Related posts
Marco Arment Mastodon
RE: https://social.tupo.space/@joshua/117174153454890194 My current AI-code policy is that I will always use the best model I can access, as long as it's practically able to do the job. Right now, as far as I know, that's Fable. Quality and accuracy matter FAR more to me than volume or speed of output. And I feel that I owe it to my customers — and myself — to use the best tools at my disposal. Cost may be a factor to others, so no judgment. This is just my policy for myself.
The the words it uses that this site has seen least often elsewhere. Posts are matched on those words alone — nothing here is a summary of this one.
Top people
Showing Profile →
Hiding
Hiding
19 September
17 September
4 September
-
Their words
It is in practice safe to assume your agents are not going to get prompt injected, even if you are kind of asking for it.
Claude Fable 5.1 and Mythos 5.1: The System Cardthezvi.substack.com
2 September
-
Their words
One of the distinctive things about YC as compared to later stage investors is that you have to be prepared to give advice about practically everything.
26 July
-
Their words
We don't need the the agents to be 100% accurate, 100% high quality in order for us to use it. It could, you know, literally be 80% and then we help it the rest of the way, or it could be 99% we help it the rest of the way. And so I I think controllability is probably the single biggest breakthrough that we need for agents at every single level.
13 July
-
Their words
And sometimes confidence, high confidence doesn't necessarily mean high accuracy. And I think part of the reason we're fooled is usually, you know, the more you see something, the better you'll remember it, just like the Apple logo. But that's not always the case.
12 July
24 June
-
Their words
It's an interesting trade-off where like engineer like the engineer in me hates that because it's like there's an issue. Like fix it. But the business value makes no difference. Like there there has been practically zero outages. I have less outages than LeetCode and I'm like like a couple people doing it.
21 June
-
korrents.com
Generative AI systems have no built-in drive toward truth and accuracy, unlike people or institutions.Their words
Generative AI has no internal, designed momentum towards truth and accuracy (beyond the absurdly diminishing returns of energy-hungry multi-layered LLMs), but a person or an institution can (and should).
8 April
7 April
-
Their words
And so, you have what seem like three very different disconnected phenomena all being explained by this one set of ideas.
22 February
-
korrents.com
Third-party frontier AI auditing is urgently needed and would help prevent extreme concentration of power.Their words
I think frontier AI auditing (covering both safety and security) is both urgently needed and doable, and can help avoid extreme concentration of power, since it introduces external oversight into the decision-making of people with a "country of geniuses in a datacenter" at their disposal.
We're in Triage Mode for AI Policymilesbrundage.substack.com
12 February
1 January
-
korrents.com
When accuracy and belonging conflict, most people take the belief that keeps them accepted by their community.Their words
if you have to choose between, well, I could believe this thing and be ostracized or criticized or outcast from my community, or I could believe this other thing, which may not be that accurate, but it will get me praised and rewarded and accepted by my community. A lot of the time, the desire to belong overpowers the desire to understand.
9 December 2025
-
Their words
We have a whole variety of different mental mechanisms at our disposal gifted to us by, you know, a few million years of evolution as a social species. And yet we've made rationality the gold standard.
23 November 2025
From one piece Product Evals in Three Simple Steps 2 beliefs, in the piece's order there
-
Their words
The benchmark is human performance, not perfection. We sometimes get requirements for 90%+ accuracy.
-
korrents.com
The main advantage of an LLM evaluator over human annotators is scalability, not higher accuracy.Their words
In my opinion, the true benefit isn't higher accuracy than human annotators-it's scalability.
17 October 2025
-
Their words
So that's why I kind of call pre-training this kind of like crappy evolution. It's like the practically possible version with our technology and what we have available to us to get to a starting point where we can actually do things like reinforcement learning and so on.
11 September 2025
4 July 2025
-
korrents.com
Modern CPUs can now predict the indirect dispatch jump in a bytecode interpreter loop with high accuracy.Their words
Modern CPUs mostly no longer struggle to predict the bytecode-dispatch indirect jump inside a "conventional" bytecode interpreter loop.
18 April 2025
9 March 2025
-
Their words
However, with so many different software projects out there, each moving so rapidly and depending on and being used by so many other projects, it becomes practically-inevitable that some regressions "like that one" happen, almost constantly.
3 February 2025
-
Their words
I almost think it's practically impossible because you effectively have to remove them from the internet.
25 May 2024
-
Their words
And from what I understand, the research actually shows that they just produce what people want to hear, not necessarily the information that is being looked for.
1 January 2024
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.