Home Archives Search Feed

12-29-2022

Outcomes

Summary

Artifacts

uid: 202212290801 tags: #journal

February 22, 2023

# Extraordinary claims require extraordinary evidence

source: https://ift.tt/w0Yyi8F tags: #literature uid: 202212101125 — In the context of the bayes theorem 202212092211 readings.

What makes evidence extraordinary?

The likelihood ratio is defined as:

Likelihood ratio = Probability of seeing evidence, assuming claim is true
                    ----------------------------------------------------
                    Probability of seeing evidence, assuming claim is false

To obtain an extreme likelihood ratio, the bottom of the fraction has to be very low. The top of the fraction being very high doesn’t help much. If the top of the fraction is 99% and the bottom is 70%, that’s still not a very extreme ratio, and it doesn’t help much if the top is 99.9999% instead.

So to get extremely strong evidence, we need to see an observation which is very improbable, given “business as usual,” but fairly likely according to the extraordinary claim. This observation would be deserving of the title, “extraordinary evidence”.

What makes claims extraordinary?

An obvious next question is what makes a claim ‘extraordinary’ or ‘ordinary’. This is a deep separate topic, but as an example, consider the claim that the Earth is becoming warmer due to carbon dioxide being added to its atmosphere.

To evaluate the ordinariness or extraordinariness of this claim:

  • We don’t ask whether the future consequences of this claim seem extreme or important.
  • We don’t ask whether the policies that would be required to address the claim are very costly.
  • We ask whether “carbon dioxide warms the atmosphere” or “carbon dioxide fails to warm the atmosphere” seems to conform better to the deep, causal generalizations we already have about carbon dioxide and heat.
  • If we’ve already considered the deep causal generalizations like those, we don’t ask about generalizations causally downstream of the deep causal ones we’ve already considered. (E.g., we don’t say, “But on every observed day for the last 200 years, the global temperature has stayed inside the following range; it would be ‘extraordinary’ to leave that range.”)
  • These tests suggest that “Large amounts of added carbon dioxide will incrementally warm Earth’s atmosphere” would have been an ‘ordinary’ claim in advance of trying to find any evidence for or against it - it’s just how you would expect a greenhouse gas to work, more or less. Thus, one is not entitled to demand a prediction made by this hypothesis that is wildly unlikely under any other hypothesis before believing it.

February 22, 2023

Japan Travel Planning

Things to do

  • book airbnbs
    • including inns
  • travel itinerary
  • buy Japan rail pass

Food

Tokyo tips:

  • Shibuya Sky
  • Shopping at shibuya, shinziku
  • Udon shin
  • Ogata

Daniel recs:

  • Teamlab immersive art. I think some times sell out, but it’s been open long enough you can probably buy just a few days out safely. This was so cool that we immediately bought tickets for a second exhibition they have after going (that one looks like it closed) https://planets.teamlab.art/tokyo/
  • Hakone. Think if Lake Tahoe were like an hour and a half train ride from SF. Great views of Mt. Fuji on a clear day. We did a day trip, but there’s ryokans up there too. I really enjoyed the loop this transit pass provides, which includes a pirate ship, several gondolas, plus trains and buses https://www.japan-guide.com/e/e2358_# 19008.html
  • There’s plenty of other great nature like this that’s relatively easy to get to. We thought about going to Nikko, but the day we considering going had a forecast of like 33F and raining.
  • T’s TanTan is great, cheap vegan ramen, conveniently in Tokyo Station, with long hours.
  • It’s kinda a pain to get to, but Naoshima was awesome. Get off the ferry, rent a bike, see a bunch of unique art https://www.japan-guide.com/e/e5475.html
  • Rikugien garden in Tokyo was nice in general, and if you’re lucky and the cherry blossoms come in, has night time viewings where they light up these giant cherry trees. Super crowded for the night viewings, but worth it.
  • Not eating fish would be kind of a challenge, especially if you’re strict about broths, but there’s a lot of food everywhere so you’ll figure it out.
  • I personally got kinda bored in Kyoto after like two days. There wasn’t much to do at night besides eat and drink (whereas we did things like Teamlab at night in Tokyo), and while some of the shrines are really excellent, they kind of all look the same after awhile. It’s worth going, but I think a lot of people end up planning more time than needed there imo.

Japan itinerary 202302121801

uid: 202212061657 tags: #travel

February 22, 2023

Raft & Paxos

Consensus 202211232125

Situations where consensus is important:

  • Leader election
    • In a database with single leader replication, all nodes need to agree on the leader.
  • Atomic commits 202211232128:
    • If you have a transaction spanning several nodes or partitions, the transaction might fail on some nodes but succeed on others.
    • You need to get all nodes to agree that either something went wrong (and we need to roll back the transaction) or nothing went wrong, in which case the transaction needs to commit.

Remember that total order broadcast 202211232155 requires messages to be delivered exactly once, in the same order, to all nodes. If you think about it, total order broadcast is equivalent to performing several rounds of consensus. In each round, nodes propose the message they want to send next and then decide on the next message to be delivered in total order.

However, if consensus algorithms are like total order broadcast, and total order broadcast is like single leader replication (which has a leader, and the leader applies the writes to all the followers in the same order), and single leader replication needs to have a leader… it seems like to elect a leader (have consensus), you need a leader. There is a bit of chicken and egg here.

  • Consensus protocols all use some form of leader, but they don’t guarantee that the leader is unique. Instead, they have several epochs and ensure that the leader is unique within each epoch.
  • Every time the current leader is thought dead, a vote is started among the nodes to elect a new leader. If there is a conflict between leaders in two epoch numbers, the leader with the higher epoch number prevails.
  • Before a leader can decide anything, it must collect votes from a quorum of nodes and see if they respond in favor of the proposal.
  • Therefore, there are two rounds of voting - one to decide on a leader and the other to vote on the leader’s proposal (to make sure there is no leader with a higher epoch value). If a proposal passes, then because of the pigeonhole principle, at least one node must have voted for both the leader’s current proposal and the leader’s initial election. That must mean there is no newly elected leader, and the leader can proceed with their chosen value.

Differences between this and two-phase commit 202211232128:

  1. In 2PC, the coordinator is not elected.
  2. Fault-tolerant consensus algorithms require votes from a majority of nodes and 2PC requires a “yes” vote from every participant.

Zookeeper 202211232246 uses a consensus algorithm like this under the hood.

Drawbacks of consensus 202211232125 algorithms:

  1. The process by which nodes vote on proposals before they’re decided is a kind of synchronous replication. Databases are often configured to use asynchronous replication, in which some committed data can be lost on failover. Many people choose to accept this risk for the sake of better performance.
  2. Consensus systems need a strict majority to operate. That means you need at least five nodes to tolerate two failures, for example.
  3. Consensus algorithms assume a fixed set of nodes that participate in voting, which means that you can’t just add or remove nodes in the cluster.
  4. Consensus algorithms rely on timeouts to detect failed nodes. In environments with variable network delays, this could lead to disruptive false positive timeouts.

Created from: Algorithms to know before system design interviews 202211231052

uid: 202211231100 tags: #distributed-systems #software-engineering #algorithms

February 22, 2023

# Why You Need Systems, Not Goals

source: https://www.instapaper.com/read/1523694907 tags: #literature #living-well #productivity #favorites #goalsetting #advice uid: 202210042101 —

What Is The Difference Between A Goal And A System? Goals are merely a representation of the results you want to achieve. Systems are about the processes. The good daily habits that will lead to those results.

Goals help you set direction, but they are nothing on their own. They are focused on a future state. Systems are about the things you need to do daily, that will make you the person you want to be. They are focused on the present.

“You do not rise to the level of your goals. You fall to the level of your systems.” James Clear

What happens when you set a system to maintain your weight? You daily exercise and you are always in shape. That’s why I define systems vs goals in the following way:

Goals are good intentions with an end date whereas systems are daily actions without an ending.

Systems people succeed every time they apply their systems, in the sense that they did what they intended to do. The goals people are fighting the feeling of discouragement at each turn. The systems people are feeling good every time they apply their system. That’s a big difference in terms of maintaining your personal energy in the right direction.

Direction: Goals are good for setting direction in your life. You can boldly acclaim that you’ll become a writer. By doing so, you’re no longer directionless. You have the base to set daily systems. Decision-making: Goals help you filter incoming information and make decisions. For instance, if you have to pick between jobs, you’ll choose the one that contributes to your goal — a.k.a., the person you want to be. Or, if you’re wondering what to read. You can focus on the material that will get you closer to your goal.

These are the main advantages of setting goals:

Direction: Goals are good for setting direction in your life. You can boldly acclaim that you’ll become a writer. By doing so, you’re no longer directionless. You have the base to set daily systems. Decision-making: Goals help you filter incoming information and make decisions. For instance, if you have to pick between jobs, you’ll choose the one that contributes to your goal — a.k.a., the person you want to be. Or, if you’re wondering what to read. You can focus on the material that will get you closer to your goal.

For instance, I write every morning between 05:00 and 08:00 AM for a simple reason: There are fewer things that can go wrong. Fewer things that can disturb my morning writing session.

It’s quiet. No one to call or message me. My wife and kid are sleeping. My mind is pure. Unviolated from the outside noise — e.g., emails, demands from others, pop-up messages on websites…

Answering “What?” clearly defines what you’re going to do.

Addressing “Why?” is about the person you want to become.

When you define “When?”, it’s no longer a vague claim. You’ve scheduled the activity.

Considering “What’s needed in advance?” sets the scene. This is something I’ve stolen from cooks. In particular, the famous mise en place technique. It’s a culinary phrase that means “putting in place”. In other words, you are organizing and arranging the ingredients needed for the dish you are going to prepare.

This question will guide you: “What’s the bare minimum I can do daily that can have a meaningful result?”

Obviously, writing 2 sentences per day won’t make me a writer. But 500 words? Much better.

Very simple: One that runs daily.

That’s what we are doing with the above questions.

We are preventing failure. Preventing the system from malfunctioning.

When you are creating your system don’t assume you’ll want to do what you’re planning to do. Assume the opposite — assume you won’t want to do what you’ve planned.

February 22, 2023

Thoughts On Twitter

I hate stream-of-consciousness tweets - that’s what journals are for! at the same time, I don’t hate when people share their thoughts, half-baked though they may be. it’s just that twitters structure promotes the banalest tweets. people use it as an insta stories substitute.

Sketch notes

  • So much of discourse on twitter feels asinine.

    • it’s unfiltered, and that’s the point — Facebook / instagram are extremely filtered, people post instagram stories only when they want to show something off — twitter is for simple conversation and connection, and conversations can be about very asinine topics
    • a lot of the comments on reddit are also very asinine, but they are organized around the concept of posts. Posts have two purposes — they ground the conversation on a specific topic, and they make it so you can just scroll past the post if you think the post is kinda stupid.
    • For twitter it’s not like that — the posts themselves are the only things that you care about, so if you keep on getting irrelevant posts, you feel like you’re just wasting your time. There’s also no great way to only see tweets about a certain topic (as far as I can tell), and there’s also no way to filter out asinine, dumb tweets (rather than relying on their algorithm, which I’m not sure what I want to do). On reddit, posts are moderated, which filters out low-quality content. There’s no such way of doing so on twitter, rather than just relying on the algorithm to filter out stuff for you.

  • As a counterpoint, that sort of feels like the point of Twitter?

    can’t believe I managed to significantly elevate my material conditions, find a fantastic group of close friends, lose weight, meet my husband, and overall position myself better than my parents did at my age…mainly by incessantly posting online about my internal state

  • It feels weird for me to throw random undigested thoughts out there. I like sitting on thoughts, digesting them, and making sure that they’re well thought out before posting them. A lot of people on Twitter don’t do that.

    • Counterpoint — a lot of people on Twitter do. You can just ignore the people who don’t.

  • If you think of twitter as a place to interact with friends, and you’re getting a bunch of information from interesting friends, then your timeline will never be boring! This is what 202208262111 is saying.

  • what Pranav ranjit uses Facebook as is basically the point of Twitter. the only difference is that he has Facebook friends that are willing to engage with him in that way, and I don’t. this raises a couple of points:

    • are none of my Facebook friends interesting?
      • I think it’s more that i haven’t talked with any of them in the kind of casual way online that Twitter dictates.
    • people use different social media platforms for different things. you see a totally different side in people depending on the format they they’re communicating in (to an extent).

uid: 202207101701 tags: #blog-ideas #thoughts

February 22, 2023