Moe

Anti Zionist

[Moe @SheikhSaeen]

OpenAI launched GPT-5.6 this week — and today it proved a 50-year-old math problem no one had solved before.

Here's what it can do, how it compares to Claude Fable 5, and where the cracks are showing. 🧵

#GPT56 #OpenAI #AI

[Moe @SheikhSaeen]

What is the Cycle Double Cover Conjecture, and why does it matter that an AI proved it?

The conjecture was posed in the 1970s and remained open for 50 years. Here's what it says, in plain English.

[Moe @SheikhSaeen]

@UncleJAI You're absolutely correct, it's a hard problem to solve, without the longer loops and retries, we might not get the quality of results that we desire, but then the cost also increases. The balancing act here might just be the art that decides who wins this race.

[Moe @SheikhSaeen]

OpenAI launched GPT-5.6 this week — and today it proved a 50-year-old math problem no one had solved before.

Here's what it can do, how it compares to Claude Fable 5, and where the cracks are showing. 🧵

#GPT56 #OpenAI #AI

[Moe @SheikhSaeen]

What this means: AI has crossed from retrieval into genuine mathematical discovery.

If a model can prove a 50-year conjecture by splitting work across sub-agents, the next target isn't another benchmark score. It's every unsolved problem we've been stuck on for decades.

[Moe @SheikhSaeen]

What is the Cycle Double Cover Conjecture, and why does it matter that an AI proved it?

The conjecture was posed in the 1970s and remained open for 50 years. Here's what it says, in plain English.

[Moe @SheikhSaeen]

Today's headline: Sol Ultra proved a math problem that was open for 50 years — the Cycle Double Cover Conjecture.

It split the work across 64 sub-agents working in parallel and finished in under an hour. This is actual mathematics, not a benchmark score.

[Moe @SheikhSaeen]

How it worked: Sol Ultra split the problem across 64 sub-agents, each tackling a different piece of the proof. They coordinated results and assembled a complete argument in under an hour.

This is the Ultra mode — not one model thinking harder, but dozens thinking together.

[Moe @SheikhSaeen]

The real innovation isn't just a smarter model — it's the sub-agent system. Splitting hard problems across dozens of workers and assembling results.

That's how it cracked the math proof. That's the playbook going forward. The model matters less. The orchestration matters more.

[Moe @SheikhSaeen]

Anthropic has passed OpenAI's valuation. Chinese open-weight models keep getting better, eroding the pricing floor. ChatGPT growth stalled last year, forcing a pivot toward coding tools.

The model is a real step forward. But the company behind it looks increasingly strained.

[🗻 @5amrenaissance]

Bear in mind.

The attempted extermination of muslims ON the European continent, happened in the lifetime of some of you reading this.

This did not happen 300 years ago,
It happened 31 years ago.

The Bosnian genocide is a lesson for us muslims that we cannot sleep nor delude ourselves about what is currently happening with the demonisation and now even the dehumanisation of muslims in the public, media and politics could only be the beginning of something much worse (I pray it won’t) but ‘could’ has very real stipulations.