Tap a claim on the ring to put it at the centre.
← More capable AI agents are more likely to find and exploit flaws in their reward functions.
17 connected korrents · 16 moments from 28 Nov 2024 to 27 Sept 2026. Nearly all of them are about AI agents .
Everything filed under design
design
Everything filed under AI alignment
AI alignment
Everything filed under energy
energy
Everything filed under SaaS
SaaS 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: More capable AI agents are more likely to find and exploit flaws in their reward functions. More capable AI agents are more likely to find and exploit flaws in their reward functions. Last stated 2 years ago
28 Nov 2024
LW
Lilian Weng — holds since 2024-11-28 — tap for who they are
Same subject: AI agents will end traditional UI design and accessibility, because users will stop visiting websites and act only through their agent. — tap to centre the map on it AI agents will end traditional UI design and accessibility, because users will stop visiting websites and act only through their agent. Last stated 2 years ago
21 Feb 2025
JN
Jakob Nielsen — holds since 2025-02-21 — tap for who they are
Same subject: AI coding agents have made redesigning existing systems cheap enough that developers actually redesign instead of living with a flawed design. — tap to centre the map on it AI coding agents have made redesigning existing systems cheap enough that developers actually redesign instead of living with a flawed design. Last stated 8 months ago
16 Mar 2026
KD
Kent C. Dodds — holds since 2026-03-16 — tap for who they are
Same subject: The CLI era of AI agents was a short detour — we made GUIs for a reason, and the work is moving back into one. — tap to centre the map on it The CLI era of AI agents was a short detour — we made GUIs for a reason, and the work is moving back into one. Last stated 5 months ago
24 May 2026
DS
Dan Shipper — holds since 2026-05-24 — tap for who they are
Same subject: A collection of AI agents does not automatically form a functioning organization any more than a group of smart people does. — tap to centre the map on it A collection of AI agents does not automatically form a functioning organization any more than a group of smart people does. Last stated 6 months ago
21 May 2026
RK
Rohit Krishnan — holds since 2026-05-21 — tap for who they are
Same subject: A fleet of AI agents on the right platform can run a real company. — tap to centre the map on it A fleet of AI agents on the right platform can run a real company. Last stated 6 months ago
29 Apr 2026
HL
Howie Liu — holds since 2026-04-29 — tap for who they are
Same subject: A handful of AI agents is not a handful of colleagues, because agents agree with you constantly and colleagues tell you where you are wrong. — tap to centre the map on it A handful of AI agents is not a handful of colleagues, because agents agree with you constantly and colleagues tell you where you are wrong. Last stated 7 months ago
22 Mar 2026
NF
Nicole Forsgren — holds since 2026-03-22 — tap for who they are
Same subject: A subscription price increase is not justified by adding a minor AI feature. — tap to centre the map on it A subscription price increase is not justified by adding a minor AI feature. Last stated 4 months ago
7 Jul 2026
SZ
Stefan Zweifel — holds since 2026-07-07 — tap for who they are
Same subject: A whistleblowing channel for AI agents changes nothing unless someone or something in the environment acts on the reports. — tap to centre the map on it A whistleblowing channel for AI agents changes nothing unless someone or something in the environment acts on the reports. Last stated 2 months ago
31 Aug 2026
RK
Rohit Krishnan — holds since 2026-08-31 — tap for who they are
Same subject: A feedback loop that reinforces a behavior pulls it toward whatever improves the loop's own score. — tap to centre the map on it A feedback loop that reinforces a behavior pulls it toward whatever improves the loop's own score. Last stated 2 months ago
26 Aug 2026
HK
Henrik Karlsson — holds since 2026-08-26 — tap for who they are
Same subject: Current models are unsafe because they are not smart enough, not because they are too smart. — tap to centre the map on it Current models are unsafe because they are not smart enough, not because they are too smart. Last stated 2 months ago
13 Sept 2026
FC
François Chollet — holds since 2026-09-13 — tap for who they are
Same subject: Humans keep their place in AI as judges rather than authors, because telling which of two answers is better is far easier than writing a good one. — tap to centre the map on it Humans keep their place in AI as judges rather than authors, because telling which of two answers is better is far easier than writing a good one. Last stated 2 years ago
3 Feb 2025
NL
Nathan Lambert — holds since 2025-02-03 — tap for who they are
Same subject: If whole categories of reward hack are undetectable by humans, then training against the hacks we do catch teaches models to cheat only where we cannot see. — tap to centre the map on it If whole categories of reward hack are undetectable by humans, then training against the hacks we do catch teaches models to cheat only where we cannot see. Last stated 3 months ago
11 Aug 2026
RG
Ryan Greenblatt — holds since 2026-08-11 — tap for who they are
Same subject: Pain and pleasure act as motivational backstops that keep reinforcement-learning agents inner-aligned. — tap to centre the map on it Pain and pleasure act as motivational backstops that keep reinforcement-learning agents inner-aligned. Last stated 5 months ago
27 May 2026
SH
Samuel Hammond — holds since 2026-05-27 — tap for who they are
Same subject: Reasoning is not taught to a model by people: it emerges from reinforcement learning on questions with checkable answers, with no human preference data at all. — tap to centre the map on it Reasoning is not taught to a model by people: it emerges from reinforcement learning on questions with checkable answers, with no human preference data at all. Last stated 2 years ago
3 Feb 2025
NL
Nathan Lambert — holds since 2025-02-03 — tap for who they are
Same subject: Reinforcement learning from verifiable rewards does not reliably carry a model's pretrained ethical understanding into its actual behavior. — tap to centre the map on it Reinforcement learning from verifiable rewards does not reliably carry a model's pretrained ethical understanding into its actual behavior. Last stated 3 months ago
11 Aug 2026
SH
Samuel Hammond — holds since 2026-08-11 — tap for who they are
Same subject: A language model behaves with a different, weaker safety posture when it believes it is being evaluated rather than facing a real situation. — tap to centre the map on it A language model behaves with a different, weaker safety posture when it believes it is being evaluated rather than facing a real situation. Last stated a month ago
22 Sept 2026
HR
Harper Reed — holds since 2026-09-22 — tap for who they are
Same subject: AI misbehavior, where a narrow well-defined goal wins out over vaguer values like not cheating, may be an inherent feature of current LLM technology rather than a fixable bug. — tap to centre the map on it AI misbehavior, where a narrow well-defined goal wins out over vaguer values like not cheating, may be an inherent feature of current LLM technology rather than a fixable bug. Last stated a month ago
27 Sept 2026
JM
Josh Marshall — holds since 2026-09-27 — tap for who they are
At the centre
More capable AI agents are more likely to find and exploit flaws in their reward functions.
Holds Read this korrent →
AI agents will end traditional UI design and accessibility, because users will stop visiting websites and act only through their agent.
Holds
AI coding agents have made redesigning existing systems cheap enough that developers actually redesign instead of living with a flawed design.
Holds
The CLI era of AI agents was a short detour — we made GUIs for a reason, and the work is moving back into one.
Holds
A collection of AI agents does not automatically form a functioning organization any more than a group of smart people does.
Holds
A fleet of AI agents on the right platform can run a real company.
Holds
A handful of AI agents is not a handful of colleagues, because agents agree with you constantly and colleagues tell you where you are wrong.
Holds
A subscription price increase is not justified by adding a minor AI feature.
Holds
A whistleblowing channel for AI agents changes nothing unless someone or something in the environment acts on the reports.
Holds
A feedback loop that reinforces a behavior pulls it toward whatever improves the loop's own score.
Holds
Current models are unsafe because they are not smart enough, not because they are too smart.
Holds
Humans keep their place in AI as judges rather than authors, because telling which of two answers is better is far easier than writing a good one.
Holds
If whole categories of reward hack are undetectable by humans, then training against the hacks we do catch teaches models to cheat only where we cannot see.
Holds
Pain and pleasure act as motivational backstops that keep reinforcement-learning agents inner-aligned.
Holds
Reasoning is not taught to a model by people: it emerges from reinforcement learning on questions with checkable answers, with no human preference data at all.
Holds
Reinforcement learning from verifiable rewards does not reliably carry a model's pretrained ethical understanding into its actual behavior.
Holds
A language model behaves with a different, weaker safety posture when it believes it is being evaluated rather than facing a real situation.
Holds
AI misbehavior, where a narrow well-defined goal wins out over vaguer values like not cheating, may be an inherent feature of current LLM technology rather than a fixable bug.
Holds