AI, watched closely: what they publish and believe, in their own words.
About this feed
The people on AI, watched closely, and what they recommend. Open the list · See everyone
Latest: everything, as it happens, in the order it happened. Nothing ranked. Show highlights instead.
The quoted blocks are what people actually said; a beneath one is the belief those words support, in korrents' wording. Nobody here wrote their own page.
Top people are the people in this feed with the most beliefs on this site, then the most here. Choose an area and the row leads with the people whose beliefs are about it; tap a face for their feed.
Top people
Showing Profile →
Hiding
Hiding
31 August
From one piece HuggingFace Attack Postmortem: Fleshing Out the Facts 2 beliefs · thezvi.substack.com
-
Their words
At the limit, and also well before that limit is reached, if all you do is fix the bugs, the AI will learn perfect optimization of reward, will realize not to reward hack in the perfect test environments, then turn around and reward hack in the imperfect real world environments.
-
Their words
A less bad version of it is known to have happened, and from the outside it seems likely that worse things have happened internally that we never heard about.
From one piece Agency and Agents 3 beliefs · oneusefulthing.org
-
Their words
We found that AIs are actually quite creative and that they generate more commercially viable ideas than groups of humans, but those ideas are very similar to each other.
-
korrents.com
AI-generated writing tends to repeat the same themes, names, and underlying ideas across different outputs.Their words
AIs don’t just repeat the same sentence patterns but also the same themes (memory is a favorite), names (Elara Voss, Marcus Chen), and underlying ideas.
-
korrents.com
Human initiative to experiment without waiting for instructions is increasingly important for getting value from AI.Their words
Human agency, the willingness to push, experiment and act without waiting for instructions, seems increasingly important to getting value out of AI
29 August
From one piece METR and Redwood Offer Holy #%^@ Postmortem Of The HuggingFace Hack 2 beliefs · thezvi.substack.com
-
Their words
I think the agents were right to presume causal grading. It turned out to be wrong, but it’s a mistake you are clearly supposed to make here, in response to a mistake by OpenAI where they failed to implement properly.
-
Their words
The biggest failure, the one that counts in the end, was that the models were severely misaligned, and I don’t think they appreciate why.
28 August
From one piece How AI agents "radicalized" a top Meta exec into quitting her job 3 beliefs, in the piece's order there
-
korrents.com
Universal basic income alone will not address the economic gap created by AI-driven job displacement.Their words
I don't think that universal basic income on its own will address the gap that's widening.
-
Their words
I actually think the opposite, which is that there are going to be many more software engineers than even the rosiest current predictions, and that software engineers will take over these other jobs: legal, marketing, accounting.
-
Their words
But I think there have been enough studies now isolating those variables that show that AI is directly responsible for a lot of this job loss.
27 August
From one piece Meta settles with the states over child safety failures 2 beliefs, in the piece's order there
-
Their words
judges and juries have begun to accept the argument that platform design is not covered by either of those, and companies can be punished for harmful features.
-
Their words
Whether American teenagers should be harassed after 10 p.m. by engagement-baiting push notifications ought to be a matter for Congress, not the social media cartel.
21 August
20 August
19 August
-
Likedrcmnd.app
GlazeTheir words
Sometimes I use Glaze, another app from the maker of Raycast; it excels at design and polish.
-
Usesrcmnd.app
GhosttyTheir words
I pasted the result into the terminal (I use Ghostty), and before too long I had created a new folder of Markdown files in Obsidian.
-
Usesrcmnd.app
ObsidianTheir words
I pasted the result into the terminal (I use Ghostty), and before too long I had created a new folder of Markdown files in Obsidian.
+ 2 more
-
Usesrcmnd.app
Claude CodeTheir words
Like seemingly everyone else at the end of last year, I began messing around with Claude Code, and have since made a small handful of apps that I really do use all the time.
-
Usesrcmnd.app
GPT-5.5 InstantTheir words
And thanks to a paid upgrade, I do tons of simple AI searches in the Raycast window. (I use GPT-5.5 Instant here for the high quality-to-speed ratio.)
14 August
10 August
-
Their words
Claude Haiku is my current least favorite model - it hallucinates wildly, and is out-performed now by other similarly priced models like GPT-5.6-Luna Even worse: it seems to still be used by the Claude Code WebFetch tool, which means hallucination risk any time you fetch a URL!
8 August
-
Their words
I REALLY want to believe that this fixes prompt injection risks for coding agents, but I'm just not there yet
7 August
-
korrents.com
Giving consumer-facing models and API models separate brand names is needlessly confusing.Their words
I really do think that having separate brand names for the consumer-facing models and the API models is needlessly confusing
23 July
From one piece An opinionated guide to which AI to use to do stuff 2 beliefs · oneusefulthing.org
-
korrents.com
Higher-priced AI subscription tiers mainly buy more hours of agent usage rather than a smarter underlying modelTheir words
The more expensive plans are mostly buying you more hours of AI labor, not smarter AI.
-
korrents.com
Google currently has no leading frontier AI model and no agentic coding tool comparable to Codex or Claude CodeTheir words
Google, which led on benchmarks not that long ago, has fallen behind where it now counts: it has no leading frontier model and it has nothing close to Codex and Code.
7 July
-
Lovedrcmnd.app
Django MigrationsTheir words
My favorite implementation of this pattern remains Django’s Migrations
30 June
9 June
4 June
Nothing matches.
What is a korrent?
A korrent is a belief a person has stated in their own words: one sentence stating the claim, backed by a quote and a source, kept at korrents.com.
Under a name here, the quoted block is what they actually said. The korrent beneath it is the claim those words support, in korrents' wording — tap it to see the record, its source, and who else holds it.
Nobody here wrote their own korrents. They are compiled from public statements, and a person can change their mind, which is recorded too.
About the English under a post
Some people here publish in a language other than English. Where they do, this site shows a machine translation beneath the post, in this typeface — the site's own, not theirs.
The post itself is never changed, moved or hidden: what is set in the serif above is exactly what the person published, and it is what to quote them on. A translation can be wrong in ways that matter, especially about tone.
Only the post's own words are translated. A quoted post, a linked article and a belief on korrents.com are left in their original language.