This almost guarantees encryption is hacked.
Quantum tech, with crypto/ blockchain companies at the forefront, will be much bigger than crypto and finance in general. This will be backbone of the communications infrastructure...
Until it too gets hacked ๐
@BalaiBB If this is actually true then someone in China is getting executed. Would be easiest crime to solve in history. Assume he's already on the run but he will eventually be caught and the money will be returned. Some of it at least.
I got this from my research and for this I use Gemini, compare with $BTC, $ETH, $STRK, @ModulusZK , $CELL FRAME, $XX network,, & $ZEC maybe I am wrong but I got this result ๐ inspired by @drakefjustin 's thesis..... look at $CULT DAO friends
The current absence of evidence for an ECC break is not evidence that our existing security assumptions will survive mathematical superintelligence.
For decades, n log n was conjectured to be the limit for integer multiplication in the standard multitape model. That conjecture is now been disproven. A mathematical barrier that survived decades of scrutiny was not a barrier after all.
Just look at the results from @yukonresearch. Problems once considered exceptionally difficult are seeing remarkable progress and optimizations through collective agentic autoresearch.
I think it's becoming increasingly difficult to deny that collective AI mathematical superintelligence is emerging. Our confidence in cryptographic hardness assumptions made before 2026 should be updated accordingly.
I would not be surprised if AI discovers fundamentally new algorithms that reduce the effective security of ECC by orders of magnitude.
The apparent absence of major cryptanalytic breakthroughs from OpenAI's recent mathematical release is also interesting. Absence proves nothing, but given the strategic importance of cryptography, it raises questions about what research is being conducted privately.
In the 1940s, some of the most consequential discoveries in nuclear physics stopped being published openly. We are already seeing the beginnings of something similar in cryptanalysis, prime example being @GoogleResearch publishing major improvements to quantum resource estimates for breaking ECC while deliberately withholding the underlying circuits, in April of this year.
Nobody has publicly demonstrated a practical classical break of ECDSA. But waiting for one before seriously preparing alternatives seems like poor security engineering, particularly from someone at a company responsible for safeguarding billions in assets like @coinbase.
This mindset reminds me of the passengers on the Titanic listening to the orchestra after the iceberg struck, mistaking the absence of immediate catastrophe for evidence that everything was fine.
Being open to conservative, hash based cryptography should not be contentious.
The question is no longer whether AI will accelerate mathematical discovery. It is already happening at an extraordinary pace. Waiting for a demonstrated algorithmic breakthrough against the elliptic curve discrete logarithm problem before planning a migration seems unnecessarily risky.
We should be accelerating the transition toward hash based signatures such as XMSS and SPHINCS+ (and their variants leanSig & leanSPHINCS), and transparent proving systems such as STARKs.
I'd rather migrate a decade too early than a day too late.
Coinbase should be helping lead that transition, not waiting for the first demonstrated break to justify taking it seriously.