Hi , i am building Transparent and Controllable AI reasoning systems. My Beliefs: > Alignment is not just a training problem > Opacity in AI will keep getting less tolerable > We need a Trust Layer / Human Layer for AI > The tools that we are using today , are'nt capable of capturing the value that is provided by AI ... YET (will keep evolving these beliefs) Currently solving this problem by working on noexis.tech

What 😭

Btw codex weekly reset is not banked ig

Everyone on 100$ and 200$ plans of codex. Isn't it better now to just refer a friend(buy a cheap plan yourself) For a rate limits reset 4 times a month. Than buying another 100/200 $ subscription.

Yoo what, this is crazy ngl

Bruh

It’s even more than that. Even if I send the exact same prompt at the exact same time and branch into 2 different timelines where everything was same until I sent the prompt. External effects like other users can change the hidden path: routing, queueing, batching, GPU scheduling, sampling state, etc. From my side, nothing changed. From the system’s side, the computation wasn’t identical.

Gonna give this a shot and see what the hype is about

Tmux ad-wall

Yes 😭 @sama

🥀end is near

Why is the research and science community so closed ? Shouldn't there be something with the rigour of top journals and conferences. More open and especially free for people ?

I have been waiting for this for a while now

There exists an X - Linkedin information arbitrage

Mech interp is really cool tbh

The only way towards trust is ... Transparency and auditability. Currently solving this at Noexis

Also the monthly puzzles They're really good

Extremely bearish on this. I don't believe text boxes is the future. Rather we are moving towards a lot of innovation in the interaction layer of humans and AI , where the end goal might be BCI's

Wasn't the , keep your computer locked and still use your codex thing available since a week ago ?

To "Boil the ocean" or not ? I was watching youtube.com/watch?v=rQtrzBcf… this video by @daltonc and @mwseibel and within the first 20% of the video i realised ... is'nt this the opposite of "boil the ocean" advice by @garrytan recently. So i went back , read the article again, and completed the video. And somehow both seem right , on one hand garry challenges to let go of the previous beliefs and try to build something which is 10x. I did'nt feel it completely when i first read that post , but when i saw myself doing a whole month's worth of work in my previous job compressed into a day (and this was a few months ago) But on the other hand , Dalton talks about how he faced feature creep within two weeks and to think like an editor. Which is something i felt a few days ago , when the idea of cutting features off, thinking more in terms of UX rather than building everything that comes to mind. So what's the strategy ? Because we can't lean completely on one side , or maybe these are'nt even opposing sides. But maybe the optimal level of balance between these has shifted. I honestly don't know what shoud be the way , but i have a hunch here. It might be how "Intentional" something feels. The ease of understanding of users and clarity of what the product is, has to play a role in this. It should be engineered for the right experience, as engineering was always about good design and every feature feeling intended and not some slop.

we just kicked off s4 last night. -300 online. -20 in-person. founders booking their flights to seoul. 4 weeks to build! welcome to korea everyone 🇰🇷 ty to @arlanr for the q&a session!