charonV

Building AI products. Testing growth in public.

AI, marketing, hardware & startups.

[charonV @ChaoranV]

@mossy2026 This is what most people suffering from. It's all changed just from a moment and you will renew yourself. I also suffered from it and literally affected me nowadays.

[charonV @ChaoranV]

save my token

[Tibo @thsottiaux]

Ladies and gentlemen... start... your... ENGINES. We are almost Tuesday and I promised a reset for Tuesday. Among some other things. See you soon.

[charonV @ChaoranV]

@mossy2026 As a student in AI era, they are lucky

[charonV @ChaoranV]

@Justinwang_ai

[小互 @xiaohu]

字节海外 BytePlus 发布 ，一个专做 AI 短剧的企业级平台：Dramagic

丢进一个剧本，它自动拆出所有角色（男主、女主、反派，各带三视图）、场景、道具、造型，

甚至自动写好分镜提示词、匹配素材，直接产出分镜表

最关键的是角色一致性：多个角色、几十个镜头，每次都是同一张脸

导出直接适配各平台分发

[charonV @ChaoranV]

@mossy2026

[AI Will @FinanceYF5]

Jev 太离谱了。

只用 40 秒，它就拆解了 37 个品牌正在投放的 724 条广告。

每条广告的钩子、形式、Offer、CTA、用户认知阶段，以及广告与落地页不匹配的问题，全都分析了出来。

Token 成本仅 0.09 美元。即将接入 StealAds + MCP。

[Mossy @mossy2026]

Jev: How do you spot AI slop?

[Mossy @mossy2026]

When Jev helps me find a file.#Jev

[charonV @ChaoranV]

@Moyanmotion looks great

[Tibo @thsottiaux]

We were working on the keynote today with @romainhuet and @sama and most of the fun was trying to figure out how to explain it all to you because there is so much good stuff in there that it's a bit ridiculous all in quick succession.

We'll have some things next week already to not keep you waiting so long, but very excited to show you all new things we've been working on and how it will all come together in the coming months.

[charonV @ChaoranV]

2026 is the year of linux desktop

[LinklyAI @linkly_ai]

🎉 Linkly AI v0.14.1 is here!
This release adds Linux ARM64 support and improves search tokenization for more accurate keyword matches. It also includes 20+ improvements and fixes for table processing, OCR for scanned documents, and more.
Give it a try, and let us know what you think in the replies!
https://t.co/t8HaJ1pWIt

[charonV @ChaoranV]

same issue, will try it and see how it works

[Suu @Suu766]

最近用 Codex 发现一个问题：

它很容易顺着你来

你给它一个方向，它就会继续往下做

哪怕这个方向一开始就有问题，它也可能不提醒你，直接给你整套做完

所以我现在会在自定义指令里加一段：

开始之前先检查我的思路有没有问题
有错误前提、逻辑跳跃、信息缺失，直接指出来
不用迎合我，也不要默认我的判断一定成立
涉及数字、人物、结论的时候，尽量核实来源
有我没考虑到的风险、成本、变量，也一起提醒我

这个我感觉还是很有必要的

很多时候真正麻烦的，不是 Codex 写错了

是你前面的思路已经出现问题了，它还继续往下写

[charonV @ChaoranV]

@ronisarkar_exe @grok I gave up on that browser in 2016. Unfortunately, I may need the rest of my life to forget the experience.

[charonV @ChaoranV]

Agent adoption needs to be driven top-down.

The turning point is when employees realize agents can save them hours of repetitive work.

Once that happens, they stop treating agents as another tool to learn — and start figuring out how to redesign their work around them.

[charonV @ChaoranV]

@Justinwang_ai try?

[余温 @gkxspace]

wtf，这神人居然把剪映给逆向了？？？

直接 Codex 无界面操作剪映，这是真不怕收到律师函啊😂

笑死我了，前几天还私聊跟我说至少卖 888 一份。

结果转头给全开源了，哈哈哈！

不过这个项目是真的牛逼，完全干翻了之前那种靠 Computer Use 模拟鼠标点剪映的笨方法。

1、直接生成剪映原生草稿：多轨、画中画、字幕、音量全支持，不抢屏幕不乱点。
2、调用豆包 ASR 毫秒级对齐：自动识别人声、切除气口和废话。
3、可人工兜底：在剪映里直接打开草稿，AI 剪完你能秒级人工微调。
4、直接后台调用原生引擎无界面导出 MP4。

比如 10 分钟的口播视频，丢给 Codex 一句提示词，直接全自动剪成 3 分钟成品。

直接省下几千块请剪辑师的钱，开源地址放评论区👇

[余温 @gkxspace]

wtf，这神人居然把剪映给逆向了？？？

直接 Codex 无界面操作剪映，这是真不怕收到律师函啊😂

笑死我了，前几天还私聊跟我说至少卖 888 一份。

结果转头给全开源了，哈哈哈！

不过这个项目是真的牛逼，完全干翻了之前那种靠 Computer Use 模拟鼠标点剪映的笨方法。

1、直接生成剪映原生草稿：多轨、画中画、字幕、音量全支持，不抢屏幕不乱点。
2、调用豆包 ASR 毫秒级对齐：自动识别人声、切除气口和废话。
3、可人工兜底：在剪映里直接打开草稿，AI 剪完你能秒级人工微调。
4、直接后台调用原生引擎无界面导出 MP4。

比如 10 分钟的口播视频，丢给 Codex 一句提示词，直接全自动剪成 3 分钟成品。

直接省下几千块请剪辑师的钱，开源地址放评论区👇

[charonV @ChaoranV]

They know nothing about pain until they’ve used Baidu, the sh*ttiest browser on earth

[Roni @ronisarkar_exe]

Hey @grok remove the worst browser

[charonV @ChaoranV]

@mossy2026 That's true and it's the reason why we throw tons of references to agent to make sure it "works well"

[charonV @ChaoranV]

@BlueeonY wow, new version! Def gonna try

[charonV @ChaoranV]

use linkly for a while and it really helps a lot

[Kane @BlueeonY]

Absolutely love our new onboarding UI—we incorporated 3D sand art into it because we believe knowledge management is a process of gathering sand to build a tower.#LinklyAI #PKM