We can’t wait to share their writing with you.

Kevin Wei

I am currently a fellow in RAND's Technology and Security Policy Center. My research agenda is centered on the science of AI evals and related AI law/governance questions, with recent work published in top AI and AI ethics conferences. As an Asterisk Fellow, I'm mostly interested in exploring questions around how we can make evals more rigorous and more useful for assessment of downstream societal implications, translating technical concepts for the policy community, and the politics of AI governance.

Previously, I was a Schwarzman Scholar in Beijing (where I researched Chinese AI policy), a project manager / data scientist at cloud infrastructure provider DigitalOcean, and an organizer on Democratic political campaigns. I'm also wrapping up a J.D. at Harvard Law School and completed a master's in ML from Georgia Tech. You can find me on social media @kevinlwei or at kevinlwei.com.

John Vastola

I'm John J. Vastola, an academic researcher that studies natural and artificial intelligence. I received a PhD in physics from Vanderbilt University in 2021, and since then have been a postdoctoral research fellow at Harvard University. I'm particularly excited to write about the technical ideas behind modern AI systems, since I think you can't truly understand something without knowing how it works. I'm also interested in writing about the interplay between neuroscience and AI, and the extent to which learning more about our brains can help us build more capable machines. I write at Theories of Intelligence.

Larissa Schiavo

I work with Eleos AI Research; previously I worked at OpenAI, MURAL, and on various videogames. I write about AI welfare with an evidence-based and level-headed perspective. People say I’m funny. You can read my work at larissaschiavo.substack.com, and follow me on Twitter (@lfschiavo)

Buck Shlegeris

I'm the CEO of Redwood Research, a nonprofit that studies the risks posed by misaligned general-purpose AI, especially AI agents that conspire against their developer. To a greater extent than is reasonable, I like to draw inspiration from many different fields for my work. In this fellowship, I'm interested in exploring questions about AI misalignment risk that are of more general interest than my usual fare, and maybe I'll go even further afield than that. I usually write on the Redwood Research blog or on LessWrong.

Zilan Qian

Hi I am Zilan. I research and write about AI and geopolitics, focusing on overlooked (and fun) aspects of the U.S.–China competition, including humanoid robotics and access to banned U.S. AI models in China. For this fellowship, I hope to write more on the comparative dynamics of U.S. and Chinese AI development, including differences in the AI companion market and the geopolitical contexts.

I’m currently a research assistant at the Oxford Internet Institute, researching U.S.–China open-source governance, and a communications intern at the Safe AI Forum. I’m finishing my master’s in the Social Science of the Internet at Oxford, and I previously studied anthropology at Columbia University. I write on my Substack.

Max Nadeau

I'm Max Nadeau. I'm interested in how the abilities and usage of AI is evolving; I view it as trending toward more autonomy, reliance, and trust. I plan to blog about this and related issues during my time as a fellow — I’m very grateful to Asterisk for supporting my writing! In my day job, I'm a grantmaker on the Technical AI Safety team at Open Philanthropy. I help support machine learning research on understanding and controlling AI models, especially research that aims to scale to future, more capable AI. I've worked on grants for benchmarks of consequential AI capabilities, methods for detecting reward hacking, and techniques for finding and fixing rare failure modes in AI agents. Note: Open Phil has supported Asterisk in the past, but I was not involved in that process. The views expressed in my blogging here are my own and do not necessarily represent those of Open Phil. I write on Substack.

Richard Moulange

Hi! I'm Richard and I'm the AI–Biosecurity Policy Manager at the Centre for Long-Term Resilience, where I lead work on helping mitigate biological risks arising from advanced AI systems and understanding how we can use AI to better deter and defend against biological threats. I'm also a current Fellow with the Johns Hopkins Center for Health Security's Emerging Leaders in Biosecurity Initiative.

I recently finished a biostatistical machine learning PhD at the University of Cambridge, studying how to use deep learning to better diagnose and treat cancers using transcriptomic data. I've previously worked for the UK Civil Service and for GlaxoSmithKline R&D, as well as completing a GovAI Summer Fellowship and managing Fellows on the ERA:AI programme. My spare time mostly involves doing Park Runs and brunch with friends, playing DnD and making too many niche pop culture references.

I blog at richardmoulange.substack.com where I write about AI, biosecurity, national security and their (increasingly important!) intersections.

Michelle Ma

I'm Michelle Ma. I recently graduated with a B.A. in Economics from UChicago, where I worked on AI governance research at the UChicago Existential Risk Laboratory. Since policy is highly time-sensitive, I'm interested in examining key cruxes around the speed & nature of AI progress, focusing on practical bottlenecks that could double as governance opportunities (e.g. data, sensors). Beyond that, I'd also like to map out the robustness of AI policies under different worldviews, as well as develop a Republican perspective on AI governance. I write on Substack, mostly about AI. I'm also fascinated by progress studies, especially pharmaceutical innovation - I recently wrote a piece for Works in Progress on Journavx, a new non-opioid painkiller. Outside of writing & research, I like to run, play board games, and dance (flail around in Zumba class).

Joseph Levine

I am a PhD student at the University of Oxford, where I research finance and politics in West Africa. I've previously forecasted technological progress, designed financing for satellites and rockets, and driven ambulances. During the Fellowship, I'll be writing about AI adoption in Africa, how policymakers interact with AI tools, and the fiscal effects of AI on very poor countries. I write at https://jablevine.com/writing/.

Josh Holder

Hi, I’m Josh Holder! I graduated summa cum laude with a degree in Mechanical Engineering from Rice University, and went on to do a masters at the University of Washington, where I published research on reinforcement learning and optimization. I now work as an aerospace robotics engineer on safety critical systems in one of the most rigorous development environments on Earth. As such, I’m especially interested in how we can port some of the established safety frameworks from the aerospace industry to learning-based systems—and in turn, how automated reasoners might strengthen the safety of our existing systems. Outside of work, I'm an aspiring bird-watcher and love playing card games with friends.

Anson Ho

Hi! I'm Anson, I'm a researcher at Epoch AI, where I study the mechanics of AI progress and its societal impacts. I'm interested in understanding the bottlenecks to further AI scaling, how to better track AI progress, and what the next few years of AI developments will look like. I write on my substack.

Nate Fairbank

My name is Nate, and I’m interested in how AI works in the rain. Seriously. As an active-duty Army officer I currently command a fighting formation of 120 Soldiers in the Pacific. However, my previous assignment was to Pittsburgh, PA, where I earned two graduate degrees from Carnegie Mellon University as an Army Artificial Intelligence Scholar before serving as a Data Engineer at the Artificial Intelligence Integration Center (AI2C). At the AI2C I founded CamoGPT, a secure LLM chat client used tens of thousand of military members daily. I also prototyped HandMic, a “Siri for combat” that aims to streamline radio requests for Soldiers on the frontlines. I am deeply interested in the intersection of technology and tactics, and how they shape one another. When I’m not working I’m typically reading or running- I’m an avid endurance athlete and have run the Boston Marathon, an Ironman, and most recently a 30 mile Spartan Race. I write Edge Conditions.

Emma Erickson

My name is Emma Erickson, I’m a machine learning research scientist in the CIPHER lab, a cybersecurity lab at Georgia Tech Research Institute in Atlanta. Previously I was at the Auton Lab at Carnegie Mellon University, where my research focused on applied machine learning in specialized fields such as cardiology and predictive maintenance. I am looking forward to writing about AI in practice, particularly the interactions with domain-specific cultures which impact its uptake. I write on my Substack.

Lawrence Chan

I work at METR doing evaluations of large language models. Previously, I did research on mechanistic interpretability and adversarial training at Redwood Research. I was also a PhD student at UC Berkeley with the Center for Human- Compatible AI and InterACT Lab. I hold degrees in Computer Science and Business from the University of Pennsylvania, where I did research with the Good Judgment Project on using ML to assist forecasting. I write on my Substack.

Vince Vatter

I'm Vince Vatter, a professor of mathematics at the University of Florida, with a research background in combinatorics and graph theory. Recently, I've been interested in the intersection of math and AI, both in terms of how AI can help us to learn and do mathematics, and how AI does mathematics itself. In particular, I lead an NSF-funded project exploring how AI can help students learn to write mathematical proofs (https://hallmos.com/).

In this fellowship, I'm excited to explore what it really means for AI to "do" mathematics, whether current approaches represent genuine mathematical understanding or something fundamentally different, and what advances in AI mathematical capability might mean for the future of mathematics.

Ricky Huang

My name is Ricky Huang and I’m currently running an EdTech company in Dubai. Prior to that I dropped out of a Philosophy PhD program focused on Epistemology and Deep Uncertainty. I previously did China AI safety fieldbuilding funded by EAIF and worked as a research affiliate at Concordia’s China AI Governance Team. I write on my Substack.

We’ve extended invitations to 18 of those 900 (we initially planned for 15, but this was way too difficult), and we’ll be announcing our Fellows next week.

Until then: Sheer numbers prevent us from responding to each person individually, but we wanted to share some of the themes we saw, and explain how and why we made our selections.

We spent a lot of time — some 70-80 hours collectively — reviewing submissions. At least two of us read each application, and all three of us scored a long list of just under 200. There were many more people we would have gladly admitted into the program than we had room to accept.

Above all, we hope you use this as words of encouragement. We received some fantastic pitches (and we’ll be reaching out to some of you who weren’t accepted to see if you’re interested in writing a piece anyway). The entire purpose of this program is to get more people writing in public, and we still want to nudge more of you to do so, fellowship or not.

So here’s who applied:

In total, we received 913 applications. 120 (13%) had PhDs or are in PhD programs. There were a couple dozen tenured professors, many more folks with masters degrees, and at least one person still in high school! Applicants were heavily concentrated in the US (48%) and Europe (36%), but we were encouraged by the 8% who applied from Asia and 5% in Africa. (Note to selves: Do more outreach in Australia?)

Nearly a third of applicants work directly in ML or AI, either on the technical side (we had representation from every major lab, and many small ones) or in research. Our second largest pool worked in policy and governance. None of this really does justice to the breadth of backgrounds, though. We received applications from chemists, sexologists, hedge fund managers, neuroscientists, librarians, returned peace corps volunteers, journalists who cover cults, and even a few folks with — how do we say — sui generis theories of LLMs.

What we selected for

First, we emphasize again: We received many more qualified, suitable, and wonderful applicants than we had the space to admit. If you applied, thank you for taking the time. We’re really honored.

For several dozen of you, non-selection came down to either 1) our desire to have a cohort that varied in terms of expertise and experience, or 2) the difference of half a point on our grading criteria.1 We’ll be sharing the backgrounds of admitted fellows next week, but here are a few of the common attributes that really excited us.

A unique perspective on some aspect of AI, and an eye to write about it. With 900 applicants, some pitches got repetitive, variations on a What is a transformer? Is AI conscious? How can we regulate AI? theme.

We were most excited by applicants who pitched specific stories on areas of AI that they know well, and that we considered under-covered in the blogging sphere: what AI investment funds are doing in the Middle East, how LLMs are being used in African judicial systems, or why monitoring via chain of thought thinking is likely to fail.

Many of you we considered to have a unique vantage on AI. You described in your backgrounds how you led AI integration in your workplace, or how you were in charge of AI policy for a company or government, and we grew excited as we read those descriptions! But then your pitches were not connected to your actual work. We valued confluence between your story ideas and experience very highly.

Excitement and curiosity. An admittedly fuzzy distinction between the pitches that popped and those that were simply competently good we would best describe as “a story that you want to write” vs. “a story that you think should exist.” These obviously aren’t mutually incompatible, but in our experience, the former produces more engaged writing. It’s something we could sense in the curiosity that a writer expressed. Think: “There’s this thing I’ve been trying to work out about how LLMs solve math problems, and here’s what I think it is,” vs. “No one to my mind has written the perfect 2000 word piece on what a transformer does.” A related failure mode was when we read a pitch that seemed like it was written with the idea of “something Asterisk’s editors would like.” We like to be surprised.

A voice. We think that many aspects of writing are skill issues — hence this fellowship. However, it is an uphill battle to develop a distinctive voice, and a common feature between everyone on our shortlist is that they were able to make themselves stand out despite the brevity of the application. Some people did this via style, others through very crisp articulations of their perspectives, others through having an angle we’ve never heard before, and others through some riskier answers.

A mixture of perspectives. About half of the cohort we selected are aligned with safety research in some form (this is the magazine shaped by the philosophy of Effective Altruism), but we mean it when we say are open to all perspectives. We selected a mixture of writers because we believe that to succeed in blogging for a wide audience, it helps to understand a multitude of perspectives. AI discourse seems to slough its skin every 6 months or so, acquiring deeper nuance and creating new coalitions each time it does. We selected for writers who we think have an eye towards that process, either because they’re privy to the research driving that process, or paying attention to those who are.

What we selected against

We rejected some applicants because they were over-qualified. Writing is never easy, and so we fully appreciate the value of a writing community and accountability system for anyone.2 But if you have, for example, published a book, already have a well-subscribed blog, maintain a large Twitter following, or were mentioned as one of the intellectual inspirations that other applicants listed in their ‘Why’ sections (this happened at least three times), we thought we’d probably be less counterfactually helpful to you.

This was a pretty unique applicant pool. At an acceptance rate of less than 2% it took a lot to stand out. The most common reason we gave someone with a highly qualified background — and we are turning down academics with high h-indexes and research leads at large labs here — a low score is that we just weren’t excited about their pitches. Admittedly we can afford to be selective here, but we think this is useful information for those of you with substantial credentials who want to write more in the future: editors care as much if not more about the story than your background.

We received a number of submissions from researchers or established journalists interested in learning and writing more about AI. We chose not to accept these applicants because experience, not writing, is your bottleneck. Our best advice here is simply to start: AI is ubiquitous in the news, but there are still less than a couple dozen journalists on this beat full-time. There are many stories to be told, low-hanging fruit everywhere for those with eyes to see! You, journalist or researcher, do not need Asterisk’s imprimatur to start!

Last, while we do appreciate the irony, we unfortunately decided to turn down all obviously LLM written applications. We don’t mind if you use Claude for the copyediting & clarity assist, if the resulting application comes out really good. But there were some applications that pitched a boilerplate story that seemed eerily close to the pitches we got from GPT when we asked it to give us ideas at the intersection of AI and [your field of choice.]

Some reflections for aspiring writers

One of the most common reasons people cited for applying is that they feel held back by perfectionism, or imposter syndrome. As editors who tend to belabor a sentence, we empathize. Our best advice here is to ignore all your instincts. The internet is big, and in our experience, earnest efforts attract earnest feedback and like-minded collaborators. A blog post is never finished, only published. So this is us giving you permission. Hit submit. Just get it out there! Seriously, do it — many of you told us you have multiple pieces sitting in your drafts folder. What value do they have there? At least let Claude see them. And send them to us too — we’ll read it, even if we don’t have time to respond.

We think more people should write to your strengths. This is really a requirement for starting. The more junior you are, the more you need to narrow your scope: What do you know particularly well that no one else does? Can a CS major just out of undergrad write a transformative piece on how LLMs work, or the definitive blog post on the AI race between the US and China? We’re not saying no, but we think you’ll write a better blog post explaining the niche research project you’re working on.

The impetus of this fellowship was that there are lots of interesting conversations about AI happening offline — at lunch, at parties, on your anon discord server — that deserve to move into public view. The depth of applications we received made this clear. We’re extremely excited to start bringing some of this discussion to you on this page, Asterisk’s new AI vertical. Over the next few months, we’ll be featuring some of the best posts from our fellows here. To read them, subscribe below.

But we mean it when we say you should write the piece without us.

Thanks for reading. Thank you so much for applying. Check back here soon for more.

– Jake, Avital, and Clara

In-person conversations at AI hubs or co-working spaces are often deeper, richer, and more thoughtful than what’s available online. Important conversations don't make it onto the internet. That's because writing is hard. We'd like to make it easier.

The fellowship lasts for six weeks, from August 24-October 6. Over the course of the program, you will produce three blog posts. The fellowship is remote and compatible with a full-time job. We expect about 10 hours of commitment per week, mostly writing, but also participating in weekly workshops, and providing feedback to peers. During that time, we'll provide:

• Editorial Support: Work on your writing with Asterisk's editors as well as a full-time writing coach.

• Peer Feedback: Discuss your latest piece in weekly peer workshopping sessions and in our group slack.

• Mentorship: Fellows will have access to a network of mentors with extensive experience writing about AI for public audiences, including

  • Scott Alexander (Astral Codex Ten)

  • Jordan Schneider (ChinaTalk)

  • Dean Ball (Hyperdimensional/White House OSTP)

  • Sam Bowman (Anthropic)

  • Tim Lee (Understanding AI)

• A Platform: In addition to writing for your own blogs, your work will be cross-posted to Asterisk's new AI vertical: asteriskmag.ai. We'll work with fellows to help them develop their audience and get comfortable writing in public.

• A Stipend: $1000. (Self-explanatory)

• An Excuse: Do you need a good reason to prioritize sitting down and trying to write? This is it.

Fortunately, there's a lot of good AI writing on the internet, but we think there's a need for much, much more. Here are some areas we think are particularly under-covered: In-depth, qualitative evaluations of what AIs can and can’t do, and how they’re really being used on the ground in different fields. Accessible explanations and assessments of technical AI safety research. Explorations of what good futures with advanced AI might look like, how we could get there, and the most likely reasons we might not. We’re also especially interested in writers who can bridge the inferential gap between various AI camps with different fundamental assumptions. All that said, we're certainly not limited to fellows who want to write about these topics — they're suggestions, not requirements.

Apply here. It should take about 20 minutes. We'll close applications on August 1st, and notify those selected by August 15th. We look forward to hearing from you!

And if you’d like to receive nuanced perspectives on AI from our newest vertical, sign up below.

Subscribe now