Djadja

Success is not always about being the smartest or the networks you have now. Sometimes it’s just being patient enough to keep learning, disciplined enough to keep working, and attentive enough to keep improving.

This African dairy farmer from Rwanda 🇷🇼 went to Switzerland 🇨🇭 and was able to control cows that don't belong to him.

I want a hook that runs the "FAHHH" sound effect every time tests fail

Me voici dans le taxi, tout le monde est calme. La tantine stoppe 500 poste, elle entre devant elle met la ceinture. Le taximan éclate et lui demande si c’est la voiture de son sugar Daddy ici 😂😂😂

Tribalism. Someone from your group(tribe) will share the same values with you. This elminates the struggles from clashing views with someone from a different generational background.

Circus Charlie (1986)

Nobody has ever praised me for the good other men are doing.

@royesibe I don give you oil, just watch... Will come back in a few years to quote this.

Entrepreneurship is more about stamina than it is about genius

@Vinsmoke45 @BertinMbokish 😂😂😂

I just remembered, I'll beg people for Lahot in boarding school and they'll say no. I just made an ungodly amount of potato and egg. Fuck every single one of you who said no. I hope y'all see this tweet.

what a spiritual moment when you go out to eat alone at night in another country

We need to be serious about asking for details when people make claims like this and it starts to build virality. Sure, there is a chance they are doing what they claim, but there's almost no technical detail here about the capability of this model, how it was trained, and who did the training. > The site calls a Supabase backend and returns JSON all at once, not streamed. It does not transparently call its own model. > It returns reformatted model outputs with what looks like backend orchestration to search/tools and an LLM. > They're sending conversation history that makes it look more like app-layer than a model-serving API. At a minimum, they aren't exposing a model inference endpoint. They're sending messages to a backend that then feeds it into search/LLM elsewhere then spits a result back they modify and serve. They still could be forwarding this chat-wrapper to their own LLM but there's no evidence. It's much more likely this is either: 1. A wrapper/orchestrator on top of an existing LLM API with custom prompting. 2. A wrapper/orchestrator on a fine-tuned open model hosted elsewhere. The onus is on this team to clarify their claims and happy to update with technical details.

@BertinMbokish @Rileyfreeman420 Guyy 😂😂

@Rileyfreeman420 😂 No me oh. Her ass has to be jealous of the way shit comes out of her mouth.

Bertin Bertin😂🤣😂

The weirdest part of aging is realizing a year is not long

the essence of being a software engineer is simply not giving up when you're stuck

@BertinMbokish @_0xx_1 😂😂😂

@_0xx_1 How many times have you eaten today?

I've been at 65kg for 5 years 🤣 now

@BertinMbokish m4v files.. truly good times

I've organized thousands of songs. Good times.

@Vinsmoke45 😂😂 if na man for tok di ting, ye no go funny

😂😂😂😂😂😂😂😂

building a startup is a game of catch-22. you need funding to hire great people to build the product, but you need results to raise money. you need proof points to show prospective customers, but you need pilots to create proof points. you need to improve retention by investing in product, but the team is stretched thin just keeping the site up, so you need funds but to get funds you need retention and so many more catch 22s. Winning is figuring out how to break the paradox BREAK the Catch 22

“Then I told the horse, ‘cars won’t replace ya, but horses who drive cars will!’”