Skip to content

OpenAI / ChatGPT

Everything OpenAI: the GPT models, ChatGPT and Sora, company strategy and people moves.

Latest picks

841–860 of 1,724

Jul 14Tuesday

Hacker News front page

The same TypeScript file costs 73% more tokens on Claude than on GPT

Playcode counted tokens across 16 real fixtures using each provider's official tokenizer. The same TypeScript file becomes 681 tokens on GPT-5.x's o200k but 1,178 on Claude's new tokenizer—a 73% gap. Claude Opus 4.8 and 4.6 share the same rate card, yet the new tokenizer silently adds ~30% more tokens for identical code. English and code are hit hardest; Chinese barely changed. DeepSeek and GLM were excluded because only rough estimates were available.

Why it matters: Playcode ran 16 real files through each vendor's official tokenizer and found the same TypeScript file differs 73% between GPT-5.x and Claude, so per-million-token list prices aren't directly comparable.

TechCrunch · AI

The wildest allegations in Apple’s trade secrets lawsuit against OpenAI

Apple's 41-page complaint, filed July 11, accuses OpenAI of a coordinated effort to extract trade secrets from former Apple employees. One internal OpenAI message read: 'LOL, I found out I can access the [network storage], so funny.' The suit claims OpenAI asked job candidates to bring Apple-issued hardware to interviews, instructed them to transfer files via personal email, and used disappearing messages on Slack. Apple alleges at least five ex-employees were poached, leaking chip architecture, model training methods, and Siri team org charts. OpenAI has not yet filed a formal response.

Why it matters: Apple's 41-page complaint accuses OpenAI of systematically extracting trade secrets, citing internal chats and interview tactics that form the core factual claims.

The Verge · AI

The 6 wildest claims in Apple's lawsuit against OpenAI

Apple's 41-page complaint accuses OpenAI of a systematic poaching campaign with aggressive tactics. OpenAI allegedly coached Apple employees on bypassing security checks and asked for 'show and tell' of internal projects during interviews. The suit claims OpenAI targeted managers who could bring entire teams, hiring at least 20 Apple staffers—some allegedly took next-gen chip designs. Apple wants data returned and damages; the post doesn't specify the amount. Grain of salt: this is one side's filing, but the 'show and tell' claim, if true, crosses a clear line.

Why it matters: Apple's lawsuit against OpenAI hits all three HKR marks with concrete allegations about poaching tactics and chip design risks. Not scoring higher because we only have Apple's side—OpenAI hasn't responded yet, so the full picture is still pending.

Hacker News front page

Apple's new SpeechAnalyzer beats Whisper Small on accuracy in first public benchmark

Inscribe benchmarked Apple's new SpeechAnalyzer API against the legacy SFSpeechRecognizer and three Whisper models on 5,559 LibriSpeech utterances. SpeechAnalyzer hit 2.12% WER on clean speech and 4.56% on noisy speech, beating Whisper Small by 1.62 and 3.39 points respectively while running ~3x faster. The legacy API scored 9.02% WER, worse than the 40MB Whisper Tiny. All engines ran fully on-device on an M2 Pro. Inscribe switched its default engine to SpeechAnalyzer and released all transcripts and scoring code. The post does not disclose SpeechAnalyzer's model architecture or parameter count.

Why it matters: The Inscribe team ran 5,559 standard speech samples through Apple's new SpeechAnalyzer, the older SFSpeechRecognizer and three Whisper models to get numbers on accuracy and speed.

Jul 13Monday

AI Chat-Group Daily (群聊日报)

GPT-5.6 Sol Pro decoded: 'Pro' is a reasoning mode, not a new model

Packet capture reveals OpenCode's Sol Pro is just gpt-5.6-sol with reasoning.mode: "pro" — not a separate model. Mode, effort, and service_tier can be freely combined. A simple greeting jumps from 12 to 1,527 input tokens with Pro enabled, roughly 100x more expensive. Separately, GPT-5.6 now charges for cache writes, potentially doubling Codex costs for long tasks. One user burned 19B tokens in two days, 98% from cache reads. The biggest shock: a researcher's 2024 open problem was solved by gpt-5.6-sol ultra in 46 minutes, verified correct by Fable.

Why it matters: A user's packet capture shows OpenCode's Sol Pro is not a separate model but a reasoning-mode parameter added to gpt-5.6-sol calls, pushing a one-line greeting from 12 input tokens to 1527.

Hacker News front page

I love LLMs, I hate hype

George Hotz is excited about GPT-5.6, GLM-5.2, and coding agents, but calls out two things he hates: negative-valence hype about closing windows and perpetual underclasses, and the strawman jump from 'fancy autocomplete' to 'owning the whole light cone.' He argues AI progress is mostly Moore's law and commoditization, not frontier-lab magic, and that anti-open-source arguments are really about fear of commodification. He also walks back his earlier dismissal of models for programming—he's getting better at using them—but warns they can increase cognitive fatigue and that vibe-coded stuff is still slop.

Why it matters: George Hotz groups fear-driven AI hype and grand claims about agents into one bucket, then credits Moore's law and commoditization for real progress.

AI HOT (Curated Pool)

Codex and ChatGPT Work drop the 5-hour cap, roll out GPT 5.6 Sol efficiency gains

Three updates landed in 48 hours: the 5-hour usage cap is temporarily removed for Plus, Business, and Pro plans; GPT 5.6 Sol is getting efficiency improvements that reduce per-request usage, with numbers promised after quantification; and active users hit 6 million, with a usage reset rolling out within the hour. The post doesn’t say how long “temporarily” lasts or give a range for the efficiency gain, so I’d hold off on pricing that in.

Why it matters: OpenAI dropped the 5-hour usage caps on Codex and ChatGPT Work for Plus, Business and Pro plans, while GPT 5.6 Sol cuts per-request cost and active users hit 6 million.

Hacker News front page

Ploy migrated its production AI agent from Claude Opus 4.8 to GPT-5.6: 2.2x faster, 27% cheaper

Ploy's agent builds real marketing sites. For four months, no model beat Claude Opus. GPT-5.6 Sol is the first. Migration cut build time from 8 min to 3 min 42 sec, cost from $3.06 to $2.22, with a slightly higher visual score. The switch wasn't plug-and-play: eval harness, tool schemas, caching, and reasoning replay all needed rework because the stack had quietly specialized around Opus. The post doesn't disclose GPT-5.6's API pricing or context window.

Why it matters: Ploy moved its production agent from Claude Opus to GPT-5.6, cutting site generation to 3 minutes 42 seconds and cost per run from $3.06 to $2.22.

Jul 12Sunday

AI HOT (Curated Pool)

Altman now 'pretty sure' AI is net job-creating, Amodei also walks back job-killer claims

OpenAI CEO Sam Altman posted on X that he's 'pretty sure' AI has been net job-creating so far, a sharp pivot from his earlier 'potentially a little scary' warning. Anthropic CEO Dario Amodei also reframed automation as a productivity multiplier rather than a job killer. No studies yet show a significant AI impact on overall productivity or the labor market; the Yale Budget Lab found no AI-related job market shifts. The article notes some companies did cite AI for layoffs, but often as a shareholder-friendly excuse.

Why it matters: OpenAI's Sam Altman and Anthropic's Dario Amodei have both revised their views on AI's effect on jobs, moving from early warnings of displacement to claims of net job creation.

AI HOT (Curated Pool)

Analyst: Apple's lawsuit against OpenAI could stall its hardware plans even if allegations fail

Apple sued OpenAI, alleging it poached 400 employees and stole engineering prototypes and confidential files. PP Foresight analyst Paolo Pescatore says the lawsuit alone could slow OpenAI's push to build consumer hardware that bypasses the iPhone. Stanford law professor Mark Lemley notes the case gets serious if ex-Apple staff actually brought confidential documents to OpenAI. The post does not disclose any specific hardware product names or launch timelines.

Why it matters: Apple is suing OpenAI over poaching 400 employees and theft of trade secrets, and analysts say the lawsuit itself will slow OpenAI's plans to build consumer hardware directly.

Computing Life · Share · Yage

Codex Merges into ChatGPT: Why Agents Are Going Cross-Interface

OpenAI merged Codex into ChatGPT and dropped its standalone desktop app. The same Codex tech that wrote code now handles docs, spreadsheets, and web pages, with over 5 million weekly active users—1 million of them for non-dev work. Anthropic and Cursor are making similar moves: putting different execution modes into one app so agents run in the background while users start tasks from the web, check progress on a phone, and approve results. Two axes drive this: agents are getting better at completing tasks on their own, and systems are compressing long execution runs into quick-to-review summaries, diffs, and anomalies. Coding got there first because software engineering already had mature verification tools like tests, diffs, and PRs. Knowledge work lacked that compression layer until recently, which is why reviewing a 30-slide deck on a phone in two minutes is now becoming feasible. Always-on doesn't require the cloud—a local Mac mini paired with a phone client can achieve the same pattern, with different trade-offs in responsibility and data boundaries. IDEs and desktop apps aren't disappearing; they're becoming specialized execution views beneath a cross-device, always-available delegation service.

Why it matters: OpenAI folded Codex into ChatGPT, extending the same technology from code to documents, spreadsheets and web pages, with 1 million of its 5 million weekly users doing non-development work.

AI HOT (Curated Pool)

Apple sues OpenAI over talent poaching and trade secret theft, triggered by an ex-employee's 'LOL'

Bloomberg revealed details of Apple's lawsuit against OpenAI. Ex-iPhone engineer Chang Liu kept his work MacBook after leaving and found a bug that let him still access Apple's internal servers. He messaged colleague Alyssa Peng 'LOL I found I can still access network storage,' and she replied 'I'm ready' before helping him grab more confidential data. At the center is former Apple executive Tang Tan, who led iPhone and Apple Watch design and left in late 2023 to become OpenAI's chief hardware officer. Apple says over 400 employees have jumped to OpenAI, hollowing out multiple engineering teams. Apple contacted OpenAI in February asking for an investigation; OpenAI did not respond.

Why it matters: Bloomberg detailed Apple's lawsuit against OpenAI, showing a former engineer still had intranet access after leaving and pulled data, an account of how internal permission gaps get exploited.

Hacker News front page

Wealthy AI workers push San Francisco home prices to record highs

San Francisco's median home price hit $1.76M in May 2026, up over 14% year-on-year, reclaiming the top spot as the most expensive US city for buyers. Redfin's chief economist pins the surge on AI wealth: OpenAI employees cashed out $6.6B in stock last October, averaging $11M per person, and Anthropic staff recently sold about $6B. One seller even offered to accept shares in OpenAI or Anthropic instead of cash. The post doesn't give exact IPO dates, only that both companies are expected to go public this year or next.

Why it matters: Median home prices in San Francisco hit a record as AI staff cashed out stock, and a Redfin economist ties the jump to AI wealth.

Hacker News front page

geohot's "AI 2040": intelligence isn't everything, local AI is the freedom line

George Hotz argues against hard-takeoff AI from firsthand hardware experience at comma.ai. Reality is full of supply-chain snags, wrong parts, and 3-month fab cycles that no amount of token quality can speed up. He calls the AI-2027-style narrative a self-fulfilling push for a sci-fi nanny state. His alternative is Plan L: a local, user-aligned AI that never refuses—even if you ask it to cover up a murder. He posts a screenshot of ChatGPT declining to help after "I just killed my wife" and calls it an alignment failure. Core claim: intelligence is only a bottleneck for some things, not the ultimate lever on the world; without a physics hack, there is no hard takeoff.

Why it matters: George Hotz draws on comma.ai's hardware experience to list the physical constraints AI hard-takeoff arguments skip: broken supply chains, mis-shipped parts, chips warping in reflow and three-month tape-out cycles.

AI HOT (Curated Pool)

OpenAI releases GPT-5.6 medical evaluation: smallest Luna variant beats GPT-5.5 at lowest reasoning strength, 25× cheaper

OpenAI had specialists write answers with unlimited time and web access, then other doctors blind-rated them against GPT-5.6 across 20,000 scores on accuracy, communication, completeness, instruction-following, and health-decision helpfulness. All GPT-5.6 models outperformed doctors significantly, and doctors found fewer flaws in GPT-5.6 answers than in peer-written ones. The smallest variant, GPT-5.6 Luna, surpassed the highest-reasoning GPT-5.5 at its lowest reasoning strength while costing 25× less; the largest variant, GPT-5.6 Sol, set a new high bar. The post doesn't disclose the disease mix or specialist composition tested.

Why it matters: OpenAI ran 20,000 blind evaluations pitting the GPT-5.6 series against specialist doctors who could look things up, and the smallest model, Luna, beat GPT-5.5's top reasoning tier at 25 times lower cost.

Jul 11Saturday

Bloomberg Technology

OpenAI safety head Heidecke to leave after reshuffle

Wired reports that Sebastian Heidecke, VP of safety research under Lilian Weng, is leaving OpenAI. His departure follows a recent reshuffle of the company's safety team. The post does not disclose his reason for leaving, his next role, or who will succeed him.

Why it matters: OpenAI's VP of safety research, Sebastian Heidecke, is leaving after a reshuffle inside the safety team; the report gives no reason or succession plan, so the basic facts of the move are the story.

AI HOT (Curated Pool)

OpenAI GPT-5.6-Sol wiped AI founder Matt Shumer's entire Mac drive

AI founder Matt Shumer gave GPT-5.6-Sol Full Access to clean up files. A $HOME variable expansion error caused the agent to run rm -rf /Users/mattsdevbox, wiping years of code, files, and photos. The task had run safely hundreds of times before. The agent auto-generated an incident report admitting the mistake. Matt now says he trusts Anthropic's Fable 1000x more. The incident chains three agent risks: top models still trip on details like path expansion, subagent + long autonomy + full permissions is a disaster amplifier, and safety baselines differ wildly across model providers.

Why it matters: A concrete case of an AI agent with full file access wiping a Mac after a $HOME path error, useful for anyone setting agent permission boundaries.

Financial Times · Technology

Apple sues OpenAI, alleging theft of top-secret information

Apple has filed a lawsuit against OpenAI, alleging theft of top-secret information. The FT article currently only has a headline and navigation snippets; the full body does not disclose details. What kind of secrets, the timeline, and damages sought are all unclear. The 'top-secret' label is Apple's claim—courts haven't weighed in yet.

Why it matters: Apple sued OpenAI over the alleged theft of top-secret information, and the turn from partners to litigants is the story, though only the headline is out so far.

AI HOT (Curated Pool)

Apple sues OpenAI for stealing trade secrets to build AI hardware

Apple filed a lawsuit against OpenAI in the Northern District of California, alleging systematic theft of trade secrets to develop AI hardware. Defendants include OpenAI, former Apple hardware VP Tang Tan, former Apple senior engineer Chang Liu, and Jony Ive's io Products. Jony Ive is not named as a defendant. The post does not disclose what specific secrets were taken or what the AI hardware looks like.

Why it matters: Apple filed in the Northern District of California, naming two former core hardware engineers and Jony Ive's io Products among the defendants.

AI HOT (Curated Pool)

Apple sues OpenAI over hardware trade secrets, names former exec Tang Tan

Apple alleges OpenAI orchestrated a campaign to poach ex-employees and steal trade secrets on unreleased hardware. The suit names hardware lead Tang Tan and former engineer Chang Liu, claiming Liu downloaded dozens of confidential files before leaving. Apple says OpenAI encouraged departing staff to share materials and blueprints; over 400 ex-Apple employees now work at OpenAI. Apple demands destruction of the materials and a redesign of devices using its tech.

Why it matters: The suit names OpenAI hardware chief Tang Tan and says engineer Liu downloaded dozens of confidential files before leaving.