Top all-time · #5189
Jacob Bayless
Personal blog · jbconsulting.substack.commathscience
Jacob Bayless writes These Are Systems, a personal blog at jbconsulting.substack.com about math and science. Since 2022, 14 different Hacker News readers have linked it 19 times, most often “Inflation, part 1: On shelter futures”. The last post was in 2024, but the archive is still widely read.
Too few sampled comments (7) to say how often HN recommends it.?
Why HN links it
HN readers link Jacob Bayless's control-systems posts on right-half-plane zeros, inflation as a feedback system, housing and interest rates, Kalman filters, and LLM reasoning limits.
Summarised by Claude from a sample of 7 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
A newsletter on probability, applied control theory, and systems thinking.
Most-linked posts?
- Inflation, part 1: On shelter futures7 people · 7 mentions · first linked 2022-06-10
- It's not just statistics: GPT-4 does reason3 people · 4 mentions · first linked 2023-05-24
- How to turn your estimates into reliable predictions3 people · 3 mentions · first linked 2023-01-16
- Is the Kalman filter just a low-pass filter?2 people · 2 mentions · first linked 2022-07-28
What Hacker News says?
The person you're replying to already has answered this question in a post you've already replied to. But to be explicit, consider the first sentence of the follow up post [1]: > Now that you know what right-half-plane zeros are, in this article I'm going to begin a deep-dive into a control-systems-based analysis of a certain feedback system that is drawing particular public attention today: inflation. In my mind, t…
Comment linking /p/on-shelter-futures-part-1?s=r
on “You need to know what right-half-plane zeros are”Submitted You need to know what right-half-plane zeros are as a story · 439 points
HN discussionThe danger of "systems thinking" is that it misleads otherwise clever people into believing they've found a cheat sheet for subjects they don't have deep knowledge in. This article is a fine example of that. I didn't know what right-half-plane zeroes are. I knew some of the examples it gave (Veritasium has a fun video on the bicycle steering phenomenon), but not that there was a category they all fit into. Neat. But…
Comment linking /p/on-shelter-futures-part-1
on “You need to know what right-half-plane zeros are”Submitted Is the Kalman filter just a low-pass filter? as a story · 267 points
HN discussionLink, because I assumed it wasn't out yet since the author hadn't provided it: https://jbconsulting.substack.com/p/on-shelter-futures-part-...
Comment linking /p/on-shelter-futures-part-1
on “You need to know what right-half-plane zeros are”Submitted Inflation, part 1: On shelter futures as a story · 157 points
HN discussionThe follow-up article on inflation of shelter costs in response to mortgage interest rates is also fascinating and worth a read, imo [0]. The tl;dr is that in the medium and longer term, higher interest rates will decrease inflation, but there's a right-half-plane zero effect that will cause higher inflation rates in the short term in response to interest rate hikes (with a lot of simplifying assumptions). [0] https…
Comment linking /p/on-shelter-futures-part-1
on “You need to know what right-half-plane zeros are”Submitted It's not just statistics: GPT-4 does reason as a story · 3 points
HN discussionNot OP but here's an example of how GPT-4 can't deal with the goat/wolf/cabbage problem when things are switched up just a little. https://amistrongeryet.substack.com/p/gpt-4-capabilities Although it's interesting that if you use different nouns it does just fine: https://jbconsulting.substack.com/p/its-not-just-statistics-...
Comment linking /p/its-not-just-statistics-gpt-4-does
on “GPT Best Practices”Submitted How to turn your estimates into reliable predictions as a story · 3 points
HN discussionIt is only difficult for a LLM to sort a list of numbers if the list is longer than half of the context window. (Source: I tested this myself[1]). The sorts are not error-free every time, but with sufficient training they become error-free the vast majority of the time, even for long lists. This is not especially surprising because transformers are capable of directly representing sorting programs.[2] [1] https://jb…
Comment linking /p/its-not-just-statistics-gpt-4-does
on “Graph of Thoughts: Solving Elaborate Problems with Large Language Models”Submitted Inflation, part 1: On shelter futures as a story · 2 points
HN discussionThis blog had a great read that I half-understood on how inflation is already baked into a relatively long-term time horizon due to how interest rates affect housing costs. https://jbconsulting.substack.com/p/on-shelter-futures-part-...
Comment linking /p/on-shelter-futures-part-1
on “U.S. Inflation Eased to 5% in March”Submitted Task Estimation: Conquering Hofstadter's Law as a story · 1 points
HN discussionAn interesting take awhile back on this exact issue: https://jbconsulting.substack.com/p/on-shelter-futures-part-... . The fear (now largely realized) is that increasing interest rates mechanistically drive up inflation (because of how the price index is calculated in Canada), which the Bank of Canada combats by...raising interest rates. Includes a good discussion of just how potentially screwed the housing market i…
Comment linking /p/on-shelter-futures-part-1
on “Bank of Canada raises its interest rate to 5%”> It is absolutely trivial to show Hyp2 is false To investigate precisely this question in a clear and unambiguous way, I trained an LLM from scratch to sort lists of numbers. It learned to sort them correctly, and the entropy is such that it's absolutely impossible that it could have done this by Hyp1 (sampling from similar text). https://jbconsulting.substack.com/p/its-not-just-statistics-...
Comment linking /p/its-not-just-statistics-gpt-4-does
on “From Word Models to World Models”For software estimation, you can get a long way by just assuming that the distribution is lognormal. https://jbconsulting.substack.com/p/task-estimation-conqueri...
Comment linking /p/task-estimation-conquering-hofstadters
on “Rules of Thumb for Software Development Estimations”Another good article on the subject: Is the Kalman filter a low-pass filter? Sometimes! https://jbconsulting.substack.com/p/is-the-kalman-filter-jus... PS: I've used it to remove jitter in virtual camera movement while cropping video around faces in real-time, streaming detected face locations to a Kalman filter worker and get back stream of stable camera locations.
Comment linking /p/is-the-kalman-filter-just-a-low-pass
on “Kalman Filter Tutorial: Kalman Filter from the Ground Up”I use calculus quite a lot, basically any time I need to go beyond an existing solution I could look up in a textbook. That is, all the interesting problems! For example in writing this article, especially details in footnotes 4 and 13: https://jbconsulting.substack.com/p/voiding-the-warranty-how... When I want to write or tune a Kalman filter or some other control system, I need to work out the differential equatio…
Comment linking /p/voiding-the-warranty-how-the-kalman
on “The Calculus They Won't Teach You [video]”
