Tag
A long-time AI user reports that models across ChatGPT, Grok, Claude, DeepSeek and coding agents like OpenCode have noticeably worsened at task understanding and instruction-following over the past year, prompting questions about post-training, alignment and inference changes. The post asks the community whether others are seeing the same cross-provider regression.
European residential real estate operator Heimstaden is adopting OpenAI's tools to transform its letting and property management processes, cutting virtual staging costs by over 99% and targeting significant annual savings.
Oreja de Van Gogh 是一款 Product Hunt 上的新产品，可在 ChatGPT 说话时自动暂停 Apple Music，实现语音助手与音乐播放的无缝切换。
Guillermo Rauch shares enthusiasm for Vercel's MCP update, which surfaces deployments as interactive cards in ChatGPT and other MCP Apps-enabled clients with live status, build logs, and a visit button.
OpenAI announced a new internal benchmark evaluating how well its AI handles everyday human tasks like shopping and meal planning, reporting that GPT-6 Astra scored 93.1% on finding a kindergarten after a move and 90.4% on creating a diabetes-friendly meal plan.
OpenAI announced LifeVal, a new benchmark built from feedback by thousands of ChatGPT users on everyday tasks they most want help with, and has used it over the past year to measure and improve model performance on practical real-world workflows.
A user reflects that you don't need unlimited tokens to do great work, praising $200 AI subscription plans and arguing this is the most expensive intelligence will ever be.
OpenAI is merging ChatGPT's Chat and Work tabs into a unified experience, initially for select users, allowing seamless transitions from ideation to execution while Codex remains separate and both share a token allowance pool.
ChatGPT and Codex will share a usage limit, and users may need to wait for it to reset before they can continue chatting.
OpenAI announced LifeVal, an evaluation built to measure how well its models perform on practical everyday tasks—such as childcare, budgeting, shopping, and government services—derived from time-use studies and a new 3,995-respondent survey, and used over the past year to drive model improvements.
An exploratory case study tests ChatGPT with 110 surveillance-related prompts (55 jailbreak, 55 direct), finding substantive surveillance guidance for 40 jailbreak and 33 direct prompts, and offering a taxonomy of seven surveillance risk categories.
OpenAI is capping the number of paid ChatGPT accounts allowed per device and payment card, steering users toward more expensive credit purchases and usage resets instead of multiple subscriptions.
Parameter Labs disclosed a complex attack chain against ChatGPT's HTML preview artifacts that allowed sandbox escape, JavaScript execution, and exfiltration of private user data via crafted share URLs. OpenAI patched the flaw and paid a $5,000 bug bounty, which was criticized as too low for such a severe vulnerability.
OpenAI announced that starting in December, usage allowances for Plus, Pro, and Business plans will be shared across ChatGPT and Codex, measured in tokens, as it integrates Work's agentic capabilities into ChatGPT.
OpenAI 将于12月合并 Chat 与 Work，仅 Instant 查询不限量，其余请求都将计入 Codex 用量限额。这一变动将对用户的订阅用量和使用方式产生广泛影响。
Parameter Labs researchers disclosed a cross-frame message passing flaw that chained ChatGPT's HTML Preview sandbox with MathLive's unvalidated postMessage listener, letting a shared conversation run attacker JavaScript in a victim's signed-in session and take over their account. OpenAI fixed the P1 issue on October 26, and no user action beyond opening the link was required.
OpenAI 推出 ChatGPT 多账号切换功能，用户可在移动端和网页端同时登录最多两个账号并快速切换，各自保留独立的对话、记忆、文件和订阅设置。
An InfoQ report on OpenAI's Thomas Sarlandie's talk at AGNTCon + MCPCon details OpenAI's MCP server review rules requiring tool annotations (readOnlyHint, destructiveHint, openWorldHint), while emphasizing that correct annotations are only untrusted hints and do not replace server-side authorization and security.
OpenAI 将限制 ChatGPT 向青少年提供武器和安防系统等实用细节，并在用户表现出暴力执念时自动结束对话，同时加强了年龄预测机制。OpenAI 未公布其年龄估计和暴力执念检测的准确率数据。
Live Activities for Codex Remote are now available in the ChatGPT iOS app, letting users track remote agent tasks from the lock screen or Dynamic Island.