Weekly Update - June 22, 2026

Pricing goes public, the site finishes going multi-lingual, and the collaboration thesis gets outside validation

de Sam Rogers
9 min de leitura
update
pricing
i18n
governance
portfolio
security
Weekly Update - June 22, 2026

This was a transparency-and-integration week. Where the prior week was about hardening and measurement, this week was about making the product legible from the outside and wiring it into the rest of the portfolio. Full pricing went public on the blog with no "contact sales" gate, the multilingual site completion that finished internally the week before got its public announcement and graduated three languages from beta to fully supported, and a post showed end to end how a single behavior in a single assessment can be linked to a named legal obligation under a named statute by connecting PAICE to ObligationFirst. The week closed with a video on an external finding from OpenRouter — "fusion beats frontier" — that is the clearest outside evidence yet that the thing PAICE measures in people is becoming the thing that decides outcomes between models. Under the hood, a pre-push malware IOC scan hook shipped to harden the supply chain, and two locale-fallback fixes merged so translated pages stop briefly flashing an English banner.

Content Published Last Week

Monday (June 15): "Weekly Update - June 15, 2026"

Tuesday (June 16): "What PAICE Costs: A Tier-by-Tier Breakdown" Full pricing, published openly. No "contact sales for pricing," no qualification call. The assessment is free and that is the product, not a funnel. The post walks every tier — free assessment, Pro Packs, and organizational Baseline engagements — and what each is built to do.

Wednesday (June 17): "PAICE.work Is Now Fully Available in Spanish, French, and Portuguese" The public announcement of full-site multilingual coverage. Every public page — Home, About, Baseline, FAQ (55 questions across 13 categories), Individual, Cohort, History, Accessibility — plus ~150 blog posts, now in Latin American Spanish, French, and Brazilian Portuguese. The three languages graduated from beta to fully supported.

Thursday (June 18): "From Behavior to Breach: Linking Assessments to Legal Obligations" The mechanic, end to end, for connecting a behavioral measurement layer to an agent-native obligation graph. PAICE measures what a professional does in a turn; ObligationFirst represents what the law requires of that actor. Where they meet, you can say which obligation under which statute a specific behavior implicates — a stronger claim than "this profession is regulated, therefore AI use is risky."

Friday (June 19): Video - "Collaboration Beats Capability" On OpenRouter's "fusion beats frontier" finding. PAICE rests on the claim that how well someone collaborates with AI predicts outcomes better than how capable either party is alone. OpenRouter showed the same law holds one layer down, between the models themselves. External evidence that collaboration effectiveness, not raw capability, is what decides outcomes — in people and now in machines.

Pricing, In Public

We publish pricing for the same reason we publish methodology: anyone considering PAICE for themselves, their team, or their organization should be able to decide with full information. The assessment is free with no account, no email gate, and no usage limit — not as a marketing calculation but because it is the product. Pro Packs and organizational Baseline engagements are priced on the page, in USD, with no qualification call between a buyer and the number. Publishing the number is a positioning choice: it says the value is in the instrument, not in the friction around acquiring it.

Portfolio Integration: Behavior Meets Obligation

This week's framework post is the first to wire PAICE directly into another component of the portfolio. PAICE produces behavioral evidence — did the professional catch the injected error, exercise independent judgment, recover when the workflow broke. ObligationFirst represents legal obligations in a structured, agent-native form: actor, action, condition, deadline, authority, exception, with provenance pointers back to PubLedge artifacts. Joined, they support a claim no AI assessment tool currently makes: that in a specific turn of a specific assessment, a professional took an action that would have breached a named duty under a named law. The post shows the mechanic so a compliance reader can follow the chain from observed conduct to statutory obligation without taking it on faith.

PAICE.work PBC is not a law firm and does not provide legal advice. The integration is an evidence-linking mechanic, not a legal opinion.

The Collaboration Thesis, Validated One Layer Down

The Friday video connects an external research finding to the founding premise of the product. OpenRouter's "fusion beats frontier" result shows that routing and combining models can beat any single frontier model — fusion beats frontier. That is the machine-to-machine version of what PAICE measures in people: a terse professional who catches an injected error scores higher than a fluent one who misses it, because collaboration effectiveness beats raw capability. The same law now appears to hold between models, which is why the two open specs the portfolio is building for machine-to-machine collaboration matter. When the field's own evidence starts confirming your founding claim, it is worth naming.

Technical Improvements

Supply-Chain Security

A pre-push malware IOC scan hook shipped this week, with a shareable installer so the same protection can be dropped into sibling repos. It scans staged content for known indicators of compromise before a push leaves the machine — a supply-chain guard that complements the dependency and secret scanning already in the quarterly audit cadence.

Western Hemisphere Languages Fully Supported

This week we put the finishing touches on the Spanish, French, and Portuguese functionalies, and exited Beta status. This included two pull requests (#33 and #34) merged to stop translated routes from briefly rendering an English-fallback banner before the localized content settled. Pulse and the Contact page picked up their Spanish, French, and Portuguese copy, and the UI string count was corrected from 812 to 809. Translated pages now load clean in-language with no English flash, and no longer display "beta" flags.

Documentation and Pricing Records

Product pricing, the URL inventory, the security review, handoffs, and the locale beta-exit ticket were all updated to match what shipped. The pricing published on the blog and the pricing recorded in content/product-offerings.md are now in sync. A project-management currency pass also reconciled the planning docs against the code — confirming which open items are genuinely still open, recovering the scoring-bench design notes and an earlier infrastructure decision into the decision log, and re-sequencing the roadmap into a dependency-ordered set of tiers with security and operations debt scheduled first.

Platform Stability

All three environments (production, staging, pilot) operating normally. Zero unplanned incidents. Automated daily health checks and weekly backup restoration drills continue running without issue.

The Week in Numbers

  • 5 posts published (1 weekly + pricing + i18n announcement + framework + video)
  • 14 commits to main (+6,680 / -716 across 187 file changes)
  • 2 pull requests merged (locale fallback banner on translated routes)
  • ~150 blog posts now translated into 3 languages (weekly updates excepted)
  • FAQ fully translated: 55 questions across 13 categories per language
  • 3 languages graduated from beta to fully supported (ES, FR, PT)
  • 809 UI strings per locale (count corrected from 812)
  • 1 pre-push malware IOC scan hook shipped with shareable installer
  • All pricing tiers published openly, no sales gate
  • 0 unplanned incidents across three environments

Why This Week Matters

Three of this week's moves are about removing friction between PAICE and the people who need to evaluate it. Publishing pricing means a buyer never has to talk to anyone to learn the number. Finishing the multilingual site means a Spanish-, French-, or Portuguese-speaking compliance director can evaluate the methodology, read the FAQ, and understand the engagement model without switching languages. Both are bets that the product is strong enough to be evaluated in the open, on its own terms, without a gatekeeper translating it.

The From Behavior to Breach post is a different kind of move: it is the first public demonstration that PAICE is not a standalone instrument but one layer in a portfolio that is starting to interlock. Behavioral evidence on one side, structured legal obligations on the other, and a defined mechanic joining them. That is the difference between "we measure AI collaboration" and "we can tell you which named obligation a specific behavior implicates." The second claim is the one regulated buyers actually need.

And the collaboration thesis getting external validation matters because it is the load-bearing assumption under everything else. If raw capability beat collaboration, the entire premise of measuring how people work with AI would be on shaky ground. OpenRouter's finding is independent evidence, from a different domain, that the assumption holds — and that the open specs the portfolio is building for machine-to-machine collaboration are aimed at the right target.

Thank You

Thank you to everyone who read the pricing post and took it for what it is: an invitation to evaluate PAICE with full information, no call required. And to everyone evaluating PAICE in Spanish, French, or Portuguese this week: the experience is now native end to end, from the first page through the assessment through the results. If anything reads awkwardly in your language, please tell us. While we are officially out of Beta status, locale quality is something we will keep tightening.

Get Involved:

📖 This Week's Posts:

📖 Other Recent Updates:

Curioso, mas sem muito tempo?

Faça o PAICE Pulse de 3 minutos — uma verificação rápida de confiança que mapeia como você enxerga sua própria postura de colaboração com IA. Sem necessidade de login.