Every 分享 Astra 视觉特效实战与 AI 写作的 13 条心得

附加内容：一家创意工作室让Astra处理视觉特效，关于用AI写作的13个信念，以及一种新型模型的Vibe Check

本周三期内容的九个工作流程和提示。

“Vibe Check：Astra是独立电影制作人的突破吗？”：https://every.to/p/vibe-check-is-astra-a-breakthrough-for-indie-filmmakers 作者：Cyrus Duff：https://every.to/@cyrus.duff 和 Josh Lee：https://every.to/@josh.lee.afterimage / Vibe Check：https://every.to/vibe-check：Cyrus Duff：https://every.to/@cyrus.duff 和 Josh Lee：https://every.to/@josh.lee.afterimage，创意工作室Afterimage的联合创始人，测试了GPT-6 Astra在他们通常外包给专业人士的视觉特效工作上的表现——比如在桌子上添加一个酒杯。它擅长刚性物体，比如建筑物和玻璃器皿，并且可以独立使用专业工具，但它尚不能让生物或场景接触的任何物体看起来真实。对于一个两人工作室而言，这些工作原本是他们需要外包的。

🖥“迷你Vibe Check：TypeSafe的Jev在0.7秒内评判了我写的所有内容”：https://every.to/also-true-for-humans/mini-vibe-check-typesafe-s-jev-judged-everything-i-ve-written-in-0-7-seconds 作者：Mike Taylor：https://every.to/@mike_2114 / Also True for Humans：https://every.to/also-true-for-humans：TypeSafe的新模型Jev可以接收普通英文的“是/否”问题，并返回一个表示可能性数值，这样其他软件就无需人工审核即可使用它。Mike在不到一秒的时间内完成了对整个Every存档的处理，花费仅为几分钱。🖥观看：https://x.com/jackcheng/status/2100729670991802386 Jack Cheng：https://every.to/@jackcheng 通过说话和指向摄像头，在画布上绘制和移动形状，而Jev在不到半秒内执行每一个指令。

“关于AI写作的13个信念”：https://every.to/also-true-for-humans/ai-writing-beliefs 作者：Mike Taylor：https://every.to/@mike_2114 / 也适用于人类：https://every.to/also-true-for-humans : Mike 几乎所有发布的内容都是用AI起草的，他的第一条规则是让提示比输出更大。如果 ChatGPT 可以写，他就不会浪费时间。他仍然自己写推文和 Slack 消息，因为他的文章就是从那里开始的。我们的写作方式：https://every.to/guides/compound-writing 我们：https://every.to/p/what-writers-who-use-ai-want-you-to 立即写作：https://every.to/also-true-for-humans/ai-writing-beliefs 该系列以与 Mike 和 Katie Parrott 的现场共写活动结束：https://every.to/@katie.parrott12，时间是东部时间 9 月 23 日星期三下午 1 点，在 X 平台上。

“与 AI 玩耍教会了我关于工作的什么”：https://every.to/working-overtime/what-playing-with-ai-taught-me-about-my-work 作者：Katie Parrott：https://every.to/@katie.parrott12 / 加班工作：https://every.to/working-overtime : Katie 认为，你可以从一次实验中获取有用的东西，而不必把它变成下一个任务。她让 Astra 将她的复合写作技能：https://every.to/guides/compound-writing 变成了各种角色，而弄清每个角色想要什么让她将满足读者需求的问题加入她的写作工具箱。另外：获取“I s this anything?”，她在单次 AI 会议中寻找价值的技能。

“你可能忽视了电脑的使用”：https://every.to/context-window/you-re-probably-sleeping-on-computer-use 作者：Laura Entis：https://every.to/@laura_27bbaf_1 / 上下文窗口：https://every.to/context-window : Mike 在 9 月 3 日的 Vibe Check 中给 GPT-6 Astra 打了黄色评分。不到两周后，它成了他的日常工具，因为电脑使用帮他填写了学校表格，制作了六套课程演示，并检查了他书稿中的每一个链接。

“展示你的文件夹”：https://every.to/context-window/show-us-your-folders 作者：Katie Parrott：https://every.to/@katie.parrott12 / Context Window：https://every.to/context-window：Anthropic 将 Chat 和 Cowork 合并为一个 Claude，它可以自己为每个任务选择工具，这使得 Anthropic 和 OpenAI 走上了我们在四月预测的知识工作单一桌面应用之路：https://every.to/context-window/show-us-your-folders#signal。内容还包括：Dario Amodei 想要放慢前沿发展，而 Dan 提出了一个计划，使前沿 AI 广泛可用，从 AI 时代的防病毒软件入手，使普通人也能获取防御性代理。

“为什么你应该消耗更多代币”：https://every.to/context-window/why-you-should-burn-more-tokens 作者：Laura Entis：https://every.to/@laura_27bbaf_1 / Context Window：https://every.to/context-window：Dan 使用的代币数量是 Every 其他人三倍以上，他担心花费低意味着思考不够大胆。与 Willie Williams：https://every.to/@williewilliams 一起，他确定了一个起点：找到工程团队最大的限制，并测试 AI 是否能消除这些限制。

在《论文陈述》中阅读六个前沿建设者和思想者的更多预测：https://every.to/thesis-statements，这是有关人类伟大工作与 AI 未来的具体、可争议的主张合集。

请加入我们于2026年11月5日举办的首届论文：2027会议：https://every.to/thesis-2027

亲身体验 Every 如何使用 AI。在这些现场训练营、工作坊和聚会：https://every.to/events 中，团队成员将教授我们的工作流程。

本周内容到此为止。在 X 上关注 Every @every：https://twitter.com/every，在 LinkedIn 上关注：https://www.linkedin.com/company/everyinc/。

现在每个人都是建设者。Every All Access：https://every.to/builder-pack 为你提供完整会员资格加 Builder Pack——价值 9,000 美元以上的我们使用工具的积分。

登录即可免费阅读并获得你的 AI 评估提示。

Plus: A creative studio puts Astra to work on visual effects, 13 beliefs about writing with AI, and a Vibe Check of a new kind of model

Nine workflows and prompts from this week’s three editions.

“Vibe Check: Is Astra a Breakthrough for Indie Filmmakers?”：https://every.to/p/vibe-check-is-astra-a-breakthrough-for-indie-filmmakers by Cyrus Duff：https://every.to/@cyrus.duff and Josh Lee：https://every.to/@josh.lee.afterimage / Vibe Check：https://every.to/vibe-check : Cyrus Duff：https://every.to/@cyrus.duff and Josh Lee：https://every.to/@josh.lee.afterimage , cofounders of the creative studio Afterimage, tested GPT-6 Astra on visual effects work they would normally outsource to a specialist—adding a wineglass to a table. It’s good at rigid objects like buildings and glassware and at running professional tools on its own, but it can’t yet make living things or anything that touches the scene look convincing. For a two-person studio, that’s work they used to outsource.

🖥 “Mini-Vibe Check: TypeSafe’s Jev Judged Everything I’ve Written in 0.7 Seconds”：https://every.to/also-true-for-humans/mini-vibe-check-typesafe-s-jev-judged-everything-i-ve-written-in-0-7-seconds by Mike Taylor：https://every.to/@mike_2114 / Also True for Humans：https://every.to/also-true-for-humans : TypeSafe’s new model, Jev, takes a yes-or-no question in plain English and returns a number indicating the likelihood of yes, so other software can use it without human review. Mike ran his whole Every archive through it in less than a second for a fraction of a cent. 🖥 Watch：https://x.com/jackcheng/status/2100729670991802386 Jack Cheng：https://every.to/@jackcheng draw and move shapes on a canvas by talking and pointing at his webcam, as Jev executes each command in less than half a second.

“13 Beliefs About AI Writing”：https://every.to/also-true-for-humans/ai-writing-beliefs by Mike Taylor：https://every.to/@mike_2114 / Also True for Humans：https://every.to/also-true-for-humans : Mike drafts almost everything he publishes with AI, and his first rule is to make the prompt larger than the output. If ChatGPT could have written it, he doesn’t bother. He still writes his own tweets and Slack messages, because that’s where his essays start. Our How：https://every.to/guides/compound-writing We：https://every.to/p/what-writers-who-use-ai-want-you-to Write Now：https://every.to/also-true-for-humans/ai-writing-beliefs series closes with a live write-along with Mike and Katie Parrott：https://every.to/@katie.parrott12 on Wednesday, September 23, at 1 p.m. ET on X.

“What Playing With AI Taught Me About My Work”：https://every.to/working-overtime/what-playing-with-ai-taught-me-about-my-work by Katie Parrott：https://every.to/@katie.parrott12 / Working Overtime：https://every.to/working-overtime : Katie argues you can take something useful from an experiment without turning it into your next assignment. She had Astra turn her Compound Writing：https://every.to/guides/compound-writing skills into creatures, and working out what each character wanted led her to add questions about fulfilling readers’ needs to her writing toolkit. Plus : Get “Is this anything?,” her skill for finding value in a single AI session.

“You’re Probably Sleeping On Computer Use”：https://every.to/context-window/you-re-probably-sleeping-on-computer-use by Laura Entis：https://every.to/@laura_27bbaf_1 / Context Window：https://every.to/context-window : Mike rated GPT-6 Astra a yellow in the September 3 Vibe Check. Less than two weeks later, it’s his daily driver, because computer use filled out school forms, built six course decks, and checked every link in his book proofs.

“Show Us Your Folders”：https://every.to/context-window/show-us-your-folders by Katie Parrott：https://every.to/@katie.parrott12 / Context Window：https://every.to/context-window : Anthropic merged Chat and Cowork into one Claude that picks the tools for each task itself, which puts Anthropic and OpenAI on the path to the single desktop app for knowledge work we predicted in April：https://every.to/context-window/show-us-your-folders#signal . Also inside: Dario Amodei wants to slow the frontier, and Dan counters with a plan to make frontier AI widely available, starting with an antivirus for the AI age that makes defensive agents available to ordinary people.

“Why You Should Burn More Tokens”：https://every.to/context-window/why-you-should-burn-more-tokens by Laura Entis：https://every.to/@laura_27bbaf_1 / Context Window：https://every.to/context-window : Dan uses more than three times as many tokens as anyone else at Every and worries that low spend means thinking too small. With Willie Williams：https://every.to/@williewilliams , he settled on a starting point: Find the engineering team’s biggest constraints and test whether AI can remove them.

Read six more predictions from frontier builders and thinkers in Thesis Statements：https://every.to/thesis-statements , a collection of specific, contestable claims about the future of great human work with AI.

Join us at our inaugural Thesis: 2027：https://every.to/thesis-2027 conference on November 5, 2026.

Get hands-on with how Every uses AI. At these live camps, workshops, and meetups：https://every.to/events , team members teach our workflows.

That’s all for this week. Follow Every on X at @every：https://twitter.com/every and on LinkedIn：https://www.linkedin.com/company/everyinc/ .

Everyone’s a builder now. Every All Access：https://every.to/builder-pack gets you the full membership plus the Builder Pack—$9,000+ in credits for the tools we build with.

Sign in to read for free and get your AI assessment prompt.