Francesco Oddo | Rage Arc

New newsletter just came out, lemme know what do you think about it! It has all the latest most important AI News, and those two weeks where really packed. links.oddofrancesco.com/Newsletter-005

Bro i literally left home wtf c'mon

@sama Someone would call it loss leader, we call it just leader

come for the rate limits, stay for the best model

@DG_9_6 True, they are lobotomized but feel like geniuses. At least here we know we are all retarded

@oxfrancesco_ LinkedIn is missing yet the Twitter cringe. We need it

I love this. A guy asked me for info about a company. I replied… and then he told me to “stop messaging me.” Bro, you messaged me. You asked the questions. I was just answering 😭 And this happened on LinkedIn. Crazy times to be alive.

@CripdoeCrypto @jarrodwatts y not

@oxfrancesco_ @jarrodwatts why?

I now have: · $10,000 of Cursor credits · 10x Codex rate limits gotta use them both within the next month what do i build?

At this point I think there is nothing that GPT 5.5 cannot do if you have an IQ of at least 80 points...

@badlogicgames @zeeg Sorry I had to do it 😂

@oxfrancesco_ @zeeg oh you

is tomorrow the day i finally have an office again?

@badlogicgames @zeeg bc macos is better?

@zeeg no windows?

The new /GOAL command in Codex is how I think we will eventually reach AGI. @OpenAIDevs recently released the /goal command under a feature flag and with just a command it let's the codex harness go for HOURS (my longest one is 9hr and 42mins). How much is yours?

@ryanvogel Didn't know that homeless mans looked so good, sf is really another world

looks like they found the nearest homeless man and gave him a mic

@ryanvogel Every time people rage on Twitter about getting replaced, I'm always like this: Bro, imagine if a retarded model replaced ya, how good would you be at your job 😂

imagine getting fired from work and they replace you with an agent running llama-7b and it works BETTER than you

After a bunch of testing, I think that GPT 5.5 is the model that will need the most of the "codex" model variant. I'm finding it difficult to decide which thinking effort to use for different tasks. @thsottiaux wen GPT 5.5 Codex?

@jtbxl @ndrewpignanelli 😂

@oxfrancesco_ @ndrewpignanelli no but like actually

Announcing Cofounder 2: Run an entire company with agents. It's the infrastructure for the one person billion dollar company - orchestrating agents across engineering, sales, marketing, ops, and design. (and yes that's my real grandma in the video)

@ndrewpignanelli I think it was on the YC requests...

@oxfrancesco_ whos working on AI gfs

@Dimillian * Imagine walking around

imagine walking around with closed macbook

@krzyzanowskim Time to vibefork github, they have most if not all the primitives already I think @CloudflareDev

Cloudflare has a chance to do the funniest thing in history and ship Github replacement that limit scrapping repos

@jun_song Ye qwen no, and DeepSeek probably more R2 then v5

Ngl, which model release are you waiting for the most right now?

@saltyAom @gnukeith Bro people are so brain fried that will start to call even stuff like Bun slop

@gnukeith If Zed is a slop then I don’t know what’s not a slop

I once again ask, WHAT IS NOT SLOP TO YOU PEOPLE!?!?!?!?