i love open source and i hate these bullshit safety policies of anthropic and open ai

but whoever claim that open source models right now are at the SOTA level of intelligence of gpt and claude is just saying straight BULLSHIT
trust me

a benchmark doesn't mean anything because you can fine tune a model and this is exactly what this open source companies do in order to precisely make it score high on those specific benchmarks

these are completely false evaluations and when you then test them in the real usage you will see they are NOT as intelligent as claude or gpt

BEWARE there are scammers in DMs with the name “AI Mafia 🥷🏼”

block and report those fuckers if they DM you

btw guys let's just make the comments group a real group chat

if you are passionate or interested in AI, Business and Coding

& everything related

https://t.me/+KZmOrgB73MswYThk

join here so we can chat

https://x.com/paoloanzn/status/2027532197977329908?s=46

anthropic literally just said “FUCK OFF” to the USA gov

this is like saying: “i ain’t give a shit i’m not scared cause i know how powerful i am”

and the united states saying: “if we don’t control ai companies we are totally fucked”

idk about you - but i feel like things are about to go even worse in the coming years

they told us servers were actually real computers - just like desktops…

apparently we all have been lied to

https://x.com/paoloanzn/status/2028412796950712445?s=46

this automated job market intelligence system is now open - and you can get it for free

100% automated - there is tens of thousands of enriched data points inside

its been running full auto 24hrs - 7/7 for two months now - scraping data from upwork specifically targeting the ai automation industry

get access here:
https://gladium.ai/intel

all you actually need to know to understand about what does really mean "reaching AGI and singularity"

this is a NON-technical short article that will give you a general big picture framing about

• what even LLMs came from
• how they seem to be appeared overnight
• what is really the race to AGI that all of these big AI companies are competing to win
• where we are at right now
• what is the TRUE point of non-return that will change the history and future of humanity forever

(extracted from a late night discussion between me and chatgpt)

/ultraplan mode

custom experimental feature recovered from the re-compilation of the leaked claude code source code

ULTRAPLAN mode

https://x.com/paoloanzn/status/2039022418698907949

ENJOY

#FREETHECODE

claude code signs every API request with a cch= hash computed in compiled zig. recompiled clients just send zeros

nobody knows the signing algorithm yet. which means anthropic can tell which requests aren't from the legit client.

currently decompiling the official binary to reverse this - would be huge for all third party clients like opencode, openclaw etc to fully bypass anthropic enforcement and actually use the tokens you're already paying for

[right now using every client which is not the official version of claude code with your subscription token CAN get you banned cause anthropic can detect it]

https://x.com/paoloanzn/status/2039228672654385642?s=20

had to post cause i already saw people copy and pasting the above message as a tweet 😂

added full support for Codex

https://x.com/paoloanzn/status/2039348588741087341

we cracked it. the cch= signing system in claude code is fully reverse engineered - all credits for the work go to @ssslomp who did an amazing re work

now any opensource client can let users actually use the anthropic subscription they already paid for. with whatever tool they want

i have a new better version of free-code ready with

- full support for multi provider
- telemetry disabled
- cross provided conversation recovery
- automatic session exporting
- official claude code cli wrapping to use claude subscription plan 100% save

i’m just waiting for github to detach my repo from the parent fork. once they resolve the ticket i’ll push everything

thanks for the patience

🚨 URGENT

Someone created this fake page impersonating my company and launched a coin

please REPORT this profile to take it down immediately 🙏🏻

https://x.com/gladium_ai

I highly suggest you go and read this NOW

or you can continue believing all the bs they are feeding you about AI and AGI

https://paoloanzn.github.io/2026/04/26/agi-will-always-be-one-day-away.html

requires more attention span then the avg doomscroller

i’ve departed from twitter recently, if you follow me you know that

but i’m still writing things