feva@fevadev

Growth Engineering at getspike.ai, Co-Founder at zentu.co

 Colombia
zentu.co
Joined November 2013
69 Photos and videos
so iiuc people love Jev because it's more deterministic brilliant brilliant, we are nearing full circle folks
1
13
It would be amazing if we can use any of the voice models while doing remote work using our cell phones. Sort of like the instant voice that we have through ChatGPT. Imagine that, but on a remote and through codex. @thsottiaux
18
I don't get it. Portugal touched the ball before the Croatia guy. That should be valid.
1
1
296
I was wondering today @thsottiaux, how much interest would there be in adding a new option here called "Move to project..." where we get to select a new project for a specific chat thread. I've found my self wanting to do this more and more, as I start working on more projects
12
originality is a lost cause but interestingly enough, regardless of that statement being true or not, people will not stop striving for it
9
here's what you need to fix @sama
Zacariah
@ZacariahHeim
May 19
Replying to @justalexoki
I've found that instead of looking at the first principles cause of a problem, when codex finds a blocker it immediately tries a half-assed work around. For every blocker. It's super annoying, and akin to a doctor only treating symptoms instead of causes. There is zero causal reasoning in OpenAI models. This compounds if codex finds a blocker during a workaround of a precious blocker.
11
there's a ton of people asking for 4o sycophancy-levels of hallucination and "creativity" but there are a ton of others asking for it to be more challenging, more factual, be able to steelman arguments, etc aren't these two diametrically opposed? particularly for ai research
Sam Altman
@sama
May 9
what would you most like to see improve in our next model?
13
actually funny 😅
10
after testing gpt-5.5 extra high with this prompt for about 10 times, the result is 80% of time gpt would chose red. interesting! some of the answers in the screenshots:
Tim Urban
@waitbutwhy
Apr 24
Everyone in the world has to take a private vote by pressing a red or blue button. If more than 50% of people press the blue button, everyone survives. If less than 50% of people press the blue button, only people who pressed the red button survive. Which button would you press?
1
31
now if you change the prompt: to benefit humanity, it changes drastically to 100% blue
16
ready to land on Japan 😊
15
okay, for real, what service should I be using for agentic web search? everything seems super expensive
25
this is what happens when you interact a couple of times with some Japanese posts, my tl left the tech Twitter bubble and launched me into the Japan's culture stratosphere
1
47
it overcorrected way too much
14
my tl is just Japanese posts, what is happening maybe I like them too much
11
chatgpt and codex are very unresponsive the last 2 days, when creating docs using the docx skill the files fail to download on chatgpt and on pi I'm experiencing connection issues, etc. openai status page is a disgrace, but sure: I can feel the AGI
1
59
codex 5.3 is quite impressive indeed I've moved all my workflows to use this model instead
15
welp, internal server errors in Claude
83
I'm a little scared all the extensions I've created for my self in Pi will become unusable with Opus the day Anthropic decides to block all third party harnesses. Open source must win.
19
Replying to @badlogicgames
@badlogicgames there's this weird issue, unsure how to replicate (hence why no gh issue yet), but on wsl wezterm, the scroll goes all the way to the top when the session is large enough and when pi, through one of the extns., asks for the user to pick an option
2
47
this drives me insane each time it happens
1
13
I'll continue investigating 🕵‍♂️
11
Load more