Twitxy Saved

Mirza

Verified account This account is verified. Learn more

@b1rkh0ff

🚀 pinli.co
✨ decklepdf.com
✨ WeekBird (soon)
bosnadev.com Joined May 2008
1.6K Following 3.7K Followers 14K Posts

Latest posts 5

Mirza Verified account @b1rkh0ff
It's time to try new Google models - with early Black Friday deals offering 50% off
Vikas Kansal Verified account @vikaskansalHQ 3h
We’re kicking off our Black Friday deals today with early access to our best deals for current AI and storage subscribers: Get up to 50 percent off Google AI Plus and annual Pro plans starting today through 11/26, so you can try out our latest models, including Gemini 4 Hydrogen and 4 Neon, AND get more storage!

Note that your exact offer will depend on your current plan. Terms apply. Check it out here:

one.google.com/ai?utm_source=…
View quoted post on X
Mirza Verified account @b1rkh0ff
🚨🚨🚨 BREAKING: An Anthropic mathematician claims he and Claude may have just proved the 167-year-old Riemann Hypothesis - the most famous unsolved problem in mathematics ($1M Millennium Prize).

Here are the wildest details from today’s release by Levent Alpöge:
• He claims it also resolves the Generalized Riemann Hypothesis.
• The informal AI-generated write-up ran past 250,000 pages, so instead of publishing “slop,” he dropped a 14-part Lean 4 formal proof.
• He admits he hasn’t even mastered all the surrounding math yet and expects much shorter proofs later.

Important caveat: It is not verified yet. An independent Lean checker is running the build right now, and mathematicians still have to confirm the encoded statement actually matches the real hypothesis.

If this holds up, November 20, 2026 just became one of the biggest days in scientific history. If it breaks, it’ll be the most famous AI near-miss ever.

Which way are you betting - valid proof, or a subtle formalization bug?