Kelsey Piper
Journalist writing on AI, effective altruism and global development; contributing writer at The Argument, previously a senior writer at Vox's Future Perfect.
Kelsey Piper did not write this page. ppll compiled it from their own words, each linked to the source. Is this you? · Report a mistake
Latest x.com
What they believe
-
If AIs can now considerably improve the training of future AIs, this is very unlikely to end well.
if Anthropic is correct, if we now have AIs that are smart enough to considerably improve the process of training future AIs, and if things can be expected to speed up rather than slow down from here, then it seems very unlikely to me that this ends well.
theargumentmag.comSaid 21 Oct 2026This on korrents
-
As machine learning models improve, it is likely that minors will be fully removed from the majority of the internet.
As machine learning models improve, it's likely that minors will be fully booted from the majority of the internet.
theargumentmag.comSaid 13 Jul 2026This on korrents
The bold line is korrents’ wording; the quote is theirs. These 2: what they wrote before what they said aloud, one per source, newest first.
Beliefs
Korrents What they believe 20 beliefs — each backed by an exact quote.
Each is a — compiled by korrents.com, not by them: the one-line wordings are korrents', the quotes are theirs.
Recent
If AIs can now considerably improve the training of future AIs, this is very unlikely to end well.
if Anthropic is correct, if we now have AIs that are smart enough to considerably improve the process of training future AIs, and if things can be expected to speed up rather than slow down from here, then it seems very unlikely to me that this ends well.
The Opening Argument Said 21 Oct 2026
A world where humans lack economic value or necessity may result in the loss of significant individual rights.
I do not ultimately expect that a world where humans are without economic value, where humans are not necessary to hold power or build things, is a world where humans retain significant individual rights.
The Opening Argument Said 21 Oct 2026
The ultimate conclusion of runaway AI improvement may not be a super-powerful, super-helpful assistant.
I don't think if you let this runaway process of AI improvement spiral to its ultimate conclusion, you get a super-powerful, super-helpful assistant. I suspect you get something else entirely.
The Opening Argument Said 21 Oct 2026
Show 17 more
As machine learning models improve, it is likely that minors will be fully removed from the majority of the internet.
As machine learning models improve, it's likely that minors will be fully booted from the majority of the internet.
Jonathan Haidt is wrong about age-gating the internet Said 13 Jul 2026
It is impossible to configure a computer to allow meaningful web interaction without bombarding a user with addictive content.
I wish everyone giving this parenting advice were first required to actually try, for an hour, to configure a computer such that it gives a user the ability to meaningfully interact with the web without bombarding them with content selected to make them hooked and miserable. It can't be done.
Jonathan Haidt is wrong about age-gating the internet Said 13 Jul 2026
If advertising to minors were taxed similarly to tobacco, it would likely destroy the ecosystem of low-quality content aimed at children.
If you could not make money off advertising to minors (say, because we imposed taxes on it like we do on tobacco), I suspect that would crush most of the slop-for-kids ecosystem.
Jonathan Haidt is wrong about age-gating the internet Said 13 Jul 2026
The era when literary magazines can operate solely on trust is over.
I don't like it any more than anyone else, but the age when you can run a literary magazine on trust is over.
The literary world is sleepwalking into an AI disaster Said 26 May 2026
Refusing to engage with AI tools leads to being mistaken about their strengths and limitations.
If you are unwilling to check it out, you will be wrong about what its strengths and limitations are - and you'll miss that a growing share of the words you are reading are AI-generated.
The literary world is sleepwalking into an AI disaster Said 26 May 2026
The current approach of edtech companies making games with minor educational elements results in software that is hard to distinguish from casual games like Candy Crush.
It's not just Prodigy, Oremus observed correctly; the entire premise of edtech companies like Gimkit or Blooket is making games with a minor educational element, with the ultimate result that "schools have embraced education software that has become hard to distinguish from Candy Crush."
Homework shouldn’t be all fun and games Said 19 May 2026
The 'content plus game' setup in edtech cannot teach sustained focus on difficult tasks.
And this sort of setup can never, ever teach "sustained focus on something that looks hard at first glance," which is one of the most valuable muscles to help children exercise.
Homework shouldn’t be all fun and games Said 19 May 2026
Giving children a worksheet followed by the reward of playing video games would be more effective than using edutainment.
I think it would be a lot more effective to just give your kid an old-fashioned worksheet and then say "When you're done, you can play video games all evening." Then, they would pick whatever game they like most and probably both do more math and have more fun.
Homework shouldn’t be all fun and games Said 19 May 2026
A safe and clean environment should be a basic obligation of society for everyone, not just a privilege for the wealthy.
A safe, clean environment should not be a privilege for the ultra-wealthy; it is the basic obligation of our society to provide it to everybody.
Our moral obligation to prevent crime Said 5 May 2026
The principled liberal position on crime does not require abandoning the values of being humane and compassionate.
But I actually think that the principled liberal position on crime and public order does not require a departure from the general Democratic ethos of being humane, compassionate, and concerned for the marginalized.
Our moral obligation to prevent crime Said 5 May 2026
Trust in the criminal justice system will likely take much longer to grow than the rate of homicides takes to fall.
But while we have a plan for addressing crime - focused violence reduction programs, improved enforcement, more police on the street - and tools like fare gates to address vandalism, my prediction is that trust will take much longer to grow than homicides take to fall.
Our moral obligation to prevent crime Said 5 May 2026
AI progress from 2024 to 2026 has been faster than the progress seen from 2022 to 2024.
By almost every metric, AI progress from 2024 to 2026 has been much faster than AI progress from 2022 to 2024.
AI's biggest critic has lost the plot Said 28 Apr 2026
The cost of querying an AI model at a specific capability level has decreased significantly.
At the same time, the cost of querying a model at a given capability level has plummeted. GPT-4-level intelligence is about 1/1000th as expensive as it was when GPT-4 was released.
AI's biggest critic has lost the plot Said 28 Apr 2026
AI makes software engineers more productive.
Making them more productive is a big deal, and in 2026, AI makes them more productive.
AI's biggest critic has lost the plot Said 28 Apr 2026
Technological progress is not a guarantee that any specific technology will be a force for good.
Technological progress has been an enormous force for good in the world, but that is no guarantee that any specific technology will be.
Why hasn't AI cured cancer? Said 3 Mar 2026
The government's actions may cause AI companies to be more reluctant to engage with the government and more inclined to work in secrecy.
But I think a much likelier consequence of this behavior by the government is for AI labs to be much more reluctant to engage with the government, much less inclined to build partnerships, and much more inclined to work in secrecy.
Anthropic is somehow both too dangerous to allow and essential to national security Said 26 Feb 2026
Claude could not currently be used to power autonomous weapons; it would endanger troops and be fairly useless.
You couldn't use Claude to power a killer bot right now even if you wanted to; it would be a danger to our troops and fairly useless at our objectives.
Anthropic is somehow both too dangerous to allow and essential to national security Said 26 Feb 2026
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.
Recommends rcmnd
rcmnd What they recommend 1 thing — each backed by an exact quote.
Compiled by rcmnd.app, not by them: the quotes are theirs, and each links to the post it was copied from.
Recent
Recommends Nancy Drew
Give them the original Nancy Drew books; the ones where Nancy packed heat.
theargumentmag.com 30 Jan 2026
Where they publish
Newsletter Kelsey Piper on Substack Aggregates her bylined pieces, mostly at The Argument.
Link verified 25 Sept 2026.
x.com @kelseytuoc Personal account.
Recent
- Gemini Carbon+Neon are currently unbeatable at "run this experiment and present me with some useful analysis of the results of it". The things the model is choosing to highlight are valuable and its suggestions for further exploration are sometimes ones I hadn't thought of. 20 Nov 2026
- I don't think large scale internet outages are super likely, but I do think they're *more* likely, and I have shifted to trying to keep a couple weeks' food in the house. This isn't about prepping for the apocalypse, which I don't think is worthwhile. 19 Nov 2026
- Tomorrow is Anthropic's IPO and also a Ban RSI rally at its headquarters: 18 Nov 2026
Show 17 more
- Not seeing a lot of love for Astra 6.2 but (perhaps as a consequence of its continued improvement in no-COT reasoning) it's quite good on my tests for smart, coherent in-character video game dialogue. 17 Nov 2026
- In The Argument, I write about a different big question about AI: what are we doing here? What are we hoping for? What would it even look like for this to go well? 17 Nov 2026
- As people in the Bay have started to expect RSI to arrive sooner and sooner, some of them have basically given up on legislation or bilateral treaties, because they figure it just can't happen fast enough. 16 Nov 2026
- seems like we are (briefly?) back in the regime where prompting and having the right tools and context and examples stored means people are getting vastly different results out of the same models. 15 Nov 2026
- The debate over this paper is superficially about the drop-out rate, but I wonder for how many people that's actually decisive. If you're opposed to letting 14-17 year olds transition, and extremely high quality evidence found that 3% will regret it, would that change your mind? 13 Nov 2026
- This is awful, of course, but I actually find it kind of heartening if the attacks on Jews were coordinated by the Iranian government rather than being spontaneous expressions of organic Jew hatred 12 Nov 2026
- Really interesting how the general view on Twitter of Fable 5.2 has been "eh, it's okay" while in these sorts of benchmarks it places extremely high. 11 Nov 2026
- you can go here and put in your url to see if the Nvidia influence operation bots interacted with you. 11 Nov 2026 sockpuppet-investigation-explorer.com
- Couple thoughts: 1) After several people pointed out that Lyman couldn't substitute all-cause mortality for malaria-specific mortality like he had done, he fixed it. His initial estimate for $/lives saved was $19000; revised it’s about $8,200. 2) I think, though I could be wrong about this, that his calculation to get $8200 still has a problem: he calculates the costs of the distribution based on the whole population but the benefits based on 75% of the population (since he argues the other 25% is already sleeping under nets). But as he himself identifies, the nice thing about an intention-to… 11 Nov 2026
- Who's behind the swarms of influencer bots out commenting on Nvidia/Jensen Huang? Apparently the same accounts also post on behalf of the RSF in Sudan, and tie back to an Israeli company which pretended to shut down while under investigation for their pro-Qatar operations 10 Nov 2026
- Narrating aloud my best guess of what the baby is thinking is an extremely powerful tactic to get his siblings to flock over and patiently try to set him straight. 10 Nov 2026
- 1) great news, good for them 2) it should be actually, genuinely independent, with lots of competitive districts. The lesson of the R dummymander, and Indiana escaping the R dummymander, is that sometimes doing the right thing is also good politics. 9 Nov 2026
- Fable 5.2 is really fast. I had a sort of default workflow of setting Claude on something and then tabbing over to my inbox or edits or Twitter, which no longer works because it answers me too quickly 9 Nov 2026
- I believe the exact opposite of this worldview. I'm okay with waiving the fine on a first offense, but anyone who drives recklessly ten times a week even after repeatedly getting ticketed for it should just lose their license forever. 9 Nov 2026
- Okay find the person who created the 15 minute approval process for posts and put them in charge of a 15 day approval process for houses 9 Nov 2026
- seems like by far the worst part of your husband having an affair is the part where the internet rushes to look at pictures of you next to the woman he cheated with and debate whether you are hideous enough that he should get a pass. 9 Nov 2026
- reminding everyone that Mississippi is one of the best states in the country at teaching kids to read and the rest of us should be taking notes 7 Nov 2026
Link verified 25 Sept 2026. Recent items update automatically from the channel.
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.