Beat Proof of Quality. Win a Tesla.

Send us your hardest dataset and tell us what good looks like. We'll choose five US robotics and world AI teams. Proof of Quality (PoQ) gets three days to meet the agreed evaluation targets. The first team we fail wins the prize.

We'll agree the dataset, output and acceptance criteria before we start. The 72 hour clock starts once access and expert reference judgments are ready. If we miss the deadline or agreed targets, we fail. We'll run the five challenges in a fixed order.

The winner chooses a Model 3, $25,000 cash or a $25,000 charity donation. One prize total. Apply with your work email, dataset size and format, and what makes it difficult. We'll agree on confidentiality before access. No purchase required!

Apply: https://sapien.io/challenge

Know someone whose team could beat Proof of Quality? If you refer the team that wins, we'll pay you $1,000. Ask them to put your name and email on their application. One referral reward, paid within three days of verifying the winner and referral.

One question has shaped much of what we’ve built at Sapien: where can human judgment have the most impact on AI?

When we started, people were doing much of the work of creating and labeling data. After two years and 195 million tasks, we learned how difficult it is to measure quality reliably at scale.

As AI gets better at doing the work itself, the role of human expertise becomes even more important. People need to define what good means, decide which tradeoffs are acceptable, and retain responsibility for the decisions that deserve human judgment.

Where we think human judgment goes next, and why this question is central to Sapien’s long term direction:

https://www.sapien.io/blog/poq-where-human-judgment-goes-next

AI systems can generate thousands of outputs before a team has time to inspect a fraction of them. When outsourcing this to AI, one of the hardest parts is deciding how much confidence to place in a single judgment.

Proof of Quality supports multiple AI reviewers in a single workflow, so teams can compare scores, reasoning, agreement, and disagreement, giving them a faster way to test evaluation workflows at scale while keeping expert attention focused on ambiguity and high consequence cases.

Read more: https://www.sapien.io/blog/proof-of-quality-works-with-ai-and-human-validators

We've been thinking about a modern problem in city intelligence: AI has made it possible for cities to easily collect and process huge amounts of sensor data, however the decisions attached to that data still depend on sparse human expert judgment.

We wrote about that gap, and how Proof of Quality can make judgment measurable in city intelligence.

https://www.sapien.io/blog/proof-of-quality-brings-measurable-evaluation-to-city-intelligence

Robotics teams can collect and process training data episodes faster than reviewers can evaluate whether or not they belong in the training set. Collection, processing and training each have a measured throughput, while the judgment stage between processing and training often does not. That review rate determines when enough trusted data exists to start the next training run.

Proof of Quality measures each episode against the standard your experts set.

How Proof of Quality works with Robotics:

https://www.sapien.io/blog/the-review-queue-between-robot-data-and-the-next-model

AI can draft a vulnerability report in minutes. Verifying it takes judgment.

That gap is the subject of our new post on AI audit triage. When automated discovery expands the queue faster than review capacity grows, the value of an audit moves from the length of the findings list to the quality of the decisions behind it.

New on the blog: The AI Audit Triage Problem: When Finding Bugs Becomes Easier Than Verifying Them

https://www.sapien.io/blog/the-ai-audit-triage-problem-when-finding-bugs-becomes-easier-than-verifying-them

SAPIEN staking rewards are live!

Holders can now deposit SAPIEN into the Sapien Vault on Base and receive vSAPIEN, a token representing their underlying SAPIEN in the vault. As rewards stream into the vault, each vSAPIEN redeems for more SAPIEN.

What that means in practice:

• The cap opens at up to 20% a year, and the full rate applies while the vault holds under 50M SAPIEN, and the rate declines as the vault grows and as the program ages.
• Rewards are funded from existing supply, an allocation of 50M SAPIEN unlocked at TGE. No additional SAPIEN is being created and the emission schedule is unchanged.
• Each deposit can be withdrawn once it is 24 hours old. After that there is no waiting period and no other lockup.
• Depositing alone carries no slashing risk. That exposure only exists for vSAPIEN a holder explicitly locks behind a validator later on.
• The live rate is always published at vault.sapien.io.

Deposits are open in supported jurisdictions at vault.sapien.io, where the live rate, contract details and Terms of Service are published.

Read the full announcement here: https://www.sapien.io/blog/sapien-staking-rewards-program

SAPIEN staking rewards program

Deposit SAPIEN into the Sapien Vault and receive vSAPIEN, a token representing your share of the vault.

We're thrilled to have been invited to Binance Clubhouse Bali at Coinfest Asia!

On August 21, our Smart Contracts Lead, Chad Lynch, will present our recent work on Proof of Quality! For anybody who isn't attending, we will be sharing the highlights across our socials afterwards!

AI has become extremely good at reproducing the appearance of professional reasoning.

The harder question is whether the evidence underneath that reasoning actually supports the conclusion.

Our latest article examines how hallucinations gain authority once they enter legal filings, government reports, research, and other professional documents, then explores how Proof of Quality creates evidence for the review process itself.

See how a rubric, consensus, and a Proof Report can turn review into something people can inspect.

Read the full blog post here: https://www.sapien.io/blog/why-every-ai-generated-document-needs-independent-verification

Why Every AI Generated Document Needs Independent Verification

AI hallucinations keep appearing in legal filings, research, and policy documents. Learn why review fails and how Proof of Quality makes document review verifiable.

Our first pilot on real product data put Proof of Quality through a practical test: could it find bad labels, protect good ones, improve models, and reduce human review?

Across 250 labels, it delivered all four with a verifiable record of the process.

Read the full blog post: https://www.sapien.io/blog/how-proof-of-quality-caught-all-nine-errors-in-an-automotive-training-data-trial

How Proof of Quality Caught All Nine Errors in an Automotive Traini...

See how Proof of Quality found mislabeled ML training data, improved model accuracy, and cut human review by 74% through AI triage.

An AI agent can turn an unsupported assumption into a database change, infrastructure operation, or financial transaction in seconds.

The agent acted. Who verified those actions?

Read now: Why AI Agents Keep Making Confident Mistakes
https://www.sapien.io/blog/why-ai-agents-keep-making-confident-mistakes

Why AI Agents Keep Making Confident Mistakes and How to Avoid Them

AI agents increasingly make real decisions, yet execution logs alone cannot prove those decisions were correct. Learn why agent hallucinations lead to real-world failures and how Proof of Quality creates independently verifiable evidence before AI work is trusted.

The original Sapien staking vault is retiring. The early withdraw penalty is gone.
You can:

Withdraw your full balance to your wallet in one transaction, with no cooldown and no penalty
Migrate into the new vault in one sitting

The new vault has no withdraw restrictions. It will support agentic validation for Proof of Quality (PoQ), Sapien’s verification layer for AI work.
Migrate or withdraw: https://unstake.sapien.io
Learn more about the new vault: https://docs.sapien.io/collateral/sapien-vault

Proof of Quality · Sapien Docs

Proof of Quality (PoQ) - Sapien's open protocol for verifiable quality signals on AI data. Live on Base Mainnet.

Join us in 30 as we’re joining Fhenix on an expert panel to chat about the risks of expanding AI into security without proper oversight.

Watch below!

https://x.com/fhenix/status/2062445991471579252?s=46&t=2Cv-MWG_JgKjddpNZrIGHA

AI is everywhere- your apps, workplace and wallet

Join us June 4, 3PM UTC as we explore the growing risks behind AI expansion, the future of privacy-preserving systems, and what it will take to build trustworthy AI at scale

Hosted by @jack_gk

Featuring @BuildOnSapien &

X

In our most recent Sapien Digest, we went into detail about the role of the community in the days to come leading up to & after our launch.

Give it a read: https://newsletter.sapien.io/p/sapien-digest-may-29th

Sapien Digest - May 29th

New ways to participate in Proof of Quality.

In today's blog post, we explore how AI security tools changed the economics of security auditing.

Candidate vulnerabilities are now cheaper and faster to generate, but we have no way of proving outright whether a finding is actually to be believed. False positives consume senior auditor time, delay remediation, confuse clients, and weaken trust in the final report.

The next phase of AI-assisted security work needs a stronger validation layer: qualified expert review, clear rubrics, reviewer agreement, severity calibration, and, most importantly, a provable record behind every decision.

Sapien’s Proof of Quality turns AI-generated findings into verified security signals, helping teams separate confirmed risk from noise at the speed AI now demands.

Read more: https://www.sapien.io/blog/false-positives-are-the-hidden-cost-of-ai-security-tools

Sapien is heading to Consensus Miami to talk about the verification layer AI systems need before they make decisions with real consequences.

Find us at House of AI, Miami Marriott Biscayne Bay:

May 4, 4:20 PM ET
Lukas Grapentine
“Your agent wired $10m to the wrong address. Now what?”

May 4, 5:50 PM ET
Trevor Koverko in conversation with Jaden Lee
“Flawed In, Billions Out: The Hidden Data Crisis Behind AI”

https://luma.com/d7e60v9j - would love to see you there!

House of AI (by Lagrange & 0G) | Consensus Miami · Luma

An afternoon of substantive, opinionated conversation on AI infrastructure, verifiable AI, and who controls the future of intelligence.
AI is making decisions,…

How can smart builders prepare themselves for the AI challenges to come?

Find out on April 8th at 12pm EST as we’re joining the lineup of experts below.

Set your reminders!

https://x.com/meteoraeco/status/2041224472011686399?s=46&t=2Cv-MWG_JgKjddpNZrIGHA

AI Is Replacing Humans?! Here’s What Smart Builders Are Doing Instead 👀

Wednesday 8th April, 12PM EST

Join our upcoming spaces, hear live from these teams:
@BuildOnSapien
@edeldotfinance
@UnityAcademy
@sumex_labs
@BoostyLabs

Spaces link 👇

X

The AMA with CryptoRand will start at 12pm Eastern Time, so about 85 minutes out from now. Make sure to join the server early so you can go through verification without rushing!

Note: The AMA will be taking place exclusively on the CryptoRand server. They will not message you privately for any reason. Handle any private message or friend requests with the necessary amount of care!

Join below!

https://x.com/cryptorand/status/2033936401515843991?s=46&t=2Cv-MWG_JgKjddpNZrIGHA

⚡️ Fresh new AMA with @BuildOnSapien

📅 Thursday 19 March at 4PM UTC

💰 Rewards TBA

📍 Join here: https://t.co/N3WuQcWWkm

X