Quoting Matthew Green

Cryptographer Matthew Green, writing on Twitter and quoted by Simon Willison, offered two probability estimates about the future of public-key cryptography: a 1% chance that we live in "Minicrypt," and a 15% chance that we functionally lose confidence in the public-key encryption algorithms we rely on today.

Green framed the remark as a deliberate willingness to raise worst-case possibilities at a time when, in his view, everyone is focused on being respectable. His concern is not only the surprises themselves but the mismatch in pace: AI produces surprises quickly, while human institutions replace cryptographic standards far more slowly, even with strong AI assistance. Preparation done in advance, he argues, is what makes recovery from such a surprise possible.

Willison added context on the term: Minicrypt is Russell Impagliazzo's hypothetical world in which public-key encryption is impossible. The tag list accompanying the quote points to AI security research, cryptography, standards, LLMs and generative AI as the surrounding themes.

Why it matters: public-key encryption underpins TLS, signatures, messaging and most secure software, so a loss of confidence in existing algorithms would be an infrastructure-level event for developers and product teams. Green's numbers are rough personal estimates, not measured results, but they are a reminder that cryptographic migration is slow and mostly won in advance. Inference: teams that already inventory where public-key crypto is used and can swap algorithms behind stable interfaces will absorb such a shock far better than those that cannot.