Ola Yeenca

Absolutely. I’ll give you a quick insight: @smeanalytica we build AI-powered systems that analyze, interpret, and act on business data. For example: • MenuFlow (our ROS): analyzes restaurant sales, pricing, and customer behavior. • Real estate WhatsApp automation: collects and analyzes lead data to optimize responses and conversions. • Lead-generation & review intelligence: uses data to predict and improve ROI. But instead of building for big enterprises (they already have their own analytics teams), we focus on small and mid-sized businesses… giving them access to advanced AI technologies in a simple, practical way.

@st4ayh4rd @mattpocockuk 😂😂

@mattpocockuk code or you're gay.

COYG

@lukas_undefined @mattpocockuk 😂😂

@mattpocockuk Commit Often, Yell “Goblins”. Clearly.

@mattpocockuk How do you feel Matt 😂😂

@imohwiz Please send a dm to @smeanalytica

@Yeenccs Your DM is locked

need a motion designer. That can convey product story reel for an AI WhatsApp agent that handles real estate leads. DM rate and a reel.

Asistiré al próximo Espacio de @WeTalkBallHq. ¿Te unirás tú también? twitter.com/i/spaces/1dKrP…

@imohwiz Send me a dm

@Yeenccs I'm available vimeo.com/imohwiz

@VellySlimKing Why you wan do me like this 😭😂😂

Na you be Ola BTC ?@Yeenccs

@thsottiaux @JustinBleuel Tell me, how am I on the max plan 5x and I’m still running out of usage limit?

@JustinBleuel is just shipping things with codex and making chatgpt better. Worth following (and complaining to)

@mitchellh @yacineMTB The agent made huge progress with the iterative loop, no doubt. But even with more cycles, it stayed stuck in the original design. The 0.02ms version shows why expert human insight still matters. Cool experiment!

I've got an agent in a loop optimizing a renderer with the goal to minimize frame times (and tests to measure). It got times down from 88ms to 2ms and allocations down from ~150K to 500. Sounds good, right? Wrong. This is exactly why agent psychosis is a big fucking problem. As an experiment, I rewrote the Ghostty core render state in Go, with access to identically laid out data structures as Ghostty and the exact same validation tests. I made a purposely naive renderer (simple, correct, but slow). 88ms per frame with 150,000 allocations (horrendous, lol)! I then kickstarted a Ralph loop to bring the frame times down. I told it it can't modify input data structures or the public API or tests (they're correct), but it can do anything else it wants. It got to work. It has worked for about 4 hours. I've spent around $350 on this experiment so far. The results? 88ms => 1.5ms 150K allocs => ~500 allocs Incredible right? Nope. My hand-written renderer I ported has frame times (same benchmark) of ~20us (0.020ms) and 0 allocations in the update path. This is the problem with psychosis and lacking systems understanding. If you don't understand the system, you're going to accept that this is an incredible result. If you understand the system, you'll see better solutions immediately and can do roughly 75x better on throughput. The people who blindly trust agent output are in the former camp. They're sheeple, overdrinking from a fountain of mediocrity. Standard disclaimer: I use AI all the time. I like AI. The point I'm making is to not blindly accept results. Think. Analyze. Learn.

@Its_Mykehl Tunde Perry don reveal him identity 😂😂

Mistakenly disclosed I am Nigerian to this Don at work and my man cannot stop speaking Yoruba to me. 🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️🤦‍♂️

@atmoio Watching this in 2x so it’s an absolute crash out lol

I'm done. I'm f***ing done.

the north star, written on the wall: "it should be boring to operate." turning it on should never feel like a gamble.

i didn't ship a single feature this weekend. i spent it making the thing i already built solid enough that a client can switch it on, walk away, and trust it's still running monday.

@yannschaub Yeah, you’d be surprised how many security and bug fixes it would find, fix and iterate on… the goal should be concrete and sandbox for testing would help too!

@Yeenccs i see! I need to do this more often as well.

🤯 TIL in Codex you can use /side while the agent is working to open a sidechat that has all the context from the main chat. Very useful to have conversations with it and braindump without interrupting main agent

@yannschaub Continual test and fixes… hidden patterns and possible bugs. Not actually building

@Yeenccs wtf are you building that runs for 20h

@thsottiaux Love the way you noticed and tackled these issues first hand, really amazing

Some of you noticed limits drained faster in Codex, we root caused it to an optimization that we rolled back that had an impact on cache hit rates when compacting across long running sessions. We fixed this and have now reset usage limits for all accounts. Enjoy the weekend.

@yannschaub You can’t be good at all things, you gotta focus on one.

how is it possible that codex is so bad at UI? it's abysmal

@maazscript Ah, thanks…

@Yeenccs Replit Dribble Pinterest Behance Also research a lot about designs and flows using ai

But come to think of it, why do I put so much energy into building Apps compared to web apps? I have two web apps I'm working on, and whenever I'm working on them, I don’t feel much excitement. But for mobile development, I want to impress myself and everyone… Am I cooked?