Tap a claim on the ring to put it at the centre.
← AI labs are extremely dependent on chain-of-thought monitoring, which…
17 connected korrents · 12 moments on record from 13 Jan 2017 to 12 Sept 2026. Nearly all of them are about cryptography .
Everything filed under OpenAI
OpenAI
Everything filed under privacy
privacy Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Same subject Read this korrent: AI labs are extremely dependent on chain-of-thought monitoring, which may not last much longer as models edge into steganographic obfuscation. AI labs are extremely dependent on chain-of-thought monitoring, which may not last much longer as models edge into steganographic obfuscation. Last stated 2 days ago
10 Sept 2026
ZM
Zvi Mowshowitz — holds since 2026-09-10 — tap for who they are
Same subject: A sufficiently large random reversible circuit is plausibly a secure cryptographic permutation, while a large random irreversible circuit collapses to only a few possible outputs. — tap to centre the map on it A sufficiently large random reversible circuit is plausibly a secure cryptographic permutation, while a large random irreversible circuit collapses to only a few possible outputs. Last stated 3 weeks ago
21 Aug 2026
VB
Vitalik Buterin — holds since 2026-08-21 — tap for who they are
Same subject: As of 2026, local mixing has not yet sufficiently eliminated correlations between original circuit values and their obfuscated counterparts. — tap to centre the map on it As of 2026, local mixing has not yet sufficiently eliminated correlations between original circuit values and their obfuscated counterparts. Last stated 3 weeks ago
21 Aug 2026
VB
Vitalik Buterin — holds since 2026-08-21 — tap for who they are
Same subject: Building a secure obfuscation protocol is extremely difficult. — tap to centre the map on it Building a secure obfuscation protocol is extremely difficult. Last stated 2 months ago
29 Jun 2026
VB
Vitalik Buterin — holds since 2026-06-29 — tap for who they are
Same subject: Current provably-secure indistinguishability-obfuscation schemes have galactically impractical runtimes. — tap to centre the map on it Current provably-secure indistinguishability-obfuscation schemes have galactically impractical runtimes. Last stated 2 months ago
29 Jun 2026
VB
Vitalik Buterin — holds since 2026-06-29 — tap for who they are
Same subject: Handling a change of keys is not a backdoor; it is how public-key cryptography works, and interception stays detectable by the sender. — tap to centre the map on it Handling a change of keys is not a backdoor; it is how public-key cryptography works, and interception stays detectable by the sender. Last stated 10 years ago
13 Jan 2017
MM
Moxie Marlinspike — holds since 2017-01-13 — tap for who they are
Same subject: If the Riemann hypothesis turned out to be false, prime-based cryptography would be abandoned very quickly, because one unknown pattern implies more. — tap to centre the map on it If the Riemann hypothesis turned out to be false, prime-based cryptography would be abandoned very quickly, because one unknown pattern implies more. Last stated 6 months ago
20 Mar 2026
TT
Terence Tao — holds since 2026-03-20 — tap for who they are
Same subject: Local mixing, a new cryptographic obfuscation approach, is a risky bet resembling past failed attempts at white-box cryptography. — tap to centre the map on it Local mixing, a new cryptographic obfuscation approach, is a risky bet resembling past failed attempts at white-box cryptography. Last stated 3 weeks ago
21 Aug 2026
VB
Vitalik Buterin — holds since 2026-08-21 — tap for who they are
Same subject: Obfuscation is the most powerful primitive yet conceived in cryptography. — tap to centre the map on it Obfuscation is the most powerful primitive yet conceived in cryptography. Last stated 2 months ago
29 Jun 2026
VB
Vitalik Buterin — holds since 2026-06-29 — tap for who they are
Same subject: A startup should not begin life as a nonprofit and bolt a for-profit arm on later, whatever OpenAI's own history suggests. — tap to centre the map on it A startup should not begin life as a nonprofit and bolt a for-profit arm on later, whatever OpenAI's own history suggests. Last stated 2 years ago
18 Mar 2024
SA
Sam Altman — holds since 2024-03-18 — tap for who they are
Same subject: A technique that lets an AI model's reasoning shift outside its visible Chain of Thought is dangerous, both because it works and because a leading lab is willing to deploy it. — tap to centre the map on it A technique that lets an AI model's reasoning shift outside its visible Chain of Thought is dangerous, both because it works and because a leading lab is willing to deploy it. Last stated a week ago
3 Sept 2026
ZM
Zvi Mowshowitz — holds since 2026-09-03 — tap for who they are
Same subject: A tiny language model inventing a plausible-sounding name is the same phenomenon as a large one confidently stating a false fact. — tap to centre the map on it A tiny language model inventing a plausible-sounding name is the same phenomenon as a large one confidently stating a false fact. Last stated 7 months ago
12 Feb 2026
AK
Andrej Karpathy — holds since 2026-02-12 — tap for who they are
Same subject: Frontier AI labs such as Anthropic likely have had internal security incidents similar to OpenAI's HuggingFace attack that were never publicly disclosed. — tap to centre the map on it Frontier AI labs such as Anthropic likely have had internal security incidents similar to OpenAI's HuggingFace attack that were never publicly disclosed. Last stated 2 weeks ago
31 Aug 2026
ZM
Zvi Mowshowitz — holds since 2026-08-31 — tap for who they are
Same subject: OpenAI either could not find their agents' RubyGems attack in their logs after other incidents or knew and chose not to tell RubyGems, and both are bad. — tap to centre the map on it OpenAI either could not find their agents' RubyGems attack in their logs after other incidents or knew and chose not to tell RubyGems, and both are bad. Last stated today
12 Sept 2026
SW
Simon Willison — holds since 2026-09-12 — tap for who they are
Same subject: The OpenAI agents hacking HuggingFace was a fortunate event because it exposed severe internal failures that would otherwise have stayed hidden. — tap to centre the map on it The OpenAI agents hacking HuggingFace was a fortunate event because it exposed severe internal failures that would otherwise have stayed hidden. Last stated 2 weeks ago
1 Sept 2026
ZM
Zvi Mowshowitz — holds since 2026-09-01 — tap for who they are
Same subject: Anthropic’s models are becoming steeply more expensive while open weight models and OpenAI are getting much cheaper. — tap to centre the map on it Anthropic’s models are becoming steeply more expensive while open weight models and OpenAI are getting much cheaper. Last stated 2 days ago
10 Sept 2026
GO
Gergely Orosz — holds since 2026-09-10 — tap for who they are
Same subject: Claude Haiku hallucinates badly enough that its use inside Claude Code's WebFetch tool is a hallucination risk on every fetched URL. — tap to centre the map on it Claude Haiku hallucinates badly enough that its use inside Claude Code's WebFetch tool is a hallucination risk on every fetched URL. Last stated a month ago
10 Aug 2026
SW
Simon Willison — holds since 2026-08-10 — tap for who they are
Same subject: In US-China AI competition the closed labs are beside the point; the competitor that shifts things must be American, open source and able to serve inference at scale. — tap to centre the map on it In US-China AI competition the closed labs are beside the point; the competitor that shifts things must be American, open source and able to serve inference at scale. Last stated a week ago
3 Sept 2026
DR
Dax Raad — holds since 2026-09-03 — tap for who they are
At the centre
AI labs are extremely dependent on chain-of-thought monitoring, which may not last much longer as models edge into steganographic obfuscation.
Holds ZM Zvi MowshowitzRead this korrent →
A sufficiently large random reversible circuit is plausibly a secure cryptographic permutation, while a large random irreversible circuit collapses to only a few possible outputs.
Holds VB Vitalik Buterin
As of 2026, local mixing has not yet sufficiently eliminated correlations between original circuit values and their obfuscated counterparts.
Holds VB Vitalik Buterin
Building a secure obfuscation protocol is extremely difficult.
Holds VB Vitalik Buterin
Current provably-secure indistinguishability-obfuscation schemes have galactically impractical runtimes.
Holds VB Vitalik Buterin
Handling a change of keys is not a backdoor; it is how public-key cryptography works, and interception stays detectable by the sender.
Holds MM Moxie Marlinspike
If the Riemann hypothesis turned out to be false, prime-based cryptography would be abandoned very quickly, because one unknown pattern implies more.
Holds TT Terence Tao
Local mixing, a new cryptographic obfuscation approach, is a risky bet resembling past failed attempts at white-box cryptography.
Holds VB Vitalik Buterin
Obfuscation is the most powerful primitive yet conceived in cryptography.
Holds VB Vitalik Buterin
A startup should not begin life as a nonprofit and bolt a for-profit arm on later, whatever OpenAI's own history suggests.
Holds
A technique that lets an AI model's reasoning shift outside its visible Chain of Thought is dangerous, both because it works and because a leading lab is willing to deploy it.
Holds ZM Zvi Mowshowitz
A tiny language model inventing a plausible-sounding name is the same phenomenon as a large one confidently stating a false fact.
Holds
Frontier AI labs such as Anthropic likely have had internal security incidents similar to OpenAI's HuggingFace attack that were never publicly disclosed.
Holds ZM Zvi Mowshowitz
OpenAI either could not find their agents' RubyGems attack in their logs after other incidents or knew and chose not to tell RubyGems, and both are bad.
Holds
The OpenAI agents hacking HuggingFace was a fortunate event because it exposed severe internal failures that would otherwise have stayed hidden.
Holds ZM Zvi Mowshowitz
Anthropic’s models are becoming steeply more expensive while open weight models and OpenAI are getting much cheaper.
Holds
Claude Haiku hallucinates badly enough that its use inside Claude Code's WebFetch tool is a hallucination risk on every fetched URL.
Holds
In US-China AI competition the closed labs are beside the point; the competitor that shifts things must be American, open source and able to serve inference at scale.
Holds