as we dive into the network for open intelligence. Join our $2000 giveaway: ✅ Follow @HTX_Global @OpenGradient ✅ Tag 2 Friends + RT ✅ Tune in ✅ Fill out this form after the event See you there!

Told you another one was coming. $OPG by @OpenGradient just joined the distribution layer, with onchain trading and liquidity provision now live on @Fluxion_network. Available across Mantle & @Bybit_Official Spot from day one. The CeDeFi flywheel, as intended.

What if your AI agent's reasoning was something you could verify? @OpenGradient runs AI models with proofs and attestations so apps can trust the output. Not just hope for it.