Tizzy

[Tizzy @Tian_yi_]

@jasteinerman My invite code is still valid. Feel free to use it if you need a new one:
86LEXC

[Tizzy @Tian_yi_]

@hellokittyold 我也来凑个热闹，大佬的用完了可以用用我的
86LEXC

[Tizzy @Tian_yi_]

who would like a muse invite!! 🎉

use my code
86LEXC
at https://t.co/SkAUy2gM4m — we both get 1B tokens 👀

[Tizzy @Tian_yi_]

Muse手机版一直显示wait list的可以试试电脑网页版登录，我登进去了，邮箱登录然后信用卡验证年龄就可以了，需要的话可以用我的邀请码，各自有10亿token~

邀请码：86LEXC
https://t.co/SkAUy2gM4m

[Tizzy @Tian_yi_]

@theadnaankhan 86LEXC

[Tizzy @Tian_yi_]

@0xsabo_ 86LEXC

[Tizzy @Tian_yi_]

who would like a muse invite!! 🎉

use my code
86LEXC
at https://t.co/SkAUy2gM4m — we both get 1B tokens 👀

[Tizzy @Tian_yi_]

@alexandr_wang If anyone needs it, feel free to use my invite code: 86LEXC — we'll both get 1B tokens!

[Tizzy @Tian_yi_]

Muse手机版一直显示wait list的可以试试电脑网页版登录，我登进去了，邮箱登录然后信用卡验证年龄就可以了，需要的话可以用我的邀请码，各自有10亿token~

邀请码：86LEXC
https://t.co/SkAUy2gM4m

[Tizzy @Tian_yi_]

@NFTLOVER7865 me🥺

[Tizzy @Tian_yi_]

@LuBtc888 @grok 验证真实性

[Tizzy @Tian_yi_]

As a software engineer, I’ve been using ChatGPT Work for the past week on actual projects.
Yesterday I gave it a messy requirements doc + some scattered notes from Slack, and it cleaned everything up, generated a proper PRD with acceptance criteria, created a basic task breakdown in Sheets, and even drafted the initial API spec. All while respecting our team’s conventions I’d shown it earlier.
It’s not magic, but it turns hours of boilerplate + context-switching into something I can review and ship faster. Really solid for getting momentum on features.
If you’re building stuff, worth checking out.

[Tizzy @Tian_yi_]

What I love about GPT-5.6 Sol is that it’s genuinely smart and remarkably persistent. It can keep working through long, complex tasks without constant prompting, babysitting, or even an explicit goal loop. It understands the direction, maintains context, and keeps pushing until the job is done.

[Tizzy @Tian_yi_]

The scary part of AI coding is not bad code.

It is low-value work becoming cheap enough to look productive.

More deploys are not the same as more product value.

The new bottleneck is judgment.

[ClaudeDevs @ClaudeDevs]

Boris sat down with Spotify VP of Engineering Niklas Gustavsson.

Spotify ships 4,500 production deploys a day, and 73% of PRs are now AI-assisted.

[Tizzy @Tian_yi_]

@huangyun_122 第二层级是agent辅助，人类分配任务给agent去完成，人类掌握主动权和主导权。
第三层级是agent主导，agent自我发现一些机会，反向给到人类提示，甚至直接做掉。但是在现在agent能力不足时，直接到第三层级可能发现的大部分是垃圾哈哈

[Tizzy @Tian_yi_]

@geekbb 已经花完了hh

[Tizzy @Tian_yi_]

@trq212 The tool is placed at the very top of the prompt. If we close and restart the session, newly installed tools will be loaded in—could this invalidate the entire prompt cache?

[Tizzy @Tian_yi_]

@songguoxiansen 中了glm5的1000w token哈哈

[Tizzy @Tian_yi_]

@songguoxiansen 赛博诸葛召唤中！你的好友Sl62EE7623271邀请你一起来猜模型上线瓜分千亿Tokens。前去参与 https://t.co/4Fheajxi4l

[Tizzy @Tian_yi_]

@lipeng0820 赛博诸葛召唤中！你的好友Sl62EE7623271邀请你一起来猜模型上线瓜分千亿Tokens。前去参与 https://t.co/4Fheajxi4l