Steve Weis
@sweis
Followers
10K
Following
5
Media
675
Statuses
8K
Software engineer at @anthropicai interested in cryptography, security, privacy, and machine learning.
San Francisco
Joined July 2007
Bay Area Crypto Day at Berkeley @SimonsInstitute tomorrow:
sites.google.com
Location: Simons Institute/Melvin Calvin Lab, 121 Calvin Lab #2190, U.C. Berkeley Please register here
0
0
3
We believe this is the first documented case of a large-scale AI cyberattack executed without substantial human intervention. It has significant implications for cybersecurity in the age of AI agents. Read more:
anthropic.com
A report describing an a highly sophisticated AI-led cyberattack
338
3K
12K
darpa: do u even lift bro
The DARPA Lift Challenge is offering $6.5M in prizes to innovators across America who design drones that can carry payloads 4x their weight, unlocking new possibilities for military & civilian applications. Registration begins January 2026.
0
0
4
"Parfait: Private aggregation & retrieval, federated, analytics, inference, & training" Google's acronym game is strong. https://t.co/sEZZLZXGvJ
github.com
Private aggregation & retrieval, federated, analytics, inference, & training from Google. - Parfait
0
1
2
"OpenPCC is an open-source framework for provably private AI inference, inspired by Apple’s Private Cloud Compute":
github.com
An open-source framework for verifiably private AI inference - openpcc/openpcc
0
5
10
"Claude Code Can Debug Low-level Cryptography" by @FiloSottile
https://t.co/KtojnyIWFC
0
11
28
More interposer fun, this time with DDR5 memory. Breaking TDX, SGX, SEV and even Nvidia TEEs. Checkout our work at https://t.co/Jl1dpGnM6J, and get a personally-signed Intel attestation report at @TEEdotFail.
45
84
343
Such a great conversation between @FryRsquared and John "Four" Flynn on Cybersecurity and AI based on his work at @GoogleDeepMind. Definitely worth a listen (and there is a part two as well!) https://t.co/SiGSP1oWXV
2
1
6
The CIA Kryptos sculpture solution key was sitting in the Smithsonian archives:
nytimes.com
Jim Sanborn is auctioning off the solution to Kryptos, the puzzle he sculpted for the intelligence agency’s headquarters. Two fans of the work then discovered the key.
0
1
6
Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.
1K
3K
20K
"Fully Anonymous Secret Sharing" slides at Crypto'25 by @TheAlliBishop , @matthew_d_green , Ishai, @paulslou, & Jain: https://t.co/zgsnZodLlm
0
4
10
“ Here [the authors] demonstrate an unconditional quantum advantage in information resources required for a computational task” https://t.co/khyT3tj39r
arxiv.org
A longstanding goal in quantum information science is to demonstrate quantum computations that cannot be feasibly reproduced on a classical computer. Such demonstrations mark major milestones:...
0
2
4
Welcome to the Quantum Benchmarking Initiative. To learn more about the effort and companies involved, visit:
Google Quantum AI is proud to be part of the @DARPA Quantum Benchmarking Initiative! This program will evaluate quantum computing approaches and help determine whether any can lead to a utility-scale, fault-tolerant computer by the year 2033. https://t.co/2GD8bTPat3
18
31
87
“Memory Integrity Enforcement: A complete vision for memory safety in Apple devices“ https://t.co/oBvY1ms1tq
security.apple.com
Memory Integrity Enforcement (MIE) is the culmination of an unprecedented design and engineering effort spanning half a decade that combines the unique strengths of Apple silicon hardware with our...
0
1
9