Weekly calls
Every call made in The Durability Curve Weekly, worded exactly as printed, with the confidence we gave it and how it came out.
Track record: scored once 10 calls have resolved (3 so far). How we score
01 · The calls
True Argon for everyone 55% confidence Resolved
The call, as printedby 31 March 2027, any paying API customer will be able to use Argon, or a renamed Gemini 4 successor, without joining Fairwind. Google has already published a price. Against that, it is still tuning Argon's guardrails, and Anthropic's newest Mythos model is still invitation-only. Six months is a fair test of "as soon as possible".
Why it was graded trueGoogle's launch post of 13 Oct put Argon on the Gemini API paid tier at a listed price, with no Fairwind step (Fairwind now covers only Argon without cyber guardrails). Checked on 15 Oct against the post, the models page and the pricing page. No live key was tested.
True Carbon for paying customers (story 3) 70% confidence Resolved
The call, as printedby 31 December 2026, any paying Gemini API or Google Cloud customer will be able to use Gemini 4 Carbon without joining Fairwind or any other vetted programme.
Why it was graded trueGoogle made Gemini 4 Carbon generally available on 10 Nov, on the Gemini API paid tier and on Google Cloud, where its model page lists it as publicly available with no Fairwind step. Only a separate model, Gemini 4 Carbon for cyber, stays behind Fairwind. Checked on 10 Nov against Google's Cloud model page, the Gemini API's what's-new page and Google's launch post. No live key was tested.
True Anthropic says it met its line (story 3) 55% confidence Resolved
The call, as printedby 30 June 2027, Anthropic will say publicly that it has met its policy's threshold for "dramatic acceleration" of AI progress, or will designate one of its models as exceeding it. Saying it may meet the threshold soon won't count.
Why it was graded truePriced from a shown breakdown (paid desk): publishes where it stands ~90% × says met or names a model ~60% = 0.54 ≈ 55%. **Resolved:** Anthropic's IPO filing of 10 Nov (S-1/A No. 2) says it "made a determination that we have met this threshold": the AI R&D capability threshold in its Responsible Scaling Policy, the one that applies where it "observe[s] or expect[s]" the accelerated progress of scenario (2). It based the determination "on what we expect from future models and our overall rate of progress, rather than on what we have observed from current models". The 2 Nov filing had said "may imminently meet". Checked on 14 Nov against the filing on EDGAR, with the 2 Nov baseline and RSP v3.4 in the Vault's copies.
Open Disclosure deadline 30% confidence
The call, as printedby 31 December 2026, OpenAI's Australian taskforce will publish recommendations that set a maximum time, in hours or days, between a lab finding unauthorised access to a government system and telling that government. A word like "promptly" won't count, and if it publishes nothing by then, the call fails.
Open Supplier warrants 70% confidence
The call, as printedBetween now and 31 March 2027, at least five more US-listed suppliers will disclose a warrant like these, issued to a customer in connection with chip, cloud or data-centre purchases.
Open A new $1bn NVIDIA investment (story 1) 70% confidence
The call, as printedby 31 March 2027, NVIDIA, or a company it invests in, will announce a new NVIDIA investment of $1 billion or more in an AI developer or in a cloud or data-centre company. Deals agreed before 14 November, its Anthropic purchase included, won't count.
Open Memory deposits 65% confidence
The call, as printedby 31 March 2027, another US-listed memory or storage maker will report at least $5 billion of customer advances and refundable deposits on its balance sheet. Sandisk had about $2.7 billion in July.
Open Agent checkout 60% confidence
The call, as printedBy 31 March 2027, another of the ten largest US online retailers will announce a new block on an AI shopping agent, or new rules requiring agents to identify themselves. Rules already in place, such as eBay's ban on buy-for-me agents, don't count.
Open Outsiders inside OpenAI 55% confidence
The call, as printedby 31 March 2027, OpenAI will announce an agreement giving a named outside organisation ongoing access inside the company to evaluate its models while they are being trained. Talks won't count, and nor will tests of one model before its release, like those the UK and US institutes already run. Anthropic took six days. Four weeks on, OpenAI says it is finalising contracts. It has named no partner, and the details are still to come.
Open NVIDIA names a permanent chief (story 1) 50% confidence
The call, as printedby 31 March 2027, NVIDIA will name a permanent chief executive in a press release or SEC filing. Extending the interim role won't count.
Open FrontierCode 65% 45% confidence
The call, as printedby 31 March 2027, any listed configuration of a model will reach a headline Score of 65% or more on Artificial Analysis's FrontierCode 1.1 leaderboard, the 100-task Main set. The pass rate doesn't count, and nor do Cognition's own boards. A later version of the test counts only if Artificial Analysis carries the scores over.
Open Nuclear upgrades or restarts 40% confidence
The call, as printedby 31 March 2027, Microsoft or Meta will announce a new deal that pays for new output at an existing US nuclear plant, through an upgrade or a restart. Amazon and Google each announced one within a week of each other. Microsoft and Meta signed theirs earlier, and those don't count.
Open US reporting deadline 20% confidence
The call, as printedby 31 March 2027, the White House or a US federal agency will publish a time limit, in hours or days, for AI companies to disclose incidents involving their models. "Immediately" won't count, and nor will a general cyber-reporting rule that doesn't name AI models. A recommendation or a bill won't count either; it has to be a requirement the companies must meet. The Force has called disclosure not optional, named no number and set no penalty.
Open Pre-release access in writing 20% confidence
The call, as printedby 31 March 2027, the White House, a US agency, the UK government or the UK AI Security Institute will publish a new written agreement or joint statement confirming the institute's pre-release access to frontier models from OpenAI, Anthropic or Google DeepMind. Speeches, parliamentary answers and anything published before 17 October 2026 won't count, and a joint statement needs the institute or the UK government plus one of those labs or a US government body to sign it.
Open Utah steps doctors back 10% confidence
The call, as printedby 31 March 2027, Utah or Nolla Health will announce that Nolla's AI has moved past its first stage, sending prescriptions without a doctor checking each one beforehand. That takes at least 100 patients and four weeks, at least 95% agreement, a clean safety record and the state's written sign-off, which Doctronic's Utah pilot still lacks after almost a year.
Open IMF recount above $1.4 trillion 30% confidence
The call, as printedby 30 April 2027, the IMF will publish an updated count of these circular deals in AI, covering the same or a wider set, that puts future revenue commitments above $1.4 trillion. Any IMF publication counts, including the April 2027 Global Financial Stability Report, a blog, a working paper or a technical note. A figure of about $1.4 trillion doesn't count as above.
Open Agent disclosure 55% confidence
The call, as printedAustralia's AI standards bill, when it is published, will set a deadline for AI companies to report incidents in which their systems get into computer systems without permission. If no bill is published by 30 June 2027, the call fails.
Open Statewide data-centre pause in law (story 1) 55% confidence
The call, as printedby 30 June 2027, a US state will sign into law a pause or ban on building new large data centres across the whole state. Local measures and executive orders, such as New York's in July, won't count.
Open Watermarks 40% confidence
The call, as printedBy 30 June 2027, anyone will be able to check a piece of text for Claude's watermark.
Open Open-model gap 30% confidence
The call, as printedby 30 June 2027, NIST's AI evaluation centre (CAISI, now renamed CAISSI) will publish an assessment that puts a freely downloadable model within two months of the best US model on its overall cyber measure.
Open Public ad-safety rate 30% confidence
The call, as printedby 30 June 2027, DoubleVerify, Integral Ad Science or OpenAI will publish the share of live ChatGPT ads that DoubleVerify or Integral Ad Science measured as brand-safe or suitable, or as failing. Results from a test environment won't count. OpenAI says its pilots are in development, and Integral Ad Science's is a closed pilot that reports to the advertisers taking part.
Open Quantum re-count (story 3) 30% confidence
The call, as printedby 30 June 2027, a peer-reviewed paper by researchers outside Anthropic, Harvard and QuEra will report its own Toffoli and qubit counts for at least one of the eight tasks in the paper's Table 1, re-counting its circuit or comparing a new one against it. Preprints, Brown's included, won't count, and nor will anything published before 24 October 2026.
Open OpenAI monitor check (story 1) 30% confidence
The call, as printedby 30 June 2027, either OpenAI will publish a safety case, or a summary of one, setting the monitoring requirements for a named category of its internal workloads, or an organisation outside OpenAI will publish results of a new test of its agent monitoring, run after the fixes for the problems METR and Redwood reported. Promises, policy pages and anything published before 24 October 2026 won't count.
Open Anthropic's operating profit (story 1) 30% confidence
The call, as printedby 30 June 2027, Anthropic's first quarterly report after it lists, on Form 10-Q, will show an operating profit for that quarter. Its adjusted measure won't count.
Open FTC filing 20% confidence
The call, as printedby 30 June 2027, a public FTC or federal court filing arising from this investigation will name OpenAI or Anthropic.
Open Price floor 75% confidence
The call, as printedBy 30 September 2027, neither lab will raise the list price of Opus 5.5 or GPT-6 Sol, or release a new Opus or Sol model at a higher input or output price per token.
Open Reading preferences 35% confidence
The call, as printedBy 30 September 2027, an independent study where real money changes hands will find the same split, with misreading people's preferences costing more than the trading itself.
Open Biology watermark 25% confidence
The call, as printedby 30 September 2027, at least one commercial DNA synthesis company will announce that it accepts or prefers SynthID Bio, or an equivalent lab-tested watermark for AI-designed biology, in its screening of orders. Twist Bioscience, a DNA maker, called watermarking "a promising new addition to the biosecurity toolbox" on DeepMind's launch page, but a quote there doesn't count.
No calls here yet.
02 · How we score
For each resolved call, take the confidence we printed, subtract 1 if the call came true or 0 if it did not, and square the gap. The Brier score is the average of those squares. 0 is perfect, and always saying 50% scores 0.25. Lower is better. Voided calls are not scored.
A call is graded true or false by its deadline, or when the event it names happens. Misses are printed in the issue that grades them.
Calls made outside the Weekly are graded in the piece that made them and are not counted here. Your AI Grader Is Only as Good as Its Answer Key printed one at 35%, due by 30 September 2027.
The same record is at /calls.json, and Claude or ChatGPT can read it through the connector.
Calibration appears once 10 or more calls are graded. Graded so far: 3 of 28. Next scheduled resolve date: .