More from What To Cook
For the grand finale episode of this season of the What To Cook show, I bring you a fajita feast! We’re headed to my backyard to grill up the most delicious summer spread of hanger steak, shrimp, and veg for fajitas, plus we’ll make a creamy charred jalapeño sauce and corn on the cob to finish off the meal. In this episode, you’ll learn about… * Grilling over direct vs. indirect heat * Using your intuition when you’re grilling a bunch of food that has different cook times * Cutting against the grain to make steak SO much easier to chew * How to prep all sorts of peppers * Much more!! I absolutely loved filming this first season and getting to SHOW you many of the things I tell you about all the time in the newsletter. Also to show you that my kids are in and out of my kitchen at all times trying to introduce me to lizards and climb on the counter and crawl at my feet. If you have kids, your cooking process likely looks similar, and I want you to know that we’re in it together! Have any of you become Cometeer diehards like me?! The fact that I can make myself a coffee shop-grade cup of coffee or iced latte at home in about 15 seconds flat has TRULY improved my life. Cometeer is what to make when you don’t feel like making coffee (or spending money at a coffee shop on a daily basis), and if you appreciate the ease of WTC recipes, I know you will love the ease of their product, too. THANK YOU, Cometeer, for supporting this show, and for fueling me with your delicious coffee every day! If you haven’t tried it yet, you must. Remember to go to cometeer.com/caro to get $20 off your first order! OK, that’s a wrap on this season of the show! Thank you SO much for watching. I’d love any feedback you may have. Did you like the longish-form format or would a shorter show be easier to watch? Would you be interested in watching more entertaining-style content, or anything else!? As always, I’ll be back in your inbox tomorrow with the recipe itself, including a bunch of notes and substitutions. Happy long weekend! xx Caro Click here to watch this episode on YouTube! This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit whattocook.substack.com/subscribe
Hi! Woah. I leave for the East Coast to start my Make It Fast (please for the love of ME, pre-order it!!!) book tour in less than 48 hours. My literary agent just called to check in on my mental state and I’ll tell you what I told her— I’m honestly feeling great. I haven’t started packing, but I’ve got a million outfit options strewn about my closet, and I’ll get there eventually. The flights are booked and the hotel reservations are made. It’s go time, baby. I’m also thrilled to announce that I have officially entered phase 5 of the creative process: With my last book, I was firmly still in phase #4 (“I suck”) until the book had actually been published and all of you started to cook from it and share your photos and reviews and love of the book. But with Make It Fast, I hosted a party and gave the book to my friends and family two weeks before pub day! I created an Illegal Bookclub and roped in my friends all across the country to share advanced copies of the book with their communities! And, comically, Indigo (the main Canadian seller of the book) accidentally sent all of the pre-orders like two weeks early, so all of the Canadian What To Cookers out there have had their books for a week now! And Canadians are so nice — they are sending so much positivity my way! So there are a LOT of people out in the world, people who I love and trust, who have already seen it, and cooked from it, and told me how much they love it. Thus launching me into phase #5 (“this might be ok”) a few weeks earlier than my last book publishing cycle. I guess, if I’m being honest, what I’m saying is that I require external validation to feel good about my work? Is that pathetic? I don’t think so. I think it’s a real, honest part of most (all?) people’s creative process. It feels really scary and vulnerable to launch a piece of work that you’ve worked so hard on out into the world. It feels a little bit like dropping your deeply feeling, sensitive, absolutely incredible child off at Kindergarten for the first day of school. Will he make friends? Will the teachers understand how creative and kind and cool he is? Will they be patient with him and let him finish his art project because he really doesn’t like when he doesn’t get to finish his art project? Will the world love him as much as I do? This analogy is not random: I am dropping my deeply feeling sensitive child off at Kindergarten for his first day of school tomorrow, and the jitters and unease of that feel so akin to publishing a book. By next Wednesday when I send this newsletter, Make It Fast will be in your kitchens, if you’ve pre-ordered (have I mentioned I’d really love it if you could pre-order?). I’ll have already been on Good Morning America two mornings in a row (bless them, what an invitation!), spoken with Jenna Bush Hager in front of an audience of you in Brooklyn, yapped with Dan Pelosi in Boston, chatted with Kerry Diamond in DC, and I’ll be en route to Charlotte where I’ll be in conversation with our very own Molly from What To Cook. I am excited, and I am nervous, and just like I’ve been telling my five-year-old, it’s OK to feel both of those things at once. And Phase 6 (“this is awesome”) is just around the corner for both of us. There are lots of other book tour dates — check them all out here! Today’s So Into That is sponsored by FP Movement. We’ve talked about our love for their Carpe Diem shorts many times before (a truly perfect athletic short at a very good price point!), but FP Movement has many more athleisure pieces we love. From Molly from What To Cook: I’m on vacation at the beach this week and have been living in this dress and shorts combo when I’m not in a bathing suit. It’s so so comfy and easy to throw on for beach walks — it’ll also be great to wear on the sports sidelines when it’s 100 degrees outside. I’m a shortie (5’3”) and had to shorten the straps a tiny bit (worth the effort!). We also love these biker shorts with this boxy short sleeve and this oversized long sleeve! * Dr. Becky invited me to plan the menu for a REALLY special dinner celebrating Good Inside Baby on Monday in LA. I got to hang with MALLY K. GOLDMAN and Megan Roup and Elaine Welteroth and so many incredible, smart women who make motherhood safer, more fun, stronger, and more fashionable. The dilly feta salad with feta brined shrimp, flank steak with elote salad, and bbq chicken are all straight from the book! My menu concept was elegant California BBQ and I think we nailed it! PS If you’re in LA, Heirloom catered and they rule. PS because I know you all want to know, Becky is awesome IRL. She is a fellow insomniac romance book lover, she values her female friendships so highly and they come up in conversation all the time (this is always a green flag for me when I meet new people), and she’s just so incredibly smart, insightful, and cool. This is not sponsored but we’ve swapped codes for each other in the past and my Good Inside code still works — you can use code WHATTOCOOK to get 20% off a subscription to Good Inside if you could use a little parenting assist (like with how to get your deeply feeling kid to dread the first day of school less!). * I recently worked with Groundwork to hire a new team member (woohooooo, Meg!) and it was such an incredible experience that I wanted to share. Three in four U.S. workers are also caregivers, and as that caregiving work collides with mid-career years, most end up forced to choose between staying all in at real personal cost or stepping away entirely — becoming a hugely qualified, "off-market" talent pool that's actually eager to work, just on slightly different terms. Groundwork connects people like that with employers building smarter, more flexible teams. In my case, I knew I needed help in a certain space but had no idea how to define the role or find the right person. They sat down with me, helped flesh it out, wrote the job description, sourced candidates, and ran interviews, and took it all off my plate. And brought me the most INCREDIBLE CANDIDATES! Like lights out brilliant women. If you're a small business owner stuck in hiring limbo, or you're job hunting and sick of the LinkedIn black hole, Groundwork is worth a look. * From Molly: A go-to breakfast of mine this summer has been avocado toast inspired by some Caro made for our team last time I was in Carmel. Mashed avocado seasoned with salt and pepper, slathered on a thick slice of GOOD toasted sourdough, and topped with chili crisp, pickled red onions, flaky sea salt, and fresh cilantro. So good. * From Molly: I’ve been slathering on this body oil after time in the sun this week — it’s very luxurious feeling and makes my skin glowy! From Caro: Oh I love this too!!! It really does not make you feel oily at all, it absorbs so nicely. My publishers are dropping a few more advance reader copies of Make It Fast in the mail to some of my media/Internet friends this week, and I told them to save some for WTCers!! We are going to mail three of you a copy of Make It Fast, along with a Butter Pants spatula and a Caraway kitchen timer! To enter to win, comment on this post with the meal you cook most often when you absolutely do not feel like cooking. It doesn’t have to be from scratch! For example, mine has historically been frozen dumplings and peanut sauce, and Molly’s is pasta with Rao’s marinara and frozen meatballs. Comment between now and August 10 — we will randomly select three winners next Tuesday, August 11, and alert them by Substack DM! Every week, we dig into the What To Cook archives to see what we were cooking this time in years past — the recipes worth bringing back into your rotation. Here’s this week’s lineup. 1. salmorejo, 2025 Oooh, this is such a good one. Traditional salmorejo! From José Andrés! If you have boatloads of ripe tomatoes that need to be used up, make this creamy chilled tomato soup. It gets its creaminess in part from blended artisanal bread, so if you have any leftover sourdough from this week’s summer shrimp panzanella, use it here! 2. skillet cheddar turkey burgers, 2024 Juicy, cheesy turkey burgers seared in a skillet and served with a tangy, punchy burger sauce — this recipe (which is an excerpt from What to Cook the book #1!) is just so good. Go classic with buns, or shake things up and make sliders for a party, meatballs for dipping, or crumbled patties in pita for lunch the next day. 3. turkey meatballs with corn butter rice, 2023 These juicy turkey meatballs are packed with flavor, but the real star of the show is the buttery corn-studded rice. It’s sweet, savory, summer perfection. While you’re cooking, double the meatballs and freeze half for later — future you will be thrilled. 4. healthy (or not as healthy…) corn chowder, 2022 Sweet corn gets blended into a creamy, cozy chowder that tastes like summer in a bowl. Keep it light with coconut milk, or go richer with heavy cream — both versions are equally delicious. Make it on a Sunday and thank yourself all week — it reheats beautifully and somehow tastes even better on day two. 5. sheet-pan crispy pork chops with fennel and peaches, 2021 These roasted peaches are iconic — soft, jammy, and caramelized at the edges. Add in sweet, roasty red onions and fennel, plus golden, crispy pork chops, and you’ve got an easy dinner that feels delightfully fancy. The method works just as well with chicken breasts or sausages if that’s what you’ve got. We use affiliate links when sharing product recs, which means WTC may earn a small commission off of purchases you make through those links. This does not cost you anything extra and is a way to support the production of and team behind What To Cook. Sponsorships are another way we partner with brands that we authentically love and can personally vouch for. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit whattocook.substack.com/subscribe
I mentioned earlier this summer that I have been working on something BIG and it’s finally time to share what that something is: I’m launching a show here on Substack!!! What To Cook is (for now!) an eight-episode series that takes you inside our home, right into the dinnertime fray. In each episode, I race against the clock to prove that my recipes truly take the amount of time I say they do, even when your kids are acting absolutely bananas. It’s chaos cooking at its finest. Every single recipe in this first “season” takes 30 minutes or less of active cook time, just like in my upcoming book, Make It Fast. On Fridays I’ll share the new episode, then — like always! — the recipe will hit your inbox Saturday morning. I had so much fun making this for all of you, and I can’t wait to share the first episode with you tomorrow!! xx Caro PS, I got to work with an incredible team on this and I’d love to take a moment to thank them here: * My Substack darlings: Dan Stone, Hanne Winarsky, Michele Dale, Ariel Swedback, Nish Bhansali, Hamish, Jai and Chris, and all of the other Substackers behind the scenes, for believing in me and What To Cook and helping me make cool things for 5+ years now! * A New Media: Jana and Chitty, for gathering the most wonderful team of individuals to make this thing happen. * Jake and Meredith (and their son, Gamble, who attended one day of the shoot and my children adopted): My video + editing powerhouse. Adore you guys, thank you for making magic out of complete chaos. * Jess Mills: My dear friend who said yes when I asked her to be the food producer having never ever done anything like this before (she is an incredible artist, in fact!) — thank you, thank you, thank you for being a part of this and sorry I made you be onscreen so much and didn’t tell you that was a part of the gig hehe. * Zach: Your video skills rule but your willingness to do dishes and play with my children cemented your place in my heart! Do you still have the friendship bracelet Calum made you? * Jillian and Molly and Adam: For doing all of the things to make this crazy train not fall off the tracks. * Cometeer team: For answering the call when I said I’m obsessed with your coffee, you are the coffee equivalent of my recipes, so want to sponsor a show about cooking in the midst of utter chaos?! I’m sipping an Intelligentsia French Roast right this minute :) Every episode of What To Cook will star the complete-meal dinner recipe I’ll be sending to paid subscribers the following morning. If you want access to all eight recipes I’ll be showcasing in this series, upgrade to a paid or Insiders subscription here! Let’s cook together — one chaotic kitchen to another ❤️ This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit whattocook.substack.com/subscribe
Thanks to everyone who joined for today’s Substack live cooking demo — if you couldn’t make it, find a full replay above! In the demo, I cooked smoky mole chicken enchiladas with lime crema, a recipe that turns a few humble pantry staples (peanut butter, chocolate chips, tortillas, tomato paste, spices, and chipotles en adobo!) into a delicious weeknight dinner. The only fresh ingredients you’ll need are chicken breasts, white onion, a red bell pepper, some shredded cheese, sour cream, and a lime! Recipes that rely heavily on pantry staples are easy to make on nights when the fridge is mostly bare — they can also be more budget-friendly, since the cost-per-use of ingredients like peanut butter and spices is often quite low. Thank you to Thrive Market for sponsoring today’s live cook-along, which allowed us to make it free for all subscribers! Thank you also to Thrive for making it so very easy for me to keep my pantry stocked up with high-quality ingredients (at an excellent price point!). In addition to pantry staples, Thrive is my go-to source for kids snacks and clean home products — I love that everything they sell is free from more than 1,000 artificial, synthetic, and GMO ingredients. I also LOVE their app and use it to immediately add ingredients to my cart as soon as I finish them, so that I don’t have to write them down or try to remember later. Sign up here to get 30% off your first order (STOCK UP!) and a free $60 gift. Now’s a great time because everyone — new and existing members — can get 25% off many products right now during their Friends & Family sale, which runs through Thursday. Also FYI that new-to-Thrive What To Cook Insiders get $100 off their first four orders! While we’re talking pantry staples, here are a handful of ingredients that pack a TON of flavor that I always keep on hand. We’ve linked WTC recipes to cook with each! * Coconut milk — Coconut milk’s a great dairy-free sub for cream that you can use in curries, smoothies, and soups. In my house though, it’s most often used to make coconut rice! * Chipotles en adobo — We used these smoky, spicy peppers today for our mole chicken enchiladas! You can also use them for chipotle black bean enchiladas and cheesy chipotle shrimp tacos. Rarely does a recipe use the whole can, so store your leftover peppers and sauce in an airtight bag in the freezer. * Creamy peanut butter — My house is never without a jar! Again, we used it here in our mole sauce, but I also love mixing it with soy sauce, rice vinegar, and honey for a peanut sauce to serve on noodles or alongside potstickers. This peanut veggie curry is delightful, as are PB&Js, apples with peanut butter, smoothies bulked up with PB… the uses go on and on! * Curry paste — Curry paste packs SO MUCH flavor into just one spoonful! Keep your pantry (or fridge, once you’ve opened it) stocked with red curry paste (Mike’s is my favorite) and you can cook thai chicken chops, one-pot pumpkin chickpea curry, and thai tofu curry noodle soup! * Jarred marinara/tomato sauce — You could make a homemade marinara… or let someone do it for you! I’ve written many recipes with store-bought jarred marinara/tomato sauce listed as an ingredient because if you get a high-quality brand (my favorite is Rao’s), it really does taste homemade and will save you so much time and effort. Stock up on a jar to be able to cook 30-minute chicken Parm meatballs, messy meatball sliders, shortcut shakshuka, and 5-ingredient eggplant marinara. What flavor-bomb ingredients is your pantry never without?! For a complete rundown on what I consider to be a perfectly stocked pantry, click here. And for a fun take on this topic, read “If My Pantry Were to Disappear, Here’s What I’d Restock Immediately” — a Bon Appétit article written by Alaina Chou that I recently weighed in on! This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit whattocook.substack.com/subscribe
More in life
This is a weird moment. I have a hunch that we’ll look back later and tell ourselves stories in which it seemed obvious where things were going. So I wanted to lay down a little time capsule and register just how confusing everything felt at the time. This is a personal view from the trenches. 1. If you’re a normal person, you may feel that AI has been good at math for a long time.1 But for “real” math, this is roughly what it felt like: When Feeling 2024 LLMs are toys Summer 2025 LLMs are interesting toys Early Autumn 2025 LLMs can do useful stuff sometimes Mid Autumn 2025 …a lot of stuff Late Autumn 2025 …a lot Early 2026 holy god Even having been astonished by AI progress in the past, it was still astonishing to see AI move in a matter of months from a mildly-useful toy to something that many people now consider almost impossible to do without. 2. Personally, I’d say that current AI makes me around 50% more productive at “doing math”. That’s despite the fact that AI can now often do in a few minutes things that would have taken me hours to do myself. The main reason is that my concept of “doing math” involves various steps, like coming up with ideas for theorems, trying to prove them, writing things down, reorganizing everything to make it more clear, and so on. AI is good at some parts of this, but still pretty bad at others, so the overall speedup isn’t as high as you might expect. (Even if planes could go 10× faster, your door-to-door travel time from your apartment in London to your friend’s place in Rome wouldn’t decrease that much, because you still have to get to the airport and wait to board.) 3. But then, I’m still (sort of; more on this later) trying to follow my old “doing math” workflow, just automating certain parts of it. That can’t be right, can it? If we found a way to make batteries at 1% of the current cost, we wouldn’t just say, “cool, batteries are cheap now”. We’d find all sorts of new ways to use batteries that didn’t make sense before, and change the ways we generate power to account for all these new batteries. Probably our whole energy system would end up transformed. Analogously, AI being good at proofs should lead to a very different equilibrium. But what is that?2 4. It seems healthy, at the moment, to maintain a fluid view of what “doing math” means. But at bare minimum it surely involves understanding the math that “I” am “doing”. So long as that’s part of the workflow, it would seem to impose a limit on how quickly “I” could “do” math, no matter how good AI is. One obvious conclusion would be that my tiny organic brain should be removed from the workflow entirely. Maybe! But it’s interesting to speculate about how much faster things could go while still keeping a human brain in the loop. I think the answer is: A lot. 5. Contrary to what you may have heard, humans are not obsolete for math. Not yet. The most obvious reason is that AI is still bad at writing. People often complain that AI writing overuses various patterns or lacks a unique voice or whatever. I don’t disagree, but I always felt like there was a deeper issue that I couldn’t quite put my finger on. With math, this becomes very clear: The problem is it doesn’t make any goddamned sense. Really, I think it’s deeper than being bad “at writing”. It seems like AI is bad at thinking in certain ways that manifest as bad writing. 6. When novices do math, they often create “spaghetti-code math”: Imagine a long document with lots of equations, but if you point at any one equation, it’s hard to say what assumptions underpin it, or how it fits into the bigger picture. Any random idea that was mentioned 17 pages ago might be used again at any time with no warning. With experience, people learn to break complex arguments down into modular pieces with a good “API”, in the sense that the assumptions and conclusions for each part are relatively simple. Somehow, AI has developed these amazing technical skills, while still being bad at creating abstractions So the arguments it gives are long and tangled and bewildering. This is what makes AI math seem so alien. 7. Now I know what you’re thinking. Maybe AI isn’t bad at creating abstractions. It’s just bad at creating abstractions that work for you, human with tiny organic brain. From a distance, that seems plausible. And it might be true to some degree. I have high-bandwidth access to a human brain. It’s not crazy to imagine that this gives me an advantage in predicting how a human brain would react to a given presentation of ideas. But still, I claim that current AI is objectively bad at creating abstractions, regardless of if the consumer of those abstractions is human or artificial. Here’s why: If I ask AI to prove something hard, it will produce a proof that starts simple and gradually descends into a rampage of complexity. If I tell it to break that proof down into self-contained pieces, it will produce lemmas that each take a full page just to state. If I ask, “Why? Why have you done this to me?”, the AI will say, “This is as simple as it gets. It’s not that bad. What’s your problem?” But if I argue with the AI for a while, I can start to understand the basic contours of the argument. And I’ll ask, “Why don’t you first just try to prove blah using only assumptions blahblah, without worrying about blahblahblah? Wouldn’t that be easier?” Often, this doesn’t work, either because my idea was bad, or because the AI isn’t smart enough to do it. (Which one? Who knows!) But after a few attempts, I can usually find something that that works, and the AI can use it to re-write the proof in a somewhat simpler way. If I repeat this many times, I can gradually chip away at the complexity. But maybe that’s still just making the proof easier for humans to understand? No! For one, this usually makes the proof shorter, and sometimes dramatically shorter. More importantly, during this process, someone (maybe me, maybe the AI) will often realize that there’s a different proof strategy that gives a stronger result. Instead of proving x < 9.7, now we can prove x < 7.1. Thus, artisanal human abstractions are “objectively” useful. 8. Some people might claim that it’s already possible to automate some of what I’m doing with some multi-agent Rube Goldberg thing. I’m not sure, because my experience has been that investing lots of time into clever AGENTS.MD or custom skills or harnesses isn’t a great use of time. That stuff is somewhat helpful, but things are changing so fast that the knowledge has a short half-life. Also, the majority of information about these tricks seems to come from people who are excited about the technology itself. On the other hand, the automator told me to try typing, “ask independent agents for an adversarial review”, and I’ve found those seven words to be quite helpful. I wish I could find a source with boring best practices for people who are happy to stay away from the bleeding edge. 9. Another reason AI doesn’t improve my productivity more is that it is often the green light, the orgastic future that recedes before me. I’ll ask AI to do something, and it will almost seem to do it. So I’ll ask for changes, and it seems to help, but it’s still not quite right. Hours later, I’ve accomplished nothing and learned nothing. Eventually, I start over, either doing it myself or with AI on a very tight leash, and I’m done in 30 minutes. Each time this happens I feel foolish and tell myself not to do it again. But it’s very hard to judge in real-time. 10. Of course, I’m not claiming that AI is intrinsically bad at writing or thinking or abstraction-making. It’s gotten better at all of these and I expect it will continue to do so. All I’m saying is that as of Autumn 2026, humans still have a lot to contribute, and it’s not just better exposition. 11. Did AI actually get much better at math in Autumn 2025? Or perhaps, in the grand scheme of things, human math ability falls into a narrow range? 12. Some people are responding to the new reality by gleefully using AI to generate lots of semi-OK-ish papers and puking them up on arXiv. I understand the incentives that drive this. In some sense, it seems like the logical response to the new reality. (After inventing cars, we started taking a lot more low-value trips.) But I don’t think this is socially useful, and I tend to think everyone will soon agree that this is embarrassing. But… I’m not sure. Maybe it is socially useful? And I find myself feeling vaguely jealous of the people who are wired this way, who instead of feeling blinkered, see opportunity and plunder. 13. Want an optimistic pro-AI, pro-progress take? OK. If I take my old “doing math” workflow, AI makes it strictly better. A lot better. It’s like having a hyper-conscientious intern that is smarter than me in many ways, works at incredible speed, and knows approximately everything. Yes, AI sucks at writing and making abstractions. But if I’m disciplined and try to mostly do the math “myself” while asking the AI for help, I can explore ideas much faster than before, make far fewer mistakes, and produce something just as clear as before, while developing just as much understanding. I can ask, “What happens if you do a Kofuni expansion?”, and get an answer instantly. I can ask for illustrations or interactive visualizations and see them instantly. Arguably, we’re incredibly lucky to be alive at this moment, when these new tools have opened up this new territory to explore. 14. And isn’t that what I wanted? Back in 2024, I bemoaned that what AI most seemed to struggle with was getting the facts straight: Perhaps the old science fiction robots were a projection of our fantasies. I want my style and voice to be precious and uncopyable. I want an AI assistant that will tirelessly hunt down facts while I provide creativity and panache and magic. But that’s not how it is. So we should remember: In a scenario where AI is useful but humans retain some comparative advantage, we might not like where that comparative advantage lies. At least with math, and at least for the moment, I get to be Kirk, while the AI is Spock. (Even with other types of writing, my impression is that since 2024, AI has improved more along Spock dimensions than along Kirk dimensions.) So that’s nice. 15. In the last few months, it frequently happens that a collaborator will use AI to generate something, and then send it to me and ask me to check if it’s useful. I can see why this seems reasonable. (Really, I do.) But when it happens a lot, one can’t help but wonder what the point of the human collaborator is. 16. On the other hand—with creative writing, my advice is not to use AI to write, at all. Ask questions if you want, but write every damned word yourself.3 With math, I don’t feel like such a hard line makes sense. It seems that creating something good still requires a lot of human involvement, to the point of agonizing over notation, repeatedly re-organizing things, and heavily editing individual sentences. Still, I find that if people make that effort, they can create something just as good as they could by hand, while still being perhaps 50% AI written.4 So I don’t mind if people send me “centaur math”, provided they take responsibility for the contents. The problem is that when working with AI, it takes constant paranoid vigilance to avoid letting things slip beyond your understanding. I often find myself wanting to awkwardly ask people, “OK, you obviously used AI to make this… which is OK… but how much of this do you vouch for? If we have a meeting to discuss it, will you crumble the moment I start asking questions?” 17. When I’m in abstraction-creation mode, I feel like I’m doing something intelligent. When yell at AI about bad notation and writing, I feel like I’m providing taste. But I often find myself giving feedback that seems incredibly basic. Things like: This is a mess, clean it up. Or: Do you need that assumption? Or: If you can’t solve this problem, why don’t you create a simpler version and try to solve it first? On the one hand, I guess a lot of what good human managers do amounts to giving very basic feedback. On the other hand, it’s hard to imagine that I have much durable advantage in doing this kind of thing. So I wonder how much durable advantage I have in abstraction-making or taste, too. 18. A common reaction to AI getting better at proofs is that humans need to focus more on exposition. This has an obvious Bourdieu-esque criticism: Human painters used to strongly value photo-realism, but after cameras were invented they convinced everyone that, oops, photo-realism isn’t important, what matters is expression, which coincidentally cameras can do. Isn’t a shift towards exposition in math just a way for humans to move the goalposts and justify their own continued relevance? The only credible way I see to avoid that criticism seems to be to bite the bullet and say that, in fact, math should have already have focused more on exposition, even before AI. Personally, I strongly believe that this is true. 19. I think the biggest reason I’m not more productive is: What’s the point? In the past, if you came up with an elegant proof, you’d tell yourself, “I made a thing! Other people will read and use my thing! I am advancing human progress.” You can still tell yourself that now. But you have to wonder, will anyone actually read your proof? Or will they just ask another AI to explain it to them? Will your proof help their AI explain it better? For how long? If you just waited a year, would some other AI produce that same proof independently? Alternatively, if you’re a normal person, you might think that LLMs still struggle with basic arithmetic? ↩ AI progress is so fast, compared to society’s ability to adjust, that we may never actually see what “the current equilibrium” looks like. ↩ This essay, if you’re wondering, was created by writing every damned word myself and not asking AI anything. ↩ You could have a whole debate about what that precentage means. It’s not the number of words written by AI that really matters, in my opinion, so much as who is responsible for the overall structure of the ideas. ↩
I always liked stories about eccentric royals doing mad things. Queen Christina of Sweden reportedly had a miniature cannon made for firing tiny cannonballs at fleas in her bedroom. Louis XIV made a proto-reality-show of his daily routines, allowing an audience of courtiers the honor of watching him dress, bathe, and use the toilet. Peter the Great famously returned from […]
The (Re)insurance AI Landscape
to become the version of yourself needed in the next chapter