Sign in

Kevin Roose

@kevinroose.com
64K followers 151 following 102 posts

NYT tech columnist, Hard Fork co-host, best at 0.8x speed

PostsRepliesMedia
Kevin Roose @kevinroose.com · 17/06/2025
The NYT Magazine asked me and @caseynewton.bsky.social to open their AI issue by talking about how we use AI, why it's not going away, and how you can't be a serious critic if you're in denial about how useful it is. Bluesky's gonna love this one! www.nytimes.com/2025/06/16/m...
73830921
Kevin Roose @kevinroose.com · 03/05/2025
It's Hard Fork Friday! This week on the show: -- A hard look at Glazegate, and why it's bad when AI models are too nice -- I got scanned by a Worldcoin orb -- And @pjvogt.bsky.social joins us for a new segment about what's going on in our group chats www.youtube.com/watch?v=3Lum...
youtube.com
How Nice Is Too Nice for an AI Chatbot? | EP 134
YouTube video by Hard Fork
9636
Kevin Roose @kevinroose.com · 25/04/2025
My new column is drawing rave reviews like “go fuck yourself,” “die, loser” and more! Take a look and add yours to the pile!
26845524
Kevin Roose @kevinroose.com · 21/04/2025
Some news: I'm writing a book about AGI! This has been a while in the making, very excited to front-burner it for a while. I think what's happening in AI today is the most important and exciting story in the world, bar none, and I can't imagine a richer topic to write hundreds of pages about.
agichronicles.com
Why I'm Writing a Book About AGI
And how you can help
221126
Kevin Roose @kevinroose.com · 21/02/2025
It's Hard Fork Friday! This week on the show: • Grok 3 is here, and it's...woke? • An interview with Robinhood CEO Vlad Tenev about whether we're becoming a nation of gamblers • And I build Casey a hot tub time machine using AI www.nytimes.com/2025/02/21/p...
nytimes.com
How ‘Based’ Is Grok 3? + Robinhood C.E.O. Vlad Tenev on Markets for Everything + Vibecoding 101
“Elon Musk is willing to spend a phenomenal amount of money and basically do everything he can to stay with the head of the pack on A.I. progress.”
3814414
Kevin Roose @kevinroose.com · 14/02/2025
It's Hard Fork Friday! This week's show is a Valentine's Day spectacular! • What happened at the big AI summit in Paris • Is AI safety dead? • How our producer Rachel uses AI to find love • And Hinge CEO Justin McLeod on the AI-filled future of online dating www.nytimes.com/2025/02/14/p...
nytimes.com
A.I. Accelerates in Paris + Can A.I. Fix Your Love Life?
It’s a Valentine’s Day Special!
341588
Kevin Roose @kevinroose.com · 01/02/2025
I spent a week testing Operator, OpenAI's new AI agent that can browse the web and take actions on your behalf. It ordered me dog food, booked a haircut, took online surveys for cash, and more. Spooky stuff! Even spookier: it registered me for a LinkedIn webinar. www.nytimes.com/2025/02/01/t...
nytimes.com
A Look at OpenAI’s Operator, a New A.I. Agent (Gift Article)
Operator, a new computer-using tool from OpenAI, is brittle and occasionally erratic, but it points to a future of powerful A.I. agents.
4634222
Kevin Roose @kevinroose.com · 31/01/2025
It's Hard Fork Friday! This week on the show: • A DeepSeek DeepDive with the great @jordanschneider.bsky.social, who explains where DeepSeek sits in China's AI ecosystem • We try OpenAI's new AI agent, Operator • And choo choo, it's the Hot Mess Express! www.nytimes.com/2025/01/31/p...
nytimes.com
DeepSeek Deep Dive + Hands-On With Operator + Hot Mess Express!
“DeepSeek is a really odd duck.”
1819810
Kevin Roose @kevinroose.com · 24/01/2025
It's Hard Fork Friday! This week on the show: • TikTok is both dead and alive, Schrödinger's app • How tech CEOs are trying to do business with Trump • Memecoins are back, are you rich yet? • And @chrislhayes.bsky.social joins to talk about his new book on attention www.nytimes.com/2025/01/24/p...
nytimes.com
Quantum TikTok + Memecoin Mania + Chris Hayes on the Attention Wars
“We are starting to see how Silicon Valley wants to do business during the Trump administration.”
2729319
Kevin Roose @kevinroose.com · 23/01/2025
I wrote about a new AI evaluation called "Humanity's Last Exam," a collection of 3,000 questions submitted by leading academics to try to stump leading AI models, which mostly find today's college-level tests too easy. www.nytimes.com/2025/01/23/t...
nytimes.com
A Test So Hard No AI System Can Pass It — Yet (Gift Article)
The creators of a new test called “Humanity’s Last Exam” argue we may soon lose the ability to create tests hard enough for A.I. models.
1720746
Reposted by Kevin Roose
John Herrman @jwherrman.bsky.social · 22/01/2025
on the tech industry's new "deal" nymag.com/intelligence...
Intelligencer
GIVE A GIFT
It is at least a little bit funny for one of the world's biggest beneficiaries of strategically written, time-sensitive venture-capital deals to tell the story of the last 30 years in tech in terms of an amorphous counterparty welching on an unwritten contract; it's the familiar "I didn't leave the party, the party left me" story, given new life in translation to the language of start-ups and funding. At least ask for a term sheet next time!
It's also the sort of pat, slightly contrarian story that dares you to ask obvious follow-up questions. Even if you accept and account for real ideological changes in the Democratic Party since the age of peak economic liberalism, is it possible that anything else happened in the tech industry between 1994 and 2016 that might have complicated its relationship with politicians and the general public? How about the economy in general? Were Republicans and conservatives vehemently criticizing and antagonizing and legislating against the tech industry at any point during this time, and if so, why didn't it matter? Did - and I'm just spitballing here — a "giant economic boom" actually come to pass, leading to the creation of some of the largest firms on earth, associated with some of the richest people on earth, whose exercise of power was felt by nearby institutions and everyday people in rather direct ways? Is it possible that the varied experiences and concerns of hundreds of millions of actual people with these companies
- not just the activist impulses of "communist" tech workers whose employers, according to Andreessen, "felt like they were hours away from full-blown violent riots on their own campuses" - might have had some downstream effects on this unspoken arrangement?There are less obvious aspects of this new unspoken deal, however, that might also prove relevant. For example: While the new Trump administration is widely expected to reduce regulatory burdens and taxes on small and large tech companies, president has also, on previous occasions, threatened to throw some of his new allies in jail, to investigate and prosecute tech firms on ideological and personal grounds, and to wield threats of regulation in targeted and politicized ways. This new deal also seems to include an awful lot of ego-tending and loyalty-signaling under duress.
In this sense, it resembles the fondly remembered '90s, in which a dynamic, growing industry was largely left alone, less than it resembles the conservative tech horror story of the last four years, in which a vastly more powerful but still dynamic industry was very much not left alone, trapped between an unpredictable and intransigent government and whipsawing public opinion. Al Gore's tech-nerd mascots have been replaced with Trump's trillion-dollar mogul collection, placed on proud display during the inauguration, each facing genuine and novel political uncertainty about their respective companies and fortunes: Am I going to get that launch contract, or is Elon going to shut me out? Is "everything now legal" in crypto, post-Trump meme coin? Do we have an outside chance of buying into a semi-nationalized TikTok, or will we end up dealing with a competitor half-owned by the United States government? Are we going to be shut out of this Al infrastructure project? What does our chatbot say when you ask it about Donald Trump?
04613
Kevin Roose @kevinroose.com · 17/01/2025
It's Hard Fork Friday! This week: • TikTok ban v12 final final USE THIS ONE • HuggingFace's @sashamtl.bsky.social joins to talk about the environmental impact of AI • And finally, put on your gold chains and pop some Zyns, it's time for MASCULINE ENERGY www.nytimes.com/2025/01/17/p...
nytimes.com
Goodbye TikTok, Ni Hao RedNote? + A.I.’s Environmental Impact + Meta’s Masculine Energy
“This is truly one of the funniest and most unexpected stories of the young year so far.”
5958
Kevin Roose @kevinroose.com · 16/01/2025
Coming tomorrow: the most masculine Hard Fork episode ever
5272914
Reposted by Kevin Roose
Casey Newton @caseynewton.bsky.social · 16/01/2025
It really is such masculine energy to blame things on a woman who no longer works at the company www.nytimes.com/2025/01/16/u...
Mr. Zuckerberg was amenable. He signaled to Mr. Miller and his colleagues, including other senior Trump advisers, that he would do nothing to obstruct the Trump agenda, according to three people with knowledge of the meeting, who asked for anonymity to discuss a private conversation. Mr. Zuckerberg said he would instead focus solely on building tech products.

Mr. Zuckerberg blamed his former chief operating officer, Sheryl Sandberg, for Facebook’s culture, according to one of the people with knowledge of the meeting. He said new guidelines and a series of layoffs amounted to a reset and that more changes were coming.
1672289054995
Kevin Roose @kevinroose.com · 14/01/2025
Normal company!
118010
Reposted by Kevin Roose
Max Read @maxread.info · 10/01/2025
on zuck's gold chain open.substack.com/pub/maxread/...
It’s been clear for a while now that Zuckerberg has been Up To Something. Bloomberg and The Wall Street Journal have described it as the “Zuckaissance”: Over the past eighteen months or so he grew out his hair; he replaced his hoodies with boxy tees; he got really into M.M.A. and wakeboarding. And, yes, he started wearing a gold chain. New Zuck is undeniably less off-putting than old, sweaty-hoodie, Caesar-cut Zuck. But he’s also unmistakably fratty, butch, and (to borrow an overused Twitter phrase) “right-coded,” partaking in the aesthetic and the hobbies of people you would expect to own crypto, listen to mindset podcasts, and vote for Trump (or, at least, refuse to vote for Biden).

It’s not exactly groundbreaking that a rich 40-year-old man has started wearing expensive streetwear. But usually this kind of personal journey culminates in divorce or hilarious/gruesome police body-cam footage or an announcement that you’re moving to Africa for a year, not in a video about new content-moderation guidelines. Zuck seems to have slowly transformed himself into a Dana White hanger-on who dresses like a Kick streamer in order to make to make this post-election right-wing turn seem authentic and deeply felt, rather than merely convenient.I wouldn’t object if you wanted to conspiracize this--a man with access to incredibly granular, up-to-the-minute data on shifting socio-cultural and political trends makes himself a human Trending Topics module in his mode of dress and choice of hobbies, a reflection of the activity and sentiment rippling across his vast empire--but I also wouldn’t object if you wanted to insist to me that this new makeover reflects an authentic political evolution. Certainly, Zuckerberg would like everyone to think it does; Kevin Roose reports that “several friends and associates of Mr. Zuckerberg’s” have told him that the aesthetic overhaul is accompanied by a genuine change in beliefs: “the billionaire’s personal politics have shifted sharply to the right since 2020, and that his embrace of Mr. Trump may stem less from cynical opportunism than real enthusiasm.”

But the interesting thing to me is less the question of authenticity around Zuck’s evolution and more the public-facing strategy. It would be easy enough, after all, to meet Trump at Mar-a-Lago a few times, send a few letters to House Judiciary, and call it a day. Why is Zuck so invested in, effectively, “vice-signaling”? Why is it so important to him to build cultural credibility as a--well, if not a conservative, precisely, then as an “anti-establishment”-coded “classical liberal”? Why is he dressed like that?
37592115
Kevin Roose @kevinroose.com · 11/01/2025
Inspiring words on innovation from the inventor of Stories, Reels, and Threads.
1081654202
Kevin Roose @kevinroose.com · 10/01/2025
It’s Hard Fork Friday! This week on the show: • The gadget robbers did NOT take from my house • What’s really behind Meta’s MAGA pivot • Deepseek, o3, and catching up on a wild month of AI progress • And 2025’s first game of HatGPT! www.nytimes.com/2025/01/10/p...
nytimes.com
Meta Goes MAGA Mode + a Big Month in A.I. + HatGPT
“I think this set of changes that the company announced this week are the most important series of policy changes that they have made in the past five years.”
8868
Kevin Roose @kevinroose.com · 06/01/2025
Everyone assumes Zuckerberg is a cynical operator who is cozying up to Trump for business reasons, but I wouldn’t discount the possibility that spending so much time around MMA guys/in the Rogansphere has truly pilled him.
2220234
Kevin Roose @kevinroose.com · 04/01/2025
5 minutes after giving my toddler kinetic sand: this stuff is incredible, a marvel of material science day 3 of finding lumps of sand in every furniture crevice in the house: whoever invented this should be tried at The Hague
4472527
Kevin Roose @kevinroose.com · 31/12/2024
Every year, I compile a list of my favorite tech projects, and honor them in a column called the Good Tech Awards. This year, I was excited to feature LLMs for genomics, AI brain mapping, open-source maintainers, the NASA heroes who saved Voyager 1, and more. www.nytimes.com/2024/12/30/t...
nytimes.com
The 2024 Good Tech Awards (Gift Article)
In a year of continued A.I. progress, “founder mode” drama and a Trump election win, a few tech projects stood out for their clear benefits to humanity.
617131
Kevin Roose @kevinroose.com · 25/12/2024
the feeling when the scissors glide through the wrapping paper >>>>>>>>
71845
Reposted by Kevin Roose
Casey Newton @caseynewton.bsky.social · 20/12/2024
"All intuition about AI capabilities will need to get updated for o3," says the founder of the ARC prize — designed to be very difficult for LLMs to solve — after OpenAI took the benchmark from 5% with o1 to 85% today with o3 arcprize.org/blog/oai-o3-...
This is a surprising and important step-function increase in AI capabilities, showing novel task adaptation ability never seen before in the GPT-family models. For context, ARC-AGI-1 took 4 years to go from 0% with GPT-3 in 2020 to 5% in 2024 with GPT-4o. All intuition about AI capabilities will need to get updated for o3.
1315218
Kevin Roose @kevinroose.com · 20/12/2024
It’s Hard Fork Friday! This week, @caseynewton.bsky.social and I went to Arizona to see Amazon’s overhauled drone delivery program, which brought us a jar of Brazilian Bum Bum Cream in 45 minutes. www.nytimes.com/2024/12/20/p...
nytimes.com
Is Amazon’s Drone Delivery Finally Ready for Prime Time?
We flew to Arizona to test the recently debuted service ourselves.
14867
Kevin Roose @kevinroose.com · 20/12/2024
Putting together my annual Good Tech Awards list of the tech projects/startups/nonprofits that contributed to human flourishing in 2024. Past lists have included things like AI for accessibility, carbon capture, AlphaFold, etc. What should be on it this year?
17455
Kevin Roose @kevinroose.com · 13/12/2024
It's Hard Fork Friday! This week: • TikTok is maybe going to get banned for real? • Julian Kelly from Google explains their new Willow quantum chip (which I almost understand) • And the cult of Claude! Podcast: www.nytimes.com/2024/12/13/p... YouTube version: www.youtube.com/watch?v=cyUV...
youtube.com
Will Trump Ban TikTok? | EP 113
YouTube video by Hard Fork
61007
Kevin Roose @kevinroose.com · 13/12/2024
I wrote about the most eligible bachelor in San Francisco's tech scene: a chatbot named Claude, who is helping AI insiders with everything from legal advice to relationship drama. www.nytimes.com/2024/12/13/t...
nytimes.com
Why Anthropic’s Claude Is a Hit with Tech Insiders (Gift Article)
A.I. insiders are falling for Claude, a chatbot from Anthropic. Is it a passing fad, or a preview of artificial relationships to come?
3712615
Kevin Roose @kevinroose.com · 07/12/2024
It’s Hard Fork Friday! This week on the show: 😮 Legendary chips reporter Don Clark joins to talk about Intel’s surprise CEO ouster and the state of chips 🤖 A recap of our trip to the Curve AI conference 🎁 And the Hard Fork gift guide! www.nytimes.com/2024/12/06/p...
nytimes.com
Is Intel Cooked? + What’s Your P(Dyson Sphere)? + Hard Fork Gift Guide
The sudden departure of Intel’s C.E.O. reignites concerns about U.S. dependence on foreign manufacturing.
5969
Reposted by Kevin Roose
Casey Newton @caseynewton.bsky.social · 06/12/2024
One of the most important debates in tech right now is the group of folks who think AI is fake and sucks vs. the people who think AI is real and dangerous. I wrote about why I'm in the latter camp, and talked about my differences with Gary Marcus www.platformer.news/ai-skeptics-...
Most people know these systems are flaws, and adjust their expectations and usage accordingly. The “AI is fake and sucks” crowd is hyper-fixated on the things it can’t do — count the number of r’s in strawberry, figure out that the Onion was joking when it told us to eat rocks — and weirdly uninterested in the things it can.

And that’s a problem, because just as these systems are more honest and helpful than they have ever been, they are also causing greater harm. And to name a real harm, already happening today, I offer the chief security officer of Amazon, CJ Moses, who had this to say about how generative AI is being used in efforts to disrupt critical infrastructure in an interview with the Wall Street Journal last month:

We’re seeing billions of attempts coming our way. On average, we’re seeing 750 million attempts per day. Previously, we’d see about 100 million hits per day, and that number has grown to 750 million over six or seven months.
This is the ongoing blind spot of the “AI is fake and sucks” crowd. This is the problem with telling people over and over again that it’s all a big bubble about to pop. They’re staring at the floor of AI’s current abilities, while each day the actual practitioners are successfully raising the ceiling.
10548667
Kevin Roose @kevinroose.com · 29/11/2024
It’s Hard Fork Friday! Today, in a very special, extra-manic Thanksgiving episode, we count down the 100 most iconic technologies of all time. Did sundials outrank the Walkman? Is penicillin more iconic than Waymo? Only one way to find out! www.nytimes.com/2024/11/29/p...
nytimes.com
Hard Fork’s 100 Most Iconic Technologies
“I just love imagining all of the angry emails we’re going to get from people who are like, ‘Why did you put the fulcrum on this list?’”
121334
Kevin Roose @kevinroose.com · 27/11/2024
Excited about this week’s special Thanksgiving ep of Hard Fork, mostly because I finally got to talk about ✨ aqueducts ✨
151433
Reposted by Kevin Roose
Will Oremus @willoremus.com · 27/11/2024
Musk admitted X throttles links, and Threads is only a little better. But social media's quiet war on links has been going on for years. I wrote about how big tech has been draining the open web's lifeblood and giving us "news influencers" instead of the news. www.washingtonpost.com/politics/202...
washingtonpost.com
Analysis | Musk admits X throttles links as ‘news influencers’ take over
Hard times for the humble hyperlink and those who depend on it.
9738801036
Kevin Roose @kevinroose.com · 26/11/2024
And only 34 million were Instagram users who accidentally clicked on a dark pattern button while trying to buy a cabbage peeler
1026620
Reposted by Kevin Roose
Erin Griffith @ering.bsky.social · 21/11/2024
Pretty shocking to me that Instagram is encouraging users to chat with an AI “therapist” Especially after a character.ai chatbot has been accused of encouraging a teen’s suicide
57517
Kevin Roose @kevinroose.com · 26/11/2024
Competition is good! It’s been a while since Meta has had a competitor it couldn’t buy or copy (Threads can feature-match around the edges, but the big apps can’t decentralize in any real way)
111639
Kevin Roose @kevinroose.com · 26/11/2024
Both of these are better than Zoom's original name which was (not making this up) "Saasbee"
5842
Kevin Roose @kevinroose.com · 25/11/2024
nothing has ever made it to my bookmark bar faster!
5949
Kevin Roose @kevinroose.com · 25/11/2024
The displacement of entry-level programming jobs by AI is pretty clearly happening, even though it's not showing up (much) in economic data yet. www.nytimes.com/2024/11/24/b...
nytimes.com
Do Coding Boot Camps Make Sense in an A.I. World?
Coding boot camps once looked like the golden ticket to an economically secure future. But as that promise fades, what should you do? Keep learning, until further notice.
78915
Reposted by Kevin Roose
Gwen Snyder is uncivil @gwensnyder.bsky.social · 22/11/2024
TIL that the CEO of Bluesky is legally named Bluesky
Lantian "Jay" Graber (born 1991) is an American software engineer who has served since 2021 as the CEO of Bluesky, a microblogging social platform created in 2019 by Twitter's Jack Dorsey.Lantian Graber was born in 1991 in Tulsa, Oklahoma, the daughter of an acupuncturist mother of Chinese descent and a math teacher father of Swiss descent.[1][2] Her mother grew up in China during the Cultural Revolution and emigrated in the 1980s, and named her "Lantian," meaning "blue sky" in Mandarin Chinese, as a wish for her to have "boundless freedom".[1] The similarity between Graber's given name and Bluesky is purely coincidental; Jack Dorsey had chosen the name "Bluesky" for the research initiative before Graber became involved with the project.[3]
201054194
Kevin Roose @kevinroose.com · 23/11/2024
I get the argument that journalists still need to be on X/Threads, but boy is it refreshing to spend time on a platform whose leaders do not overtly hate you
2755643
Kevin Roose @kevinroose.com · 22/11/2024
Gotten into poker recently. Ready to graduate from playing online to a (small-stakes) home game. Who’s got a good one in SF/East Bay?
16321
Kevin Roose @kevinroose.com · 22/11/2024
It’s Hard Fork Friday! This week: 👮‍♂️ Podcast nemesis @reckless.bsky.social joins to talk about Trump’s FCC pick 🩻 Dr. @adamrodmanmd.bsky.social talks about his study finding that ChatGPT beats human doctors at medical diagnosis 🚨 And the Hard Fork Crime Division www.nytimes.com/2024/11/22/p...
nytimes.com
Trump’s Next Online Speech Cop + Doctors vs. ChatGPT + Hard Fork Crimes Division
“I feel very strongly that the First Amendment is under the most direct threat that any of us will ever really experience.”
61014
Kevin Roose @kevinroose.com · 22/11/2024
I wrote about this app! It’s good! (But you know that, you’re here) www.nytimes.com/2024/11/22/t...
nytimes.com
Bluesky Is Turning Into a Strong X Alternative
Bluesky has a hint of the old Twitter magic, but the feeling of freedom it offers might be even better.
771823300
Kevin Roose @kevinroose.com · 21/11/2024
just did this too! took 2 minutes. no idea what happened on the back end but it's fun to have a custom domain as a handle.
8581
Kevin Roose @kevinroose.com · 19/11/2024
This is a fascinating study, and speaks to the vast perception gap between what LLMs are capable of and what people think they are capable of. Doctors using ChatGPT for diagnosis performed *worse* than ChatGPT alone, because they kept second-guessing the model. www.nytimes.com/2024/11/17/h...
nytimes.com
A.I. Chatbots Defeated Doctors at Diagnosing Illness
A small study found ChatGPT outdid human physicians when assessing medical case histories, even when those doctors were using a chatbot.
1529436
Kevin Roose @kevinroose.com · 16/11/2024
It’s Hard Fork Friday! This week on the show: 🤑 Crypto is so back! 🤖 We talk to ChatGPT head of product Nick Turley about his MOST INSANE CHATGPT HACKS (and other stuff) 🦋 And we answer a listener’s pressing question: what social network should I be on? www.nytimes.com/2024/11/15/p...
nytimes.com
Crypto Congress + HBD ChatGPT + What Social Media Platform Should I Be On?
Cryptocurrency is poised to have a huge year in 2025.
131535
Kevin Roose @kevinroose.com · 13/11/2024
Just used Claude to make a "long video summarizer" app I've been dreaming about for a year (it converts videos to mp3s, transcribes them using Whisper, and uses Llama 3 to summarize the transcripts) in 20 minutes without writing a line of code, feeling absolutely drunk with power
202282
Reposted by Kevin Roose
Casey Newton @caseynewton.bsky.social · 05/05/2023
It's Hard Fork Friday! This week on the show — can Bluesky go the distance?? PLUS: how ChatGPT wrecked Chegg, and your AI ethics dilemmas answered www.nytimes.com/2023/05/05/podcasts…
nytimes.com
13719
Kevin Roose @kevinroose.com · 02/05/2023
studying up on decentralized/federated social media to talk about it with @caseynewton.bsky.social on the pod tomorrow. what should i read/who should i talk to?
8462