Hey all :).
I was biking back home and feeling kind of meh, and I’ve been feeling kind of meh all week. Which isn’t so much of a problem in itself—surprise, I didn’t get over the sleep deprivation in a few days like I was expecting!—but then I started being depressing to other people, and that’s just unimpressive. One of my labmates last week commented: “oh, there’s the smile,” and one of my labmates today told me: “That ‘good’ [in reply to “how are you?”] was unconvincing.” And in the latter case I stumbled through something that basically amounted to: yep, nothing’s gone wrong, life’s good, I’m just not up to usual bounce-levels for unknown reasons. Laammmmmmeee.
Especially since I’ve had all sorts of really really cool things happen recently :). So I figured I’d walk myself through a gratefulness checklist, as one does.
Fun-crazy-awesome things that have happened, off the top of my head:
— Justin and Linhchi came and visited me last week! I met them both at the conference COSYNE two years ago now. COSYNE stands for Computational Systems Neuroscience Conference, and so we all have somewhat related interests, in the intersection of computational and neuroscience / cognition in my case. We had a wonderful dinner and Justin had some really great suggestions for my womenincocosci website. I’m finally going to sit down and update that hopefully this weekend, and it was wonderful to see them and help answer questions and hear so much enthusiasm :).
— I’m going to have to rerun data collection and analysis for the paper I submitted to the conference… but then I’m going to try to submit it somewhere! That’s going to be great. I’ve never submitted a paper before, so I get to look forward to all of the nitty gritty stuff and rejections and eventual acceptance and then having it be online in print with my name on it. I’m super psyched :).
— I’ve started a new rotation! Rotations here are 10 weeks long, and we do three of them during our first year. My first rotation was with Prof. Anne Collins, my second was with Prof. Tom Griffiths, and my third is with Prof. Anca Dragan. Most people don’t decide their labs until they’ve finished their rotations, but I’m lucky in that I already know where I’ll be permanently (with Prof. Griffiths). My friends are WAY tired of me waxing poetic about him :P. But like: Tom Griffiths knows my name. Tom Griffiths cares about me and asks how I’m doing. Tom Griffiths edits my papers and gives me life advice even when he’s sick and cares about my research future. Tom Griffiths explains things spectacularly and knows so much and is so competent and makes-interesting-connections and is a great mentor all the time. AHHHHHHHHH. That’s always one of my massive and first grateful-ness things. I’m getting to work with him and I’m at UC Berkeley. That was on my nametag this weekend—”Monica Gates, UC Berkeley”—and I was kind of cooing at this inside.
— Whoops, got sidetracked, because I meant to say that my new rotation adviser, Prof. Anca Dragan, is also SO AWESOME. She does human-robot interaction stuff, trying to figure out how robots can infer human intentions by their actions and cooperate and interact better with us accordingly. It’s very, very cool. She and her student Dorsa have done a lot of work on simulated self-driving cars recently, developing the idea that people aren’t just obstacles, but will adapt to robot behavior. I just met with her today to discuss what project I’ll be working on with her, and the ideas we have are really cool, the kind of questions I want to be working on in the far-future.
— I’ve been missing talking with two of my friends, Christine and Carson, and everyone recently changed rotation labs so it’s harder to get everyone together than it was previously. That said, we’d just finished our first Ethics lecture (…you know, I think I have skipped one year of this since I got to college! But for those who don’t know, ethics in science modules are required in pretty much every research setting. I wonder if they’re required in other fields?) and I was getting ready to go tromp down to my bike by myself, but then Christine and Carson are standing by the door and down to get food. Gahhhh when the things you wish for work out exactly, and you do no work to make them happen. That’s the good life.
(Note: I’m also missing talking with other friends, and those have also been happening with beautiful convenience. The Griffiths lab has been wonderful for coincidental run-ins recently, and I end up chatting with people before Prof. Griffith’s lab pretty frequently too.)
— Speaking of friends… I got to go to the AAAI Conference this weekend (Association for the Advancement of Artificial Intelligence) and hung out a LOT with my friend Smitha, Daniel, and some with Dylan. There were some great talks, and it was great to get to know Daniel especially better, since we went on some long public transportation journeys getting to and from San Francisco. It was great to see another side of AI, since the AI at that conference was of a different style than that practiced by a lot of the Berkeley faculty and students!
— Heh. I was kindly invited to a Rationalist gathering at Ryan’s house, and it was super fun. It was exactly the kind of social interaction that I consider optimum: one-on-one long conversations, for about half an hour each, with lots of interesting, thoughtful people, standing right next to a food table with easy and natural drift between conversational partners. I just really enjoyed talking to everyone I met there, and they all had well-formed, different interests, and lots to teach or joke around about. As for the “Rationalist gathering” means—ehhhhh, that’s something I need to figure out: how to describe the Rationalist community. If you know me, I’m told that it is no surprise at all that I enjoy the mindset, but I don’t think I’m the best example. I guess that right now I’d say that it’s a group of people who say that they want to learn how to make humans think better, but there are all sort of non-consequent features and associated groups that make me tentative in a description.
— I got to see Cirque du Soleil with my friend Alyssa last weekend. It was awesome. Since we bought tickets months in advance, we were in a front row seat, and I very much would repeat that decision. There was this part where the performers were on this gigantic swing, and they were facing us, and then they shot up into the air and it was like you were flying with them. Also this part where the xylophone guy kept on taking time off of the xylophone to blow into a tuba, and it was spectacularly well-timed with the juggling, and just wonderful.
— Progress things! Like, one stupid thing right now is that I’m really enjoying the fact that I can cut my fingernails. It’s immensely satisfying, and I’m still doing well with my decision to not bite my nails since coming here. Also, I was able to keep up with the Rationalist community discussions: there are actually a lot of concepts you need to know to keep up with the baseline there. Consequentialism / dualism, cryonics, HPMoR, the classic Rationalist literature, the primary people in the movement, general sci-fi authors, philosophers, AI safety, all of the organizations people usually work at or believe in, attitudes… It’s not like you need to know this stuff—people will explain it—but I had no idea how much I had acquired. And talking with Prof. Dragan today, I knew much much more about Markov Decision Processes and reward functions and cooperative inverse reinforcement learning than when I’d met her at the same time last year for interviews. Sitting at the AAAI Conference, I also knew much more than I had even a few months ago. I like the little signs of progress.
— I’ve been exposed to some great new reading materials. Since HPMoR (Harry Potter and the Methods of Rationality), I’ve been searching for similarly wonderful books. HPMoR is seriously in contention for my favorite book ever (it’ll probably exist along Ender’s Game, which has great personal significance apart from the content), so it won’t be matched. But I’m currently reading “Crystal Society“, which has been fun (albeit pretty sexist up to this point), and every time I read silly stories I’m still so grateful to the authors for taking the time to create something that I and many others can enjoy. I’m so grateful for stories and the people who write them.
— Prof. Dragan’s lab is great and extremely welcoming and enthusiastic. Also, I just met one of her senior graduate students who is back from a hiatus, and we have a lot of overlapping interests and he sounds like he’s going to be fun to work with. You know those people who you meet and watch interact and are like: yep, they’re one of the graduate student rockstars. I have my suspicions.
— When you’re sitting next to a professor who was sitting next to Elon Musk within a few days and you’re like uhhhhhhhhhhh. (And then expected to say something intelligent.) (And actually expected not to be impressed, because we’re contributing ideas, here.) (…It is a strange life.)
— Someone contacted me about the blog two days ago, and wants to meet up! I am singularly delighted. I’m always so touched by blog readers, and one-on-one meet-ups with new people is my favorite. He mentioned he liked the how-to-manage-people paragraphs from the last post (my dad did too, thank you) and general introspection, which is good to know. One of my other penpals who contacted me through the blog tells me she likes hearing stories, so that’s good feedback. Just fyi though, I love hearing from people, and also really enjoying hearing what you like about this blog so I can do more of it. I feel like blog writing feels pretty whiny a lot of the time, as a function of writing a blog for personal reasons and emotional stressors being the things that I usually most need to write about. So I’m always glad to push elements of it that don’t feel that way to other people.
Hm :). Well now I just successfully made myself happy with the world again. I take a certain amount of pride in that (the emotional regulation) too, which isn’t exactly fair given that I have this ridiculously wonderful life—objectively, and to an even higher degree subjectively, given my goals and values. It kind of amazes me that I do still go through these periodic downswings: it’s not unusual for me to feel meh. That could just be a me-thing, but it seems to me that generally we have some really strange wiring that keeps people from being in a state of low-energy contentment, always, if that’s a reflection of how you perceive your environment. I frequently come back to this point that I don’t think it’s harmful for people to be at a low-level pleased. Whyyyyyyyyyyyy.
I also feel weird writing about this, given that happiness seems like such a high-level problem. Then again, I went to a lecture by Prof. Rosalind Picard that was talking about how the most prevalent and costly medical phenomenon in the coming years is going to be depression, given its prevalence and rates of increase now, in both the developed and developing world. She is a wonderful speaker and I really wish I had her slides so I could provide some citations. Hopefully the AAAI talks will be posted, since I know they were filmed… but in any case, human moods and emotions and brains are universally something of importance.
(I was setting my bike up for the ride home today, clipping all the lights on and buckles and locks, and I saw the woman with the hunched back walking up to the bus stop shelter near the gym. She always sleeps there, head bowed, stringy hair hanging over her face. She apparently has a pretty colorful backpack in addition to her plastic bags. A student hopped off a scooter-bike and strode around her on his way to the gym. The contrast, always. When I was on my way to AAAI, there was a guy sitting across from me eating chocolates out of a Valentine’s day box. Just popping them out and into his mouth, one by one. He had a low baseball cap and layers of zippered jackets and was leaning forward in the two seats he was taking over, eating these chocolates. At AAAI, there are lines out the door for the mens’ bathrooms; presenters are overwhelming Caucasian. On occasion they’re those guys, who think they’re the s***, just in how they lean back and hand out advice to others, like kings, like: yeah, I know my place, I know where I stand in this field, let me just speak loudly and oh-so-confidently and come swarm around me. Back in the subway, there’s a bunch of boys walking around with a plastic green bag with a smiley face on it, collecting money for the dance they did involving hanging off the railings. They were pretty good. I turned away when they came over.)
…and apparently it’s random stories time! Here goes.
I was mentioning to someone how the proportion of graduate students who have academic parents is absurd. It’s not everyone, or even the majority, but the rates are nonetheless absurd. I just found someone who won: his two biological parents, and his two stepparents, all teach in academia. (He’s one of those graduate school rockstars, by the way. I asked if he had any siblings. “Not in academia!” he said. Apparently, his brother is a software engineer in one of the big tech companies around here. …As it goes.)
(It’s so cool to have the tech population be so big here, in that I actually know people who work in industry. This is probably the biggest cultural difference from the east coast.)
Leaning on the rail on the train, listening to two guys tell me about what a “light cone” is. This had been explained to me before by another friend, but I’d forgotten, so they’re trading off trying to explain it until I exclaim “ooooh, in the time dimension.” (As I understood it, light cones are the area of space you could have physically affected. This is a small enough region, but increases at the speed of light for every unit of time, expanding in a cone in the time dimension.) They’re good at it, the explanation, and then we go back to the joke I’d interrupted by asking what it was. I used to sound normal when taking public transportation, I think.
Apparently people don’t remember their dreams? Apparently people don’t wake up at night? I’ve started tracking it, and I’m up 2-3 times a night, when the dreams switch. I dream about things that are very close to reality, and get flashbacks during the day. I said a swear word a few days ago during conversation for emphasis, which I don’t do unless I’m echoing someone, and then the dream the subsequent night was that I was uncontrollably swearing in front of small children. My subconscious is excellent at worrying about things that don’t quite make the threshold for daily thinking :P.
I have figured out my next long-term goal. And that goal is: work with some of the graduate student rockstars! There are about four of them—I know who they are—who have very similar interests to mine, and are going to be great to work with. I’ve just been intimidated by their awesomeness. But as a friend says: fight the fear, and it’s been making me feel bad for a while that I haven’t stepped up. I’m not going to set a concrete date on this, but I want to put it out there that this is something I should be looking for opportunities to do.
The associated long-term goal is to come up with projects to work with them on… which involves figuring out the questions that I’m interested in. GAH. I just had to give an elevator pitch to two amazing people, and fell flat on my face. It was so bad. I was literally stuttering at points, and was generally incomprehensible. The last time I did this badly was at Harvard Med School my senior year, when I really should have prepared something but didn’t. This time, I had prepared something, but it was the wrong thing. I was told “No, don’t tell us what you think we want to hear, tell us what you’re actually interested in.” And “Don’t tell us what techniques you think will be useful, tell us what the question is.” …So basically I have to figure out what questions I’m interested in, which involves knowing what’s been done in the literature so I know what has been done and how to frame it, but also just essentially involves knowing what I’m interested in.
I have all sorts of key words. Social inference, optimization, mechanism design, computational cognitive science, preference inference, value-alignment, AI safety, intent recognition, reinforcement learning, decision-making, biases and heuristics, priors, Bayesian inference, human-human interaction, etc. Lots of keywords, not many concrete problem spaces or questions. Mehhhhhhhh. I’d say that people aren’t generally good at this, laying out a research program, but then again I know two people in my cohort who are good at it, so it’s not like it’s not possible.
One of my friends said that what would help would be the ability to think on my feet, since if I plan badly then I generally don’t do that well. But then I immediately started thinking of methods on how to plan better, which I think was missing the point. In any case, within my current framework of just trying to plan for everything, having a cohesive research statement—with concrete problems and questions, which I now know to include—is something that needs to happen. Preferably before I work with the graduate student rockstars, but probably I can’t assign that as a pre-requisite otherwise I’m not going to get anywhere. …That said, time has helped a lot on developing interests already…
In any case, I think it’s a good goal to work with the crazy-competent people. I have this internal dialogue that continually goes: oh, they’re very good, I kind of wish I wasn’t competing with them, and in response: shut up, this is the best, we want to be with all the really really great people, wouldn’t have it any other way. I’m pretty annoyed with the “ooh I’m intimidated and wish I were a big fish in a small pond” kind of thinking that’s pervasive in my head currently, so I think thinking about it in a more positive “this is a Monica project” kind of way will help a lot. It already feels like a far more productive use of emotions and time :).
Another thing that’s been bothering me about my thinking is about being normal. I feel like I might have discussed this on the blog recently, but don’t quite know, so apologies if I’m repeating myself. This biased line of thinking can be focused on “purple hair”.
Two summers ago, one of my friends dyed her hair purple. I thought it was cool, was happy to walk around with her. Recently, I had an occasion where, under significant cognitive load, I refused to walk around with someone wearing oddly-shaped sunglasses. I also didn’t want them to get a full-arm tattoo when I was thinking normally. Additionally, I am not willing at all to dye my own hair bright colors, but was hoping that two other people would, and said “they’re not being weird properly”. For almost all of the sentences in this paragraph, starting with “happy to walk around with her”, ending with “weird properly”: …WTF?
Seriously, though, WTF is this thinking.
First of all, there’s obviously some conscious / unconscious stuff going on here, where I’m top-level trying to control for some messed-up thinking underneath. The fact that under cognitive load I refused to walk with someone with oddly-shaped sunglasses is particularly condemning, because that felt like an instinctual baseline response. Also, so much of the thought process here reveals some deep uncomfortableness with not presenting as normal (what does “weird properly” even mean?). Which is absolutely not okay, and I’m obviously aware of that, because I’m trying to hide that uncomfortableness when it occurs, and also embracing any occasion where I’m not having this response.
It feels a lot like my experience of being racist or sexist, actually. A LOT like that. The way I deal with the fact that I’m both sexist and racist is to try to overcompensate, give people a lot of extra credit upfront when I know I’m going to weight them less subconsciously. It doesn’t work particularly well, but I think it helps some. I think I’ve been trying to do the same thing here, and it’s again had mixed results.
This is a problem. It’s a problem because people should be allowed to express themselves however they wish and not be judged negatively for it, that I have no right to judge anyone anyway or, horror of horrors, inform someone of my stupid nonsensical opinion, and that I hold some normative really stupid opinion that’s not the right opinion to hold regardless (weird people are pretty much always more interesting than non-weird people).
Ug. I read a study once that said that racism in the US is temporarily reduced around Olympics time, with lots of positive exposure. But I already like all the people who present as not-normal (…whatever normal means) and think they’re really excellent people and actually admire them in most cases for not presenting at normal. Or at least I think I do. Ugggggg.
Not sure how to fix this one, honestly. Admitting it’s a problem is probably a good first step, but I’m really happy to take suggestions about this or any of the –isms my hindbrain dumps on me. And I’ll continue to be very sorry to all of the people who I’m stupid to because… well, I’m being incredibly stupid.
(…Well, talking about my happy stories now feels awkward since I just finished talking about how I’m a terrible person, but I’m just going to keep on going. If it helps, I think this brand of stupidity can underlie a lot of hateful behavior, and thus is likely not unique?)
I love watching people who are aware of the problems that can arise in groups and go and solve them. One of the usual leaders of our meetings stopped the discussion at some point, said “there are only a few people who have been talking for a while, let’s stop here and take those discussions offline, and allow anyone else who has questions to ask them now. High-level questions, like why do we care, or how this is relevant, or anything.”
Master move, right there. Got almost everyone in the room involved, very nice phrasing with example questions to ask at a high level, very very nice. That kind of leadership and moderation is definitely something to strive for.
AHHHHH love stories. This is from “Unsong“, written by Scott Alexander. (It’s a Rationalist kind of love story.)
Talking to someone who also goes to Human-Compatible AI meetings: “You’re not so quiet.”
Me: hands up in the air in the universal symbol of achievement. (Even blind people, who have never seen anyone else do this, use this gesture). Goal unlocked: have something useful to say in HCAI meetings.
Quote from my postdoc:
“if you want to griffithsify the paper, make the last paragraph of the introduction begin with the sentence ‘The plan of the paper is as follows.'”
And then he provided a link to the Google search terms: “the plan of the paper is as follows” “thomas griffiths”
This is kind of amazing. (Prof. Griffiths definitely leaves his mark on his writing.)
When somehow, you’ve managed to almost equally divide your data analysis streams between the programming languages Matlab, Python, and the program Excel, and now know far too much about how to port in between them. Badly done, Monica, badly done :P.
When you get stuck on a stats problem, and reach out to the social group of the first-year neuro students, and they all help you through it on a Saturday morning.
One of my favorite people to read, Bill Hayes, is coming to Berkeley! He writes Humans of New York-style blog posts, and is one of the people who I’d like most to emulate in writing. I like the feel of them—calm, open, fascinated, aware. He has a new book out, and I’m looking forward to hearing him speak.
A professor walks out of his office with a visiting professor, walks by me and a fellow graduate student. The graduate student stands up, introduces herself, offers her hand. I stay seated. The professor introduces me also a graduate student, and I stand up, before they walk on.
All the small signals on how to improve :).
“People don’t stop being scientists when they leave the lab!”
But aren’t we allowed to? I mean, we don’t stop, not really, but aren’t we allowed to?
One of my housemates came back from break, asked one of my other housemates Alex, “Are you cooking mashed potatoes with sausage? [That’s so stereotypically German]” and to me: “Interesting choice for a smoothie!”
Well-meant, of course, and valid things to comment on. It’s just that me and Alex had already established these facts so many months ago, and if we mentioned things like that it was for teasing purposes, long-standing jokes. It was jarring-funny, how three housemates could have such uneven relationships based on how much time we’d spent together over the course of six months.
I said something, and looked at my friend’s face, and there was a whole conversation in her skeptical look, one that we’d had several times before. Another time I said something general to a different friend, and was just waiting for the reply to come back about how what I’d said was not strictly accurate, and sure enough, he delivered :). It’s lovely to be able to predict these things, to fall into the ease of it.
Hmm, bed time :). Thank you as always so much for reading, penpals are always much appreciated, and best wishes to you all.