Michael Celia

Prediction Markets are hard. There is definitely foul play involved, yet statistically you can beat the market over time. @precisionalgo is launching a beta product soon that commoditizes ML models for free to help you beat these markets. Sign up today at precisionalgorithms.com #PredictionMarket

@vivoplt Cursor. Composer 2.5 is fast and cheap.

Which do you think provides best value for vibe coders? - $20 claude - $20 cursor - $20 codex

@pmddomingos Pass that shit broski

If it needs a harness it’s not AGI.

@afrax777 Order of operations

@sama why can we still not use 5.6 in codex when connected to a remote codebase...

@Tech2Wild You need a cpu that supports a lot of pcie lanes and a whole pc built around them.

So what is the CATCH with these ? You could grab these over 4 x 3090s at the SAME PRICE. And you get 32 GB a GPU vs 24 GB of VRAM. These seem like the better buy...

@jun_song None of these open source models are actually close in real world use. I think the lack of a really good harness is a reason why.

People actually think Open Source AI is months behind, while their models match the performance with only 1/3 to 1/4 the parameter size of US labs? - Kimi rivals Fable with just 2.8T, while Fable is estimated at 10T through reverse math. - GLM matches Opus with just 0.75T, while Opus is estimated at 3T through reverse math. Open your eyes.

I pay for cursor ultra. Working directly in the IDE with the agents in parallel. I don’t like being so far away from the code in the case of CC desktop or Codex Desktop. I know the CLI is great for those, and I use them, but for some reason I really like cursors harness. Grok 4.5 and composer 2.5 are amazing too. Something about how fast they run makes me much more engaged.

I don't know anyone on my feed who pays for Cursor or Perplexity anymore. How are these companies still around?

@UmbraAtrox_ @turingpi I downloaded 1TB of vram today wdym

@turingpi No amount of linux can conjure vram.

Cloud is a tax on not knowing Linux.

@Aracal5 @marcvidal Doesn’t count enterprise users probably

@marcvidal ¿Cómo es posible que Claude tenga tan pocos usuarios con respecto a los competidores y sea un modelo tan potente?

así va la cosa...

@md_kasif_uddin Def not kali lmao

Which Linux distribution is the best choice for first time users?

@jun_song Source?

Not many people know this : Jensen Huang predicted AI and started CUDA on 2006. Steve Jobs predicted AI after they launch Apple computer on 1983. These two geniuses predicted AI very early, and their companies produce best hardwares for personal inference now.

@leerob @cursor_ai Composer 3 plz

We just doubled the included usage of Cursor models on all plans. Enjoy more access to Grok 4.5 and Composer 2.5!

@verrsane @bu7emba Isn’t this kind of just marketing hype? 30% more tokens on 43k tokens is cents. An index beats hitting a couple APIs one at a time. I feel like this could be obvious without benchmarks.

@mcelia @bu7emba glean.com/blog/cowork-mc… heres the blog post, mb gang

7 figure renewal coming up Client did an internal eval of Glean MCP inside Claude vs Native Connectors for retrieval tasks 60% cost reduction in tokens with neck to neck completion RAG is dead, wrap it up

@verrsane @bu7emba Thank you!

@verrsane @bu7emba You need to sign up just to read the whitepaper?

@bu7emba interesting, how would a benchmarking tool work for comparing rag setups? we do have this x.com/jainarvind/sta…

@supabase Center a div

What's one thing AI still can't do and you have to by hand?

@verrsane I think even if you are an engineer?

If ur not an engineer 90% of your tasks can be done with a smaller cheaper model

@japan_nobunaga 🤣🤣🤣 felt the exact same thing my first time at Buc-ee’s

In Texas they told me to stop at Buc-ee's for gas. I have been to shrines. I have stood in temples that took two hundred years to build. I was not prepared for the gas station. There were one hundred and twenty fuel pumps. I counted them because I did not believe them. A man beside me was filling a truck the size of my first apartment, and he was not filling it because it was empty. He was filling it because he was here, and here is where a man fills things. Inside was a hall so vast I lost the horizon. A wall of jerky. A wall of fudge I did not know the country produced. A brisket sandwich handed to me by a man in a beaver costume. And I want to be clear, the beaver is not a mascot. The beaver is a saint. The people speak of him the way my grandmother spoke of the mountain behind her house. And the bathrooms. I had been warned about the bathrooms and I had dismissed the warning as the pride of a loud people. I was wrong to dismiss it. The bathrooms are famous across the whole state and they have earned it. I have slept in worse hotels. I nearly bowed upon entering. A janitor was polishing the floor with the devotion of a man tending a garden he loved, and when I thanked him he said "welcome in," which I have since learned is what Texas says instead of hello, and also instead of I am glad you exist. I went in for gas. I was inside for ninety minutes. I came out with fudge, a shirt printed with a joke I do not fully understand, forty dollars of jerky, and a feeling I can only describe as having been to church. I did not need any of it. I needed all of it. I have walked through the great cathedrals of the old world. I lit no candle there. I lit no candle at Buc-ee's either. But I did fill the truck. And I understand now that in Texas, this is the same thing.

@jun_song Is a DGX Spark worth it? Just say yes or no because if so I’m buying one right now.

Now Open weight AI models are only 2 months behind. And Gemini, Grok, Meta are at least 6 months behind.

@robj3d3 @instagram @Meta Happened to me and I got in contact with them through my business account and they still were no help. All AI at this point lol

Instagram just banned my account. No warning. No reason. No email. 8 years of posts. 400 days straight of posting. Gone. Does anyone have a contact at @instagram @Meta who can help?