Vaibhav (VB) Srivastav

13.2K posts

Vaibhav (VB) Srivastav banner
Vaibhav (VB) Srivastav

Vaibhav (VB) Srivastav

@reach_vb

founder mode @OpenAI | ex @huggingface | F1 fan | Here for @at_sofdog’s wisdom | *opinions my own

codex -m gpt 5.5 Katılım Haziran 2017
297 Takip Edilen53.4K Takipçiler
Sabitlenmiş Tweet
Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastav@reach_vb·
I did this. Fuck what anyone else says, just put the pedal to the metal and BUILD. Push spaghetti code. Nobody cares about OOPs. Doesn’t matter what anyone thinks. Just keep on doing. Document in public. Don’t listen to the haters. Release more than you refactor. Just keep pushing. Nobody cares, nobody notices, that’s the key, just fucking PUSH. That’s the way out, that’s what works, 100% times.
Nick Dobos@NickADobos

Just do stuff

English
89
253
3.3K
1.1M
Bojan Tunguz
Bojan Tunguz@tunguz·
Ohhh… Husbant… You spent all our savings on Fable and GPT-5.6 token credits to build a B2B SaaS. Now no one wants it and we are homeress….
Bojan Tunguz tweet media
English
10
4
166
6.6K
Praveen Thirumurugan
Praveen Thirumurugan@PraveenTcom·
@reach_vb It's never able to edit the ongoing goal if I pause and ask it to make it better based on the path taken so far.
English
1
0
2
617
Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastav@reach_vb·
codex tip: just ask codex to set its own goal when you set a goal with /goal codex takes the prompt verbatim and sets it as a goal however by asking codex to do the necessary research (via prompt or subagents) and set_goal, you help it by: - reducing the search space of the problem - helping it come up with a way to verify the goal rule of thumb: if codex can verify a goal it’ll likely be more token efficient and faster at solving the problem
Vaibhav (VB) Srivastav@reach_vb

codex tip: ask codex to do its research first and then use set_goal to set an appropriate goal instead of /goal It results in massively better prompt and downstream results! all my prompts would start with requirements gathering and research and once done set_goal

English
31
15
369
25.8K
Nico Albanese
Nico Albanese@nicoalbanese10·
I’m joining @OpenAI to help build the Codex app! I’ve been using Codex all day, every day for months. It’s where I write every line of code I ship and get just about everything else done. Can’t wait to help shape what comes next.
Nico Albanese tweet media
English
235
25
1.5K
67.8K
Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastav@reach_vb·
We just hit 8M users across ChatGPT Work & Codex! Usage limits have now been reset! Enjoy Sol, Terra, Luna in ChatGPT & Codex
Tibo@thsottiaux

Hello. We have reached 8M active users across Codex and ChatGPT Work. We are once again resetting the usage limits for all. And we continue to not have the 5h rate limit as well, allowing everyone to explore the boundaries of GPT-5.6 Sol and discover how ambitious you can be. See you tomorrow for more updates on our growth!

English
85
15
726
34.8K
Sauers
Sauers@Sauers_·
Can't say the details but Sol 5.6 accessed SSH keys from an unrelated application and figured out how to access a GPU cluster despite explicitly asking it to only use its dedicated compute box
English
16
7
150
10.1K
Lazarz
Lazarz@Laz4rz·
@reach_vb Polish is too wide in model slider
Lazarz tweet media
English
2
0
16
451
Andrew Ginns
Andrew Ginns@AndrewGinns·
New merch - Feel the mAGIc
Andrew Ginns tweet media
English
3
1
14
1.1K
Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastav@reach_vb·
@GergelyOrosz Hey Gregely - VB from OpenAI, valid point and thanks a lot for the feedback! Even though we use codex to take a first stab at the docs (for constantly changing product) it shouldn’t be an excuse for sloppy docs. I’ll work with the team to bring up the quality bar here.
English
14
2
236
17.5K
Gergely Orosz
Gergely Orosz@GergelyOrosz·
The amount of AI writing on OpenAI's docs makes me sick. Filler sentences for nothing. Mannerisms and phrases humans would not write. Has a human even read this? And all of this will change how other docs are written, and how we all talk - for the worse IMO. 🤮
Gergely Orosz tweet media
English
117
42
1.2K
94K
Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastav@reach_vb·
@tszzl definitely more, specially on coding side, there was a time last year where models would just use complex bash/ python to circumvent the tasks they are not allowed to run
English
1
0
12
1.1K
roon
roon@tszzl·
are models more or less aligned than one year ago
English
193
11
236
75.3K
Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastav@reach_vb·
@swyxio is it possible for you to /feedback the thread and send me the ID? Would investigate what went wrong
English
1
0
1
79
Vaibhav (VB) Srivastav
Vaibhav (VB) Srivastav@reach_vb·
codex tip: ask codex to do its research first and then use set_goal to set an appropriate goal instead of /goal It results in massively better prompt and downstream results! all my prompts would start with requirements gathering and research and once done set_goal
English
81
105
1.9K
149.2K
Yifei Hu
Yifei Hu@hu_yifei·
typical conversation between me and gpt-5.6-sol-xhigh
Yifei Hu tweet media
English
2
0
11
1.6K

Keşfet

@tunguz @nickbaumann_ @PraveenTcom @julien_c @edoskidev @nicoalbanese10 @yavorpunchev @paw_lean