Effective Altruism Forum
EA Forum

Hide table of contents
by Yegreg, alexflint
4 min read 0

5

AI safetyAI safety resources and materialsAnnouncements and updates
Frontpage
This is a linkpost for https://www.lesswrong.com/posts/sPAA9X6basAXsWhau/announcement-learning-theory-online-course

Hey everyone! Gergely and Kōshin here from ALTER and Monastic Academy, respectively. We are excited to announce a 6-week, 4-hours-per-week mathematics course designed to introduce a few early topics in what we are calling mathematical AI alignment (roughly, classical learning theory up to embedded agency and infra-Bayesianism, though this course will cover just an initial portion of this) aimed at folks with the math proficiency level of a university math major (formal university degree not required). What this means is that if you have some mathematical dexterity, and wish that you could answer difficult questions on MDPs, POMDPs, learnability, sample complexity, bandits, VC dimension, and PAC learning, then this course may be for you.

Lectures will be given over video call by Gergely, who has a math PhD from Stanford and has been working with Vanessa Kosoy over the past two years on the Learning Theoretic Agenda. Kōshin is helping with logistics. We are planning to have a one-hour video lecture each week, a one-hour discussion session each week, and a weekly problem set that should take around two hours to complete.

The cost of the course is $200 per participant, which we’ll refund up to the full amount based on attendance.

There is a question set for which we ask you to submit solutions in order to apply to join the course. This is to help us select a cohort with reasonably uniform mathematical proficiency, and to help us design material for that level of proficiency. We are looking for a cohort who are willing to make a strong commitment to joining all the lectures and discussion sessions, and solving all the question sets for the entire 6-week duration. You do not need to make that commitment in order to apply, but we’ll ask for it before the course begins.

To get started right away, see the entry questions and apply for the course.

The remainder of this post is organized as follows: First, we’ll pitch the larger vision for the course. Second, we’ll discuss the entry question set and why it makes sense for anyone interested in the course to submit their best attempt at it. Third, we’ll go over dates and costs.

Pitch for the course

We feel that the best way to pitch the course is to present the question set that motivated us to organize the course in the first place. The question set below was prepared by Gergely for an AI safety workshop in the summer of 2024. At the end of the workshop, we decided to design a course that would teach the subject material needed to solve it. We realized that such a course would be quite long, so we decided to run a 6-week trial to gauge interest and gather information. We therefore present the following question set as a vision for the long run journey that we hope to go on, supposing that the course is successful and that we go on to run further courses. It goes significantly beyond the material that will be presented in the 6-week trial. Here is an excerpt (Also see the full question set):

Entry criteria

Our goal in the course is to bring participants to the point of being able to directly make use of the subject matter that we introduce. The course will therefore revolve around solving problem sets, as this is what we’ve seen bring people to this point. The topics we’re introducing do not, for the most part, depend on prior knowledge of advanced branches of mathematics. We believe we can give a first-principles introduction to these topics in about 4 hours per week of participant time. To keep up with the pace of the course will, however, require the kind of mathematical dexterity (not domain knowledge) that one would gain over the course of a university math major.

In the entry question set, the first question is intended to be fairly straightforward, while the second is intended to be challenging. This is so that we can calibrate the difficulty of the course in accord with the submissions, as well as selecting participants with similar levels of proficiency.

We therefore encourage anyone interested to submit their best shot at answering these questions. Partial solutions are welcome.

We hope to find an initial cohort who will all commit to taking the 6-week course in its entirety – turning up to every lecture and discussion session, and as far as possible solving every problem set. You do not need to make such a commitment in order to apply, but we’ll ask for such a commitment, subject to finalization of the exact schedule, before the course begins.

If you are interested in participating, please write up solutions to the entry question set in LaTeX, and submit a PDF along with your details. Feel free to ask for clarifications in the comments below, but for obvious reasons please do not post solutions.

Topics

The course will cover the following topics:

  • Cartesian framework for learning theory
  • MDPs/POMDPs
  • Learnability
  • Sample complexity
  • Bandits
  • VC dimension
  • PAC learning

Dates

The course will begin in the week of March 17th, 2025 and run through the week of April 21st, 2025, with exact dates and timing to be decided based on timezones of folks in the initial cohort.

Video call

We will be available by video call for questions about the course on February 17th at 9am Pacific. Register for the call. We will also respond to questions asked in the comments section on this post.

Cost

Due to the generosity of ARIA, we will be able to offer a refund proportional to attendance, with a full refund for completion. The cost of registration is $200, and we plan to refund $25 for each week attended, as well as the final $50 upon completion of the course. We’ll ask participants to pay the registration fee once the cohort is finalized, so no fee is required to fill out the application form below.

Next steps

5

0
0

Reactions

0
0
Mentioned in

More posts like this

Comments


No comments on this post yet.
Be the first to respond.
Curated and popular this week
LintzA
 ·  · 15m read
 · 
Cross-posted to Lesswrong Introduction Several developments over the past few months should cause you to re-evaluate what you are doing. These include: 1. Updates toward short timelines 2. The Trump presidency 3. The o1 (inference-time compute scaling) paradigm 4. Deepseek 5. Stargate/AI datacenter spending 6. Increased internal deployment 7. Absence of AI x-risk/safety considerations in mainstream AI discourse Taken together, these are enough to render many existing AI governance strategies obsolete (and probably some technical safety strategies too). There's a good chance we're entering crunch time and that should absolutely affect your theory of change and what you plan to work on. In this piece I try to give a quick summary of these developments and think through the broader implications these have for AI safety. At the end of the piece I give some quick initial thoughts on how these developments affect what safety-concerned folks should be prioritizing. These are early days and I expect many of my takes will shift, look forward to discussing in the comments!  Implications of recent developments Updates toward short timelines There’s general agreement that timelines are likely to be far shorter than most expected. Both Sam Altman and Dario Amodei have recently said they expect AGI within the next 3 years. Anecdotally, nearly everyone I know or have heard of who was expecting longer timelines has updated significantly toward short timelines (<5 years). E.g. Ajeya’s median estimate is that 99% of fully-remote jobs will be automatable in roughly 6-8 years, 5+ years earlier than her 2023 estimate. On a quick look, prediction markets seem to have shifted to short timelines (e.g. Metaculus[1] & Manifold appear to have roughly 2030 median timelines to AGI, though haven’t moved dramatically in recent months). We’ve consistently seen performance on benchmarks far exceed what most predicted. Most recently, Epoch was surprised to see OpenAI’s o3 model achi
Dr Kassim
 ·  · 4m read
 · 
Hey everyone, I’ve been going through the EA Introductory Program, and I have to admit some of these ideas make sense, but others leave me with more questions than answers. I’m trying to wrap my head around certain core EA principles, and the more I think about them, the more I wonder: Am I misunderstanding, or are there blind spots in EA’s approach? I’d really love to hear what others think. Maybe you can help me clarify some of my doubts. Or maybe you share the same reservations? Let’s talk. Cause Prioritization. Does It Ignore Political and Social Reality? EA focuses on doing the most good per dollar, which makes sense in theory. But does it hold up when you apply it to real world contexts especially in countries like Uganda? Take malaria prevention. It’s a top EA cause because it’s highly cost effective $5,000 can save a life through bed nets (GiveWell, 2023). But what happens when government corruption or instability disrupts these programs? The Global Fund scandal in Uganda saw $1.6 million in malaria aid mismanaged (Global Fund Audit Report, 2016). If money isn’t reaching the people it’s meant to help, is it really the best use of resources? And what about leadership changes? Policies shift unpredictably here. A national animal welfare initiative I supported lost momentum when political priorities changed. How does EA factor in these uncertainties when prioritizing causes? It feels like EA assumes a stable world where money always achieves the intended impact. But what if that’s not the world we live in? Long termism. A Luxury When the Present Is in Crisis? I get why long termists argue that future people matter. But should we really prioritize them over people suffering today? Long termism tells us that existential risks like AI could wipe out trillions of future lives. But in Uganda, we’re losing lives now—1,500+ die from rabies annually (WHO, 2021), and 41% of children suffer from stunting due to malnutrition (UNICEF, 2022). These are preventable d
Rory Fenton
 ·  · 6m read
 · 
Cross-posted from my blog. Contrary to my carefully crafted brand as a weak nerd, I go to a local CrossFit gym a few times a week. Every year, the gym raises funds for a scholarship for teens from lower-income families to attend their summer camp program. I don’t know how many Crossfit-interested low-income teens there are in my small town, but I’ll guess there are perhaps 2 of them who would benefit from the scholarship. After all, CrossFit is pretty niche, and the town is small. Helping youngsters get swole in the Pacific Northwest is not exactly as cost-effective as preventing malaria in Malawi. But I notice I feel drawn to supporting the scholarship anyway. Every time it pops in my head I think, “My money could fully solve this problem”. The camp only costs a few hundred dollars per kid and if there are just 2 kids who need support, I could give $500 and there would no longer be teenagers in my town who want to go to a CrossFit summer camp but can’t. Thanks to me, the hero, this problem would be entirely solved. 100%. That is not how most nonprofit work feels to me. You are only ever making small dents in important problems I want to work on big problems. Global poverty. Malaria. Everyone not suddenly dying. But if I’m honest, what I really want is to solve those problems. Me, personally, solve them. This is a continued source of frustration and sadness because I absolutely cannot solve those problems. Consider what else my $500 CrossFit scholarship might do: * I want to save lives, and USAID suddenly stops giving $7 billion a year to PEPFAR. So I give $500 to the Rapid Response Fund. My donation solves 0.000001% of the problem and I feel like I have failed. * I want to solve climate change, and getting to net zero will require stopping or removing emissions of 1,500 billion tons of carbon dioxide. I give $500 to a policy nonprofit that reduces emissions, in expectation, by 50 tons. My donation solves 0.000000003% of the problem and I feel like I have f
Recent opportunities in AI safety
3
Apply to become a Futurekind AI Facilitator or Mentor (deadline: April 10)
Jay Luong
· · 2m read
0
0
4
Last week to apply for the Futurekind AI Fellowship! (deadline: April 1)
Jay Luong
· · 1m read
0
0
35
Apply to MATS 8.0!
Ryan Kidd
·
0
0
View more