HomeWritingCrosswordsNow

College Advice for People Like Me

Henry Josephson
April 11, 2025

I'm graduating from UChicago in around 60 days, and I've been thinking about what I've learned these past four years. I figured I'd write it all down while it's still fresh.

This isn't universal advice. It's specifically for people like me (or who want to be like me). High-agency, motivated types who hate having free time. When I sent drafts of this post to people for comments before publishing, one told me "you're probably the highest-agency person I know." Maybe the nicest thing anybody's ever said to me. People who'd rather risk making mistakes than risk missing out, who want to control more than they initially think they can, and who are willing to go all-in relatively quickly. If you're reading Ben Kuhn and Alexey Guzey or have ever heard of the Reverend Thomas Bayes, you're probably one of us.

So here's at least some of what I've figured out. Take what's useful, leave what isn't — maybe do the opposite of everything I've said.

Mindset and Personal Growth

Find your mission

Find something you're fully pointed towards. Something you direct yourself towards. It's tough to know if you've really found it, which is where epistemic humility comes in, but it's a bit like falling in love, where you realize that it's happened after it's happened / you've been pointing at it for a bit.

The mission is useful both inherently (because the mission should be something you really care about for itself) but also instrumentally as something that guides action. Vagueness in the inherent goodness can be kinda bad, but vagueness in the instrumental means you lose action-guidingness.

Don't make your mission everything, though. The failure mode is putting too many of your eggs into any basket, even when that basket is awesome. Have at least one non-mission-aligned friend. Your mission can be vague / big / lofty, but the things you do to get closer to it can't be. If both mission and steps are vague, you end up just fucking around instead of actually making progress.

Recognize that you can always be better

This isn't about feeling emotionally bad — that's unsustainable. It's a principled refusal to plateau. Not that you're bad, but that you affirmatively should take the room to grow that you have. You stand in the shadow of who you could have been and who you could ultimately be. The best time to plant a tree is 20 years ago. Second best time is now.

Make more mistakes

Most mistakes that your could make probably aren't very high-cost. Forgetting people's names, getting rejected from some applications, asking a question in class and sounding dumb for 10 seconds, trying out for a club or team and not making it, starting a side project and abandoning it halfway, telling someone you admire their work and getting ignored, wearing something a little weird and getting a weird look, sending a cold email and getting no reply, posting writing online and getting 12 views.

People (including me!) fear failure so much that they don't try, on relatively stupid small stuff where the cost of failure rounds down to the status quo. If you aren't making mistakes, you're underexploring. Most low-stakes failures cost nothing and reveal where the edges are. The real mistake is being so allergic to error that you stop gathering data.

Things only get done to the extent that you want them to get done

Very often, more often than you might think, Really more often than you might think! things will only get done to the extent that you want them to get done. This feels kind of like a banal platitude, but it's very eye-opening. If you want something to happen, make it happen.

You don't need to ask permission for most things. If they only depend on you wanting them to get done, you don't need to ask other people for them to get done. Too often people take the fact that they might have to ask permission as an implicit no, which it just isn't.

For a lot of choices, the worst-case scenario is just that the status quo persists. Worst-case scenario, you do it and nothing happens, and you're in the same world as if you had never done it. So clearly, if the upside is at all positive, the EV is positive. When I was in the fall of my third year, I took a class called AI for Public Policy with Jens Ludwig, professor at UChicago's Harris School, cool guy, and I wanted to do research, and I wanted to do the research that he did. So I reached out to him, I asked if he had time to talk about this. I went to his office hours. I met with him in the next quarter and was like, hey, do you have research projects I can work on? Do you have things that I can do? And he did. And that wouldn't have happened if I were less effortful. I think it only happened to the extent that I was motivated to keep sending him emails and keep showing up at his office hours.

There are no adults

This is a corollary to the above — there is no secret group of smart, qualified people behind the scenes making sure everything goes well. There probably isn't anybody watching over you to make sure you don't fuck up. This is liberating (you can do just great things) and terrifying (you can do terrible things).

Deadlines are mostly fake

Deadlines are mostly fake, and this is both good and bad. The good part is that deadlines are only "official," I find that "officially" often means "not really". see the sears tower vs world trade thing. and most professors, if you ask nicely enough or have a good enough reason, will delay the deadline. Which means that if you want to change the deadline, you can ask the professor, and in most cases, they'll change the deadline.

What is a deadline, anyway? It's just the moment when consequences kick in. But you have leverage in that. There are broadly two ways you can change the incentives that apply to you — by changing the external pressures that act on you, and by changing how you react to those external pressures. Good agency involves recognizing that the defaults aren't necessarily right, or built for you. e.g. set artificial deadlines before things are actually due by setting an office hours appointment to go over a draft you haven't written yet.

Put yourself in positions where you'll be lucky

Luck is distributed unevenly. Lightning doesn't strike twice, but you're way more likely to get hit atop the Empire State Building. Find where chance encounters happen, where cool people who could introduce you to your next job hang out, and go there.

It's mostly about proximity to opportunity, but also a bit of shots on goal. Be available to influential professors / mentors / movers and makers. This sort of thing is most obvious in its converse — good things will not happen to you if you spend all day reading in your room. Read somewhere where your friends will see you if they walk past, and invite you to grab a drink with them. Make sure the right people know you before they need to know you, so that you're at the top of their mind once you have to be.

Luck favors the prepared

Once you've positioned yourself where luck can find you, you need to be ready when it arrives. OK, so you were in the room and you got picked. Now what? Are you ready to go? Do you have a plan ready to use? If the previous step is all about getting you to the free throw line, but you're not sure when, you'd better be practicing your shot.

Spend 5-10 minutes thinking about what you'd do if big things happen. You don't need to have everything fully worked out, but the first steps especially should be ready to go.

Test your fit at lots of things

One of the biggest things you should be doing in college is figuring out how you want to spend your career. What's the best way to figure out whether you'd be able to work in something? Actually trying the work. Find cheap ways to test whether you'd be good at something, whether you'd enjoy it, whether you have an on-ramp to do it.

Toy models and vibe checks are still useful insofar as they provide verisimilitude. Try writing policy memos or taking notes on legislative sessions before you gun for a leg internship. Check for enjoyment, aptitude, motivation, status alignment. Can you see yourself doing this for years? It's not weird for you to spend your three summers doing three different things.

Do side projects

Find things that seem cool and do them. They're fun. They're also how you'll do most of your learning. Much of my learning didn't come from classes — it came from thinking "this is interesting!" about something adjacent to coursework. Homework teaches specific skills. That's necessary. But to learn how to build things, you have to build things.

Don't spend most of your time on side projects. The optimal amount isn't zero, but it's not fifty percent either.. Side projects are exploration, not exploitation.

Things that seem irrelevant often become relevant later. They might give you unexpected skills, help you make connections, or serve as credentials. Being able to say "I build things" is a powerful signal to people with biases toward action.

Get good at introspecting

This is much tougher to get operational advice on, but the closest I can get is having practice kicking your thoughts up one level of abstraction from having thoughts to noticing that you're having the thoughts. Especially when so much of the world is adversarially optimizing to keep you locked in to things you don't want to be locked in on — looking at you, YouTube shorts. See also Algotransparency's manifesto. You often do things without realizing that you're doing them. Analogously, you often do things without realizing you're doing them, or how much time you're spending on them. Track your time in detail for at least a day; you'll be surprised what you learn.i

Feelings are slippery. If one is flopping around (wrong verb), trapped in your mind, recognizing it means you can let it go and clear your mind, but being able to snag it before it leaves means you can hold it in your hands and examine it. Where, physically in your body, do you feel the emotion? Why, causally, do you feel it?

Be honest

Honesty is incredibly important, and very, very rarely is there an advantage to be gained interpersonally or intrapersonally by lying. I mean this not only instrumentally — lying corrupts.

There is no belief without action. If you say you care but don't act, you don't actually care. Turn vague ideals into hard commitments — ones you could be called out on. If your conviction can't be falsified, it probably isn't real.

Put your money where your mouth is

Revealed preferences >> stated preferences. If you think something is worth doing, you should be doing it. If you aren't doing it, you almost certainly, deep down, don't think it's worth doing. Are you lying to yourself? In some cases, be literally willing to put real money on the line for your beliefs. If you think it's important, it's worth actually giving your time to it. Spending your money on it. Devoting your career to it.

Interpret others charitably

I subscribe to Hanlon's razor, but replace "incompetence" with a combination of "having something else going on, inattention, or indifference." People won't think the same way you do and won't do things the way you do. How much you get done is very different from actual value as a human being, though, and a danger of buying in too wholeheartedly to the "get stuff done" mindset I've been illustrating is that you end up looking down on people who don't.

Be the kind of person others can come to for help

Be emotionally available and nonjudgmental in moments of stress. Cultivate enough slack or surplus that you can afford to help others. Avoid performative busyness so people feel they can approach you. Make it clear (implicitly or explicitly) that people should reach out when they need help.

Default to yes. Being the kind of person others can come to means being ready — emotionally, practically, morally — to say yes when asked. Not unconditionally, but reflexively. Especially when it's easy for you and hard for them. The world works better when you are generous by default and when "person who helps" is core to your identity.

Productivity and Focus

Go to bed early

Go to sleep when you're done, not when you're exhausted. If your evenings devolve into aimless scrolling, https://en.wikipedia.org/wiki/Bedtime_procrastination don't treat them as free time — treat them as wasted time. More broadly, if you're doing something that isn't useful, stop doing it. You could be asleep. You could be waking up sharp, ahead of everyone else. I'm a huge morning person. Very annoying to me that UChicago's gym didn't open until 7 am. You can stop doing things that aren't helping you. That includes the end of your day.

Brick your phone

Grayscale your phone. Set screen time limits. I use ScreenZen. It's free. You were not built for this device. Your phone is a superstimulus. https://www.lesswrong.com/w/superstimuli If you don't fight back, you lose by default. Brick your phone — grayscale it, On an iphone, settings > accessibility > display and text size > color filters > grayscale. strip notifications, Only have notifications on for things you wouldn't mind someone literally interrupting your work and tapping you on the shoulder to tell you about (and this probably isn't most things you get notifications for). hide addictive apps. I use Smile. It's free. Treat attention as sacred, and add friction to things you don't want to do. Protect it like you mean it.

The optimal amount of slack time is not zero, but it's close to zero

If you're like me, you hate having free time. You're happiest on days where you have things back-to-back-to-back-to-back. You're willing to go all-in relatively quickly. But the downside for this is having no time for others, to be curious, or to explore. Having to say no. Confusing what's urgent with what's important.

Have time blocked off to do stuff that matters… and have a little time blocked off for stuff that doesn't (I shoot for 2-3 hours a day). Slack is necessary for the good stuff that isn't on your calendar. Maximal scheduling can feel productive, but it crowds out curiosity, serendipity, and people. The goal isn't zero free time — it's deliberate slack. Block time for rest, for wandering, for non-urgent joy. Without it, you'll optimize into a corner.

If you aren't getting work done, pick up your shit and go somewhere else

Have multiple places you can get work done. If you're just arriving to college, find three or four out-of-the-way study spots. If you're coming to UChicago, email me and I'll let you know mine. If you find yourself being unproductive in one spot, change scenery to give yourself a mental reset, then restart wherever you've ended up. New place, new start.

Don't take your phone to the places you're studying

You can full-on leave your phone in your room. Nothing bad will happen. You'll be fine. It's much more likely to be a distraction than a good thing, and it'll be right where you left it when you come back. If you need, tell someone where you're going. Or tell them to email you.

Do not try to do more than one important thing at once

You get one focal point at a time. You can juggle, sure — but real focus is zero-sum. When two important things compete for your attention, neither wins. Pick one. Block time for it. Let the rest wait.

Offload difficult things to automations, habits, or other people

There are three big ways to offload the light mental loads. Automate them. Have Claude write you code to do it. Unironically, your little frictions go away. Make the things second-nature and habitual, so they don't need to take up any part of your perceptual space, or have friends help. Your working memory is finite — don't waste it remembering to buy toothpaste.

Make your bed

Disorder compounds — fight it early. Little messes become big ones. Visual clutter becomes mental clutter. If something can be put away in 30 seconds, do it. Make your bed. Give everything a home. Everything has a home. Keep the entropy outside your mind.

If it takes less than 5 minutes, do it now

Light tasks get heavy fast. A two-minute chore becomes a backlog becomes a burden. If it'll take less than five minutes and you're not deep in something else, just do it. If you are focused — don't switch. Write it down somewhere you'll actually check. Then return when the runway clears.

The floor is actually way lower than you think

I mean this more as "the activation energy is way lower than you think it is. I got some of this when I TAed a class I thought was easy with some studying and saw how little studying some people did. I was getting emails the night before the exam saying "oh, I never studied and now I'm freaking out!" crazy. You can be far more exceptional than you think you can". Don't shift it into motivation-speak, but most people just don't give anywhere near 100%. Just doing the readings at all is enough to catapult you to the front of most classes.

Planning and Goal Setting

Make sure your goals are falsifiable

Set goals for yourself, and make sure you'll be able to tell when you've failed them (or when you haven't). Good goals are also achievable. Success and failure tend to snowball inside you. A parallel skill is getting good at task decomposition (and is very helpful for research management).

Set goals you can fail at. If a goal isn't falsifiable, it isn't actionable. "Get better at writing" is mushy; "publish one blog post a week for a month" is real. Make your targets crisp enough that you'll know if you missed — and then you can try again.

Vague ambitions don't guide action. Sharp ones do. Achievable ones snowball. And big ones? They get built one small, testable success at a time. Learn to break things down. That's the only way anything gets finished — especially research.

Credibly pre-commit to things you care about getting done

Find ways to tie yourself to the mast. Future you is their own person and they won't always want to sit down and do the work. But you can control their incentives. Tell a friend you'll send them the thing once you're done with it. Put money on the line. Make it so that, if future you behaves suboptimally, they actually feel it. Don't get principal-agent problemed Link by yourself.

Track your progress

It feels great to see yourself improve! And to know that the improvement is because you're choosing to put the effort into improving. Track what you want to get better at. You'll never find a number for how good you are I've tried. there isn't one. — but you can log your workouts, save your essays, journal your mood. Progress feels better when you can see it. And it hits different when you know it's because you made it happen.

Make a 5-year plan google doc

Do not decide your future based on vibes. Think through. Think by writing, because you can learn a lot by having to write it out. If you make a copy of this Google doc and email me the Google doc link set so anyone can edit, and I'll leave you comments (hi@henryjosephson.com). I'm offering because I know none of you will do it. prove me wrong. Seriously, block time off.

Consider graduating early

Consider doesn't mean do, but as a rule, if you aren't doing something big, if you're not doing it, make sure it's because you thought about it and ruled it out, not because it never crossed your mind. Most of the value of your degree Note — I'm not saying most of the value of your time in college. is just as a signal that you're smart. But graduating in two years can be an even stronger one — if you have something better to do next. And "better" includes your health and happiness.

Relationships and Community

Build your own community

You can absolutely make friends through your dorm or classes, but that's very passive. Don't wait for the things you want to come to you. Do community-building! Lots of good community-building resources here. Organize a club! Get the practical skills it gives you, but also learn what activities and advertisements attract people like you. It's also a great excuse to reach out to people.

Fall in love at least once

Fall in love at least once. Not for the plot. For what it teaches you. You'll learn what kind of partner you are, how you handle closeness, how you grow alongside someone — and how you mess up. It's hard to describe the upside if you haven't felt it.

Also learn to see past the love-blindness, though, because your partner is the person you're most inclined to see positively. Learn to look inside, because, when you're really in love you'll love this other person more than yourself, and it'll be tough to recognize when you're uncomfortable or the relationship is going downhill.

If you aren't happy and excited and exciting single, you won't be happy or excited or exciting in a relationship

Some people you know will think love will all of a sudden make them happy. This is, in general, false. Not only is being sad and boring a red flag that'll make it tougher to get dates, but healthy relationships don't turn you into someone else. If you're sad and boring alone, you'll still be sad and boring in a relationship — plus now it's someone else's problem too. Don't bet your happiness on getting a girlfriend.

This remains true once you're in the relationship. Keep a real life outside your partner! Have your own hobbies. Friends. Competitions.

Friends are people you can talk to for hours

Friends are people you can talk to for hours. It's a litmus test: if the conversation loops, stretches, deepens, you've probably found someone worth keeping. If they have the opportunity to end the conversation many times, but they don't, it's becuase they're having fun. This is also flirting advice, insofar as conversation is a skill. I have had multiple exes say that they loved how easy I was to talk to. These conversations are how relationships grow and how real learning happens. But it's not just about finding people like that. Be someone who's easy to talk to.

Throw parties

Parties are such a social bottleneck!!! Everybody likes them, nobody wants to throw them, and frats suck. Make sure your roommates are okay with it, then invite your friends and acquaintances. Try to send invites out at least a week and a half in advance. Experiment with themes and sizes.

If you have a fire escape, see if the people in the apartment above you also want to throw a party, then throw a double-decker. Bonus points if you can get the apt below you in on it, too, and then you can do three. You will never have more people your age around you who want to go to parties than in college.

Hang out with people who are better than you at the things you care about

Spend time with people who outmatch you. If you care about getting better at something, hang out with people who already are. Not just for advice — for the ambient learning. Watch how great thinkers ask questions. Watch how charismatic people move through rooms. You become what you're around. Seriously, frats suck. Choose your average deliberately.

Professors are people. You can just make friends with them

Professors are people. You can just befriend them. Not in a networking way. Just in a human way. They like talking about the things they love. They like when students care. And you might find you have more in common than you think. Reach out. Say hi. Follow up. Sometimes it turns into mentorship. But either way, you'll get more than just lectures.

Academic Success

Go to office hours

Go to office hours. Seriously — just go. Nobody else does. When I TAed a 60-person intro class, like 5 people showed up. That's why it works. You get the grader's eyes on your work. This is a cheat code, especially when writing papers at higher levels. Your profs aren't just good at building papers -- they decide what grade your paper gets! You get real-time feedback. You build relationships with smart people who want to help. And sometimes they'll like you enough to root for you. That helps too.

Be careful how you use AI

Be honest with yourself about how you're using AI. Lots of this post was voice memos before Claude helped me clean it up, for example. AI is now good enough to one-shot every lower-level CS class and is a strong writer below ~2 pages (8 or 9 if you know how to do it well). But if you have the AI do the work for you, you won't learn the thing you're actually working on. If you're only doing college for the signal, maybe you don't care, but the signal is probably only useful for getting a job, and not being able to do a SQL join is much more damning.

Read things, everywhere

There is information and amazement everywhere. College towns usually have multiple great bookstores Hyde Park's best one is Powell's. where you can find things interesting, and fill your bookshelf. Curate newsletters. I'm working on another blogpost about this, eventually. Podcasts only kinda count. Go the library and wander through the stacks until you find one or two things that're interesting. Sneak peeks at your professor's shelves when you're in office hours. Ask people who are where you want to be for book recommendations at the end of your coffee chats. Keep a folder on your computer of interesting pdfs.

Write in your books! Star things and highlight them and summarize them in the margins every couple pages to make sure you're tracking things. Don't treat books like sacred objects. Treat them like training grounds. Highlight. Argue in the margins. Write summaries every few pages to make sure you're following. If the book changed you, it should look like it.

Don't be afraid to skip the boring parts

Just don't sunk-cost yourself into parts of a book that aren't being helpful. You don't owe a book your obedience. If a section isn't teaching you or holding your attention, skip it. Skimming isn't always cheating — it's strategic. The goal isn't to finish. The goal is to learn. Don't let sunk costs trap you in pages that aren't serving you.

Learn how to read quickly

Expect to face a larger volume of reading in college than you initially think you can handle; this is normal. Part of navigating this workload is learning how to read efficiently, which involves skills like triaging material, skimming for main points, zooming in on crucial sections, moving fast when appropriate, and crucially, knowing when to slow down. Like any skill, reading efficiency comes with practice, so start building it now, even if you're still in high school. A great way to get reps is by engaging with journal articles — try sorting by 'most downloaded' on platforms like PhilPapers, SSRN, JSTOR, NBER, or ArXiv. Use AI tools like Claude to help explain complex paragraphs you encounter along the way.

Write things

Get good at expressing yourself, and also get good at making points cogently. Remember how I talked about making important things second-nature? You can do this with persuasion and explanation and entertainment and all the important things writing does. If you can't put it into words, that's good evidence that you don't understand it. And if writing is hard every time, you'll do it less — so make it second nature. Practice turning thoughts into sentences. Practice circumlocution when you're stuck. You can't always wait for the perfect phrasing.

Read widely based on curiosity, not just relevance

Read widely based on curiosity, not just relevance. Follow your gut and read things just because they seem interesting, even if they feel totally unrelated to your classes or goals right now. It's enjoyable, and you never know when that random knowledge might spark a cool idea or turn out to be useful later. Some of your best insights might come from sources you weren't even looking for. It won't always pay off right away, but this habit will shape the questions you ask and how you connect different ideas, which is more valuable than you might think.

Have an easy way to capture ideas

Good ideas, interesting quotes, or book recommendations often pop up at inconvenient times – maybe walking to class or in the shower. Don't let them vanish! Find a simple, quick way to jot them down. Use whatever works best for you: a dedicated app like Jotlog, your phone's built-in notes app, or just a small notebook and pen you carry around. The main thing is making it effortless.

Health and Lifestyle

Go to the gym

If you're like me, you're happiest and most fulfilled when you're active. You get a clear thing that you can see yourself make progress on, and it also helps you live longer, reduces stress, and makes everything else easier. You'll also make gym friends who you'll talk to every day and not fully know their names.

Your body isn't totally yours to shape — but it's more shapeable than you think. And the feedback is fast. If you don't look the way you want to look, you can get there (within reason).

The only place to work yourself to failure is the gym

Burnout sucks. Actual full failure is encouraged at the gym, but you shouldn't work yourself to the bone everywhere.

The gym is the only place to push to failure. Not everything in life rewards total effort. Burnout is real. But in the gym, failure is part of the process — it's how you grow. Save your full-force, leave-it-all-out-there energy for a space built to hold it. Don't live every day like a max-rep set.

If something isn't making your life better, change it

You can change your life. That sounds dramatic, but it's literal. You choose when you wake up. Who your friends are. What you consume. Where you spend your time. College might be the first time you're far enough from your defaults to really see that. So look. And if something isn't making your life better, change it. You can.

Leave campus

Don't spend your whole time within the few blocks / acres / in some cases literal walls of your campus. Leave. It's a bubble. Meet people who go to different schools. Cool things are far away. Go places on a whim. Walk home to downtown. Bike to Milwaukee. Find the best bagel store on the other side of the city. Buy a cheap plane ticket to Oregon and walk around Mount Hood. If you don't explore now, when will you?