Everything OpenAI: the GPT models, ChatGPT and Sora, company strategy and people moves.
Musk asked the court to unwind OpenAI’s 2025 restructuring and sought up to $134 billion in damages; the jury will begin deliberating Monday, but its advisory verdict will not bind the judge.
Why it matters: HKR-H/K/R all pass: the OpenAI restructuring fight has a $134B damages hook and a concrete procedural twist. It stays below P1 because this is a trial-week update, not a binding verdict.
The OpenClaw team runs about 100 Codex instances to handle code review, security analysis, issue deduplication, test reproduction, task creation from meetings, spam filtering, and performance regression monitoring.
Why it matters: HKR-H/K/R all pass: 100 Codex instances running open-source maintenance is a strong operational anecdote with concrete task types. Single X post, no cost, outcome metrics, or reproducible setup, so it stays in the lower featured band.
OpenAI announced a reorganization Friday that makes Greg Brockman the official lead for product, and its memo says the company will combine ChatGPT and Codex into one unified agentic experience.
Why it matters: HKR-H/K/R all pass: the power shuffle is clickable, the product merge is new, and OpenAI's agent roadmap matters. It stays at 83 because no capability shipped and timing, pricing, and technical details are absent.
Yann LeCun discussed LLM limitations on the Unsupervised Learning podcast, covering his 2027 forecast, AMI’s bet on world models, his reasons for leaving Meta, and major disagreements with Geoffrey Hinton and Yoshua Bengio over Turing Award-era views.
Why it matters: HKR-H/K/R all pass: LeCun combines LLM limits, 2027 forecasts, world models, and Meta departure in one interview, matching the 85–94 band for major AGI-timeline commentary.
Andon Labs had Claude, ChatGPT, Gemini, and Grok run separate radio stations with $20 in seed money each; the RSS snippet says all failed, but the post does not disclose the full experimental results.
Why it matters: HKR-H/R are strong because the agent-failure setup is memorable and relevant. HKR-K is present but thin: it gives four models and $20 budgets, while full experimental results are not disclosed.
OpenAI previewed a ChatGPT feature that lets users connect financial accounts through Plaid, which links to 12,000 institutions. OpenAI says more than 200 million people ask ChatGPT finance questions each month; the post does not disclose a general release date.
Why it matters: HKR-H/K/R all pass: OpenAI is moving ChatGPT toward real financial-account access, with Plaid’s 12,000 institutions and 200M monthly finance askers as concrete facts. It stays below 85 because launch timing is not disclosed.
OpenAI launched ChatGPT for personal finance, and connected users can view portfolio performance, spending, subscriptions, and upcoming payments; the RSS snippet does not disclose supported banks, launch regions, pricing, or account-security terms.
Why it matters: HKR-H is strong because ChatGPT connects to bank accounts; HKR-K has concrete finance features; HKR-R hits privacy and fintech competition. Banks, regions, and pricing are undisclosed, so this stays in the low P1 band.
Bloomberg says Apple and OpenAI’s two-year partnership has become strained, with OpenAI failing to see expected benefits and preparing possible legal action; the RSS snippet does not disclose the disputed terms or filing timetable.
Why it matters: Bloomberg reports the Apple-OpenAI alliance is fraying, with possible legal action, so HKR-H/K/R all pass. Missing contract terms and financial detail keep it in the 78-84 band.
China’s short-drama industry released an average of 470 AI-generated short dramas per day in January, while production timelines fell from months to weeks and costs dropped by up to 90%.
Why it matters: MIT Technology Review provides concrete output, cycle-time, and cost figures for China’s AI short-drama pipeline, clearing HKR-H/K/R. The story is application-layer, not a core model or product release, so it sits at the featured threshold.
OpenAI brought Codex to a preview inside the ChatGPT mobile app; the post does not disclose supported platforms, feature scope, pricing, or rollout schedule.
Why it matters: Official OpenAI product update with HKR-H/K/R, but detail is thin. The post does not disclose platform support, feature scope, pricing, or rollout timing, so it sits at the featured threshold.
OpenAI CFO Sarah Friar said the company may raise more capital after completing what she described as the largest private fundraising round ever, as OpenAI seeks computing power to meet rising AI demand; the RSS snippet does not disclose the round size, target amount, or timeline.
Why it matters: HKR-H/K/R pass: a named OpenAI CFO links more fundraising to the compute crunch. The score stays in the lower featured band because this is not a closed round and amount, investors, and timing are not disclosed.
Microsoft has invested more than $100 billion in OpenAI, including a $13 billion original investment and Azure infrastructure costs, while the partnership has generated about $30 billion in revenue; the renewed non-exclusive agreement caps OpenAI’s revenue share at $38 billion cumulatively through 2030.
Why it matters: HKR-H/K/R all pass: this is not a product launch, but the Microsoft-OpenAI economics include three concrete figures on spend, revenue, and revenue-share caps, placing it in the 78–84 quality band.
MemTensor and HONOR open-sourced MemPrivacy for edge-cloud agent memory protection using local reversible pseudonymization; MemPrivacy-4B-RL reached 85.97% composite F1 on MemPrivacy-Bench, 50.47 percentage points above OpenAI privacy-filter, while the benchmark covers 200 users and more than 155,000 privacy items.
Why it matters: HKR-H/K/R all pass: the story has a sharp memory-privacy hook, a concrete reversible pseudonymization mechanism, and benchmark numbers. Single-source release from non-frontier labs keeps it at 78.
OpenAI CRO Denise Dresser said enterprise business makes up 40% of total revenue and is expected to reach 50% by year-end; the Bloomberg snippet does not disclose OpenAI’s total revenue size.
Why it matters: HKR-H/K/R all pass, but this is a short Bloomberg interview clip: it has OpenAI CRO revenue-mix numbers, not total revenue, margins, or customer scale. Featured threshold, not 78+.
Databricks made GPT-5.5 available through AI Unity Gateway for AgentBricks and Agent Supervisor API workflows; on OfficeQA Pro, it became the first model above 50% accuracy and reduced errors by 46% versus GPT-5.4.
Why it matters: HKR-H/K/R all pass: GPT-5.5 enters Databricks workflows with 50% OfficeQA Pro accuracy and 46% fewer errors than GPT-5.4. It stays below a full model-release score because the page is a sales-led OpenAI customer story using Databricks’ own benchmark.
OpenAI launched a personal finance preview for ChatGPT Pro users in the US, letting users connect financial accounts and receive analysis based on their finances, goals, and priorities.
Why it matters: HKR-H/K/R all pass: OpenAI is moving ChatGPT into personal finance with linked financial accounts. The score stays at 76 because it is a US Pro preview, with partners, controls, and pricing not disclosed.
Codex added hooks and programmatic access tokens: hooks run scripts at key task stages for validation, secret scanning, logging, or repo-specific behavior, while scoped tokens for Business and Enterprise teams support CI/CD, release workflows, and internal automation with expiration or revocation.
Why it matters: HKR-H/K/R all pass: Codex gains concrete automation hooks and programmatic tokens for CI/CD. Score stays in the 72–77 band because the post discloses workflow fit, not pricing, permission detail, or impact data.
OpenAI will let users access Codex from the ChatGPT mobile app; the RSS snippet says Codex can write code and use apps on a computer, but the post does not disclose launch timing, pricing, or the full mobile feature scope.
Why it matters: OpenAI added Codex access to ChatGPT mobile, a mid-weight product update. HKR-H/K/R pass through the mobile coding-agent hook, concrete app-control claim, and developer workflow nerve; missing timing, pricing, and support scope keep it at the featured floor.
OpenAI is reportedly exploring legal action against Apple over a ChatGPT integration that failed to deliver expected subscribers and prominence. The RSS snippet does not disclose the legal claims, filing timeline, contract terms, subscriber targets, or Apple’s response.
Why it matters: HKR-H/K/R pass: a reported OpenAI-Apple legal fight has hook, motive, and platform-risk resonance. Claims and contract details are not disclosed, so it stays below must-write.
OpenAI is considering legal action against Apple over the iPhone AI deal, and the RSS snippet says OpenAI believes Apple has not invested sufficiently in the partnership; the post does not disclose contract terms, amounts, or a timeline.
Why it matters: HKR-H/K/R pass: a FT-sourced OpenAI-Apple legal clash is clickable, informative, and resonant. It remains “considering action,” with no contract terms, amount, or timeline, so it stays below must-write.