Shows
On Brains, Minds, And Their Possible UsesThis week in fashionOver the Christmas holiday, I became slightly obsessed with the Netflix show "Next in Fashion." It's (probably) only a temporary obsession, nothing to worry about. But before I move on to the next obsession, I wanted to take a moment to write up what I learned1. Buckle up, this will be a weird one.Before I get into the nitty-gritty of my newfound interest in fashion, let me first address the elephant in the room - why care about fashion in the first place2?Why fashionThere are three different schools3...2023-01-2300 minOn Brains, Minds, And Their Possible UsesSimulator MusingsNerding out with the OG cyborgsA few months ago, the enigmatic janus published a write-up on what they call "simulator theory" on the Alignment Forum. The post spurred a lot of discussions and changed a bunch of minds. This outcome was not completely clear from the onset - janus has been a fixture in the alignment community and on Discord, and their thought has been in the water for a while. And the message of the simulator post is comparatively simple - “large language models can/should be viewed as simulators”. But the post ended up bein...2023-01-0300 minOn Brains, Minds, And Their Possible UsesTSPSIs and How I WriteThe letter is for you, dear reader. For you who are still here, for you who are still a reader, for the only reader of all, the author, for her, for all readers, for those who've long ago read and found the story unsatisfactory, for all of you who, as I've said, have long since passed through the gateway, through the doorway, through the opening that leads to the realms of the story and to the realm of the storytellers, who are all gone, who are all gone forever. Thank you for being here. […] Best wishes to you, dear reader...2022-12-2800 minOn Brains, Minds, And Their Possible UsesIntroduction to Hebbian Natural AbstractionsWith this sequence, we (Sam + Jan) want to provide a principled derivation of the natural abstractions hypothesis (which we will introduce in-depth in later posts) by motivating it with insights from computational neuroscience.You can find the intro post to the sequence over at Sam’s Newsletter.2022-11-2200 minOn Brains, Minds, And Their Possible UsesThe Inter-Agent Facet of AI AlignmentMeta: I know you miss me, I miss you all too! I’ll be back! In the meantime, here is a little something the fantastic Michael Oesterle came up with and that might be interesting food for thought!We argue that intent alignment and capability robustness do not capture the full “chain of command” (i.e., “alignment” without any qualifier) from the programmer's intentions to an AI agent’s real-world impact, but stops at the stage where the agent performs its actions. We argue that, especially in complex Multi-Agent Systems (MAS), this is not sufficient to actually achieve alig...2022-09-2000 minOn Brains, Minds, And Their Possible UsesVia productiva. Audio version.You can get notifications for audio releases on Spotify or Apple Music.2022-07-2400 min
On Brains, Minds, And Their Possible UsesVia productiva. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-07-2415 minOn Brains, Minds, And Their Possible UsesHello DALL-E!Meta: guest post by Ava Kiai, a very awesome individual who works with bats in her PhD and who knows a thing or two about art! Lots of pretty pictures in this one!DALL-E is a powerful little parlor trick in your pocket. But what can it do? In what way does it constitute an advance? What is it good for? And what does it mean for the future of AI’s integration with how we live and work? With the help of a couple of friends, I tried to find out.The past decade ha...2022-07-1800 minOn Brains, Minds, And Their Possible UsesA quick one while he's awayA quick updateMy earliest best friend from school is named Jan. Later in high school, three out of four people in my group of closest friends were also called Jan. In university, I had three flatmates whose name was, I kid you not, also Jan1. We all had middle names that we used to differentiate ourselves, so throughout my childhood and teenage years, I was always called Jan Hendrik. It was very weird for me when I moved further south and people called me just Jan because there were no other Jans around. I almost felt...2022-07-0300 minOn Brains, Minds, And Their Possible UsesInferring utility functions from locally non-transitive preferences. Audio version.You can get notifications for audio releases on Spotify or Apple Music.2022-07-0300 min
On Brains, Minds, And Their Possible UsesInferring utility functions from locally non-transitive preferences. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-07-0314 minOn Brains, Minds, And Their Possible UsesTask Decomposition And Scientific Inquiry. Audio version.You can get notifications for audio releases on Spotify or Apple Music.On Brains, Minds, And Their Possible UsesTask Decomposition And Scientific InquiryMeta: As this post is all about summarization (and since I value your time!), I’m experimenting with putting a one sentence summary at the beginning of each paragraph. Let me know if you hate this in the comments. Astonishing Asymmetries There is a curious asymmetry in how difficult it is to make something versus how easy it is to evaluate it…Read morea year ago · 2 likes · 2 comments · Jan Hendrik Kirchner2022-06-2600 min
On Brains, Minds, And Their Possible UsesTask Decomposition And Scientific Inquiry. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-06-2611 minOn Brains, Minds, And Their Possible Uses"Brain enthusiasts" in AI SafetyMotivationAI Safety is a rapidly growing field of research that is singular in its goal: to avoid or mitigate negative outcomes from advanced AI. At the same time, AI Safety research touches on many aspects of human lives: from the philosophy of human values, via the neuroscience of human cognition, to the intricacies of human politics and coordination. This interplay between a singular goal with multiple facets makes the problem intrinsically interdisciplinary and warrants the application of various tools by researchers with diverse backgrounds.Both of us (Sam & Jan) have backgrounds in cognitive science...2022-06-1500 minOn Brains, Minds, And Their Possible UsesPuberty as Cause X? Audio version.You can get notifications for audio releases on Spotify or Apple Music.On Brains, Minds, And Their Possible UsesPuberty as Cause X?Epistemic status: I wrote this to understand how a GiveWell-type analysis feels from the inside. I expect this would require a couple hundred hours of additional research to get to the point where I'd be confident in the conclusions. But I still think the resulting analysis is interesting, so read it like you'd read the Feuilleton. Any feedback is highl…Read morea year ago · 1 like · Jan Hendrik Kirchner2022-06-1300 min
On Brains, Minds, And Their Possible UsesPuberty as Cause X? Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-06-1325 minOn Brains, Minds, And Their Possible UsesResearching Alignment Research: Unsupervised AnalysisMeta-meta: You can also find this here on the Alignment Forum.Meta: In this project, we collected and cataloged AI alignment research literature and analyzed the resulting dataset in an unbiased way to identify major research directions. We found that the field is growing quickly, with several subfields emerging in parallel. We looked at the subfields and identified the prominent researchers, recurring topics, and different modes of communication in each. Furthermore, we found that a classifier trained on AI alignment research articles can detect relevant articles that we did not originally include in the dataset....2022-06-0700 minOn Brains, Minds, And Their Possible UsesThe Unreasonable Feasibility Of Playing Chess Under The Influence. Audio version.You can get notifications for audio releases on Spotify or Apple Music.On Brains, Minds, And Their Possible UsesThe Unreasonable Feasibility Of Playing Chess Under The InfluenceA proud history of drinking and playing chess Please enjoy this clip of the reigning chess world champion Magnus Carlsen playing a game of hyperbullet (30 sec total per player) while inebriated…Read morea year ago · 3 likes · 2 comments · Jan Hendrik Kirchner and Philipp Hummel2022-06-0500 min
On Brains, Minds, And Their Possible UsesThe Unreasonable Feasibility Of Playing Chess Under The Influence. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-06-0500 minOn Brains, Minds, And Their Possible UsesThe Brain That Builds ItselfPreviously in this series: How to build a mind - neuroscience edition, Serendipitous connections: applying explanations from AI to the brain, The Unreasonable Feasibility Of Playing Chess Under The Influence, A Brief Excursion Into Molecular Neuroscience.Constraints make creativeOriginally, my motivation for getting into neuroscience was: "Building artificial intelligence is difficult, so perhaps I can learn something from looking at how the biological brain is built. Once I've figured that out, AI will be easy".(When I read these words I can't help but add a mental "duuuuh".)Since those i...2022-05-3000 minOn Brains, Minds, And Their Possible UsesSerendipitous connections: applying explanations from AI to the brain. Audio version.You can get notifications for audio releases on Spotify or Apple Music.2022-05-2900 min
On Brains, Minds, And Their Possible UsesSerendipitous connections: applying explanations from AI to the brain. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-05-2919 minOn Brains, Minds, And Their Possible UsesAdversarial attacks and optimal controlMeta: After a fun little motivating section, this post goes pretty deep into the mathematical weeds - you might prefer to read it on LessWrong with proper typesetting instead of my creative use of Unicode.TL;DR: When we invest the appropriate effort, the probability of random catastrophic events tends to decrease exponentially with a rate given by the1 rate function.One way of estimating the rate function is to solve an optimal control problem, where we have to determine the "least surprising" way that the catastrophic event comes about. The rate function th...2022-05-2200 minOn Brains, Minds, And Their Possible UsesBelief-conditional things. Audio version.You can get notifications for audio releases on Spotify or Apple Music.On Brains, Minds, And Their Possible UsesBelief-conditional things - things that only exist when you believe in them🎵 Do you believe... 🎵 It is the season, and I just finished watching the 2003 classic "Elf" with Will Ferrell as "Buddy", a human raised among elves at the North Pole who travels to New York to reunite with his biological father and to save Christmas. The movie got me thinking…Read morea year ago · Jan Hendrik Kirchner2022-05-2200 min
On Brains, Minds, And Their Possible UsesBelief-conditional things. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-05-2200 minOn Brains, Minds, And Their Possible UsesOn (Not) Reading Papers. Audio version.You can get notifications for audio releases on Spotify or Apple Music.2022-05-1500 min
On Brains, Minds, And Their Possible UsesOn (Not) Reading Papers. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-05-1513 minOn Brains, Minds, And Their Possible UsesOn Automatic Ideas. Audio version.You can get notifications for audio releases on Spotify or Apple Music.On Brains, Minds, And Their Possible UsesOn Automatic IdeasCode first, ask questions later In On Scaling Academia I made the case that automatic research, i.e. AI-supported & scalable application of the scientific method, might be great. My argument went like this, automatic research is beneficial for academia (systematization and…Read morea year ago · 1 like · 1 comment · Jan Hendrik Kirchner2022-05-1100 min
On Brains, Minds, And Their Possible UsesOn Automatic Ideas. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-05-1114 minOn Brains, Minds, And Their Possible UsesElementary Infra-BayesianismShort summary: I apply the updating equation from Infra-Bayesianism to a concrete example of an Infradistribution and illustrate the process. When we "care" a lot for things that are unlikely given what we've observed before, we get updates that are extremely sensitive to outliers.I've written previously on how to act when confronted with something smarter than yourself. When in such a precarious situation, it is difficult to trust “the other”; they might dispense their wisdom in a way that steers you to their benefit. In general, we're screwed.But there are ideas for a cons...2022-05-0800 min
On Brains, Minds, And Their Possible UsesSlightly advanced decision theory 102: Four reasons not to be a (naive) utility maximizer. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-05-0425 min
On Brains, Minds, And Their Possible UsesThe Greedy Doctor Problem. Audio version.You can get notifications for audio releases on Spotify or Apple Music. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-05-0120 min
On Brains, Minds, And Their Possible UsesDrug addicts and deceptively aligned agents - a comparative analysis. Audio version.This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-04-2700 min
On Brains, Minds, And Their Possible UsesFrankfurt Declaration on the Cambridge Declaration on Consciousness. Audio version.This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-04-2408 min
On Brains, Minds, And Their Possible UsesApplied Mathematical Logic For The Practicing ResearcherThis is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-04-2011 min
On Brains, Minds, And Their Possible UsesHow to build a mind - neuroscience edition. Audio version.This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-04-1716 min
On Brains, Minds, And Their Possible UsesCognitive Biases in Large Language Models. Audio version.This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-04-1400 min
On Brains, Minds, And Their Possible UsesOn Scaling Academia. Audio version.Check out the written version here: This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit universalprior.substack.com2022-04-1023 min
On Brains, Minds, And Their Possible UsesSoldiers, Scouts, and Albatrosses. Audio version.An audio version of “Soldiers, Scouts, and Albatrosses”. Writing long-form is fun, but reading a long post every week is perhaps a bit much. So now you can also just listen to my voice read out the posts! I plan to release these slightly more frequently than the written essays so that I catch up eventually. Also, no worries about getting flooded with notifications about this. For the next audio releases, I won’t send an email to everyone about it. You can however get notifications for audio releases on Spotify or Apple Music. This is a publ...2022-04-0313 min
On Brains, Minds, And Their Possible UsesThe Tale Of Gandhi And The DevilCo-authored by Nadia Mir-Montazeri and Jan Kirchner.Meta: This week here is something different! The following story was written for the EA Creative Writing contest. It was inspired by (spoiler alert!) this thought experiment by Scott Alexander. If you prefer listening to reading, there is an audio version of the story at the top of this post."A painting of the devil bargaining with Gandhi.", generated by CLIP-guided-diffusion.It just annoys me, that’s all. Every century or so, one of the squishy ones thinks they are so great and that everybody should lo...2021-10-1019 min