Hacker News Reader

samscullyon Why aren't smart people happier?

The point is that you actually cannot explore a spectrum of ideas without committing immediately because you are forced to live and living forces you to commit and make moral decisions every day.

samscullyon A breakthrough towards the Riemann hypothesis

For anyone looking for an introduction to the Riemann Hypothesis that goes deeper than most videos but is still accessible to someone with a STEM degree I really enjoyed this video series [1] by zetamath.

I understood everything in Profesor Tao's OP up to the part about "controlling a key matrix of phases" so the videos must have taught me something!

[1] https://www.youtube.com/watch?v=oVaSA_b938U&list=PLbaA3qJlbE...

samscullyon Cutting a 700 carat rare valuable gemstone [video]

This video [1] shows laser cutting, 3D modelling, and hand grinding of diamonds in a diamond factory. They discuss when and why they use laser cutting machines a bit.

[1] https://www.youtube.com/watch?v=YMyPo-MEvMA

samscullyon Jaccard Index

The book Mining of Massive Datasets [1] has useful information on building an efficient similarity index using Jaccard/minhash. I would also recommend Otmar Ertl's papers on extensions of minhash that approximate Jaccard better in certain situations, e.g. superminhash [2].

[1] http://www.mmds.org/ Chapter 3 [2] https://arxiv.org/abs/1706.05698

samscullyon Blue Zones, where people reach age 100 at 10 times greater rates

It does appear to be true for humans. Though interestingly there isn't a correlation between species size and cancer risk [1].

[1] https://en.wikipedia.org/wiki/Peto%27s_paradox

samscullyon My Experience as a SWE Intern at Goldman Sachs

There are infinite shapes that cannot fall down the matching hole, but the circle is the easiest to manufacture. A circular cover is easier to manufacture than e.g. a Reuleaux triangle shaped one.

samscullyon MIT moves all classes online for the rest of the semester

My halls at Imperial College had double and triple shared rooms [1]. Looks like they don't use those buildings as halls any more though.

[1] https://www.imperial.ac.uk/centenary/memories/JoaoCabral.sht...

samscullyon SEO case study: 8k to 200k monthly organic traffic

Here are some notes I made when I worked for a company that did very well in SEO. This was a few years ago now, so this might be horribly out of date now.

Hygiene factors

- Never expose an error page to Google. - Never expose an empty page. - Make sure all content is visible on first page load (e.g. not behind pop ups). - Show the same content to Google as to users. - Make sure all relevant content is in the HTML (not loaded dynamically). - Make sure page load times are good.

Keywords

- Research keywords you need to match. - Build database of search terms you wish to match, grow and improve this database over time. - Carefully curate pages, their URLs and metatags to match these keywords. - Use this database to choose which pages to expose to Google and make sure that all links and pages have the correct follow/nofollow attribute. - Steadily grow the content you expose to Google over time, keeping the quantity and quality of content on pages exposed to Google high.

Site structure and internal links

- Organise your site in a shallow tree structure. Top level pages should have many links to lower level pages. - All levels of the tree should have keyword rich content which is useful to users. - Not every level of the tree needs to be visible in the Google search results, it is mostly to make things easy for the crawler.

External links

- Use PR and social media to generate good backlinks for your site. - Use social media like reddit, pinterest and others as well as facebook and twitter. - Do not generate fake backlinks, if Google thinks you are doing this they will give you a severe penalty.

Submit your sitemap

- Submit your sitemap in Google webmaster tools and make sure it is kept up to date at least weekly.

Bounce rate

- Carefully monitor your bounce rate and time on site for users coming from Google as these are key factors in Google rankings.

Click through rate

- Use descriptive and relevant titles and metatags to improve the click through rate from the Google search results to your site. - PR, social media, advertising and other channels can improve brand awareness that in turn improves click through rate. - Regularly check a selection of searches that you rank on to assess the quality of your Google search listings.

samscullyon To Break Google’s Monopoly on Search, Make Its Index Public

The ad business would pay the search business for ad placement. The search business would also be free to auction ad space to other providers.

samscullyon My low cost provider of GPUs has run out of capacity. Good alternatives?

Vast AI (https://vast.ai) are good, 1x GTX 1080 comes to about $170/month. You are charged by the hour unlike Hetzner.

There are obviously downsides to renting time from random individuals though, so it's not suitable for a server-like workload. Good for development notebooks or training.

samscullyon How to pick a random number from 1-10

Empirically, from running a quick script on the data in the article, it seems like you only need to sample 5 numbers to get a distribution that looks uniform using this "sum mod 10" method.

samscullyon VMware to Acquire Bitnami

I think they might mean "double trigger" acceleration, where the options vest automatically on change of control.

samscullyon The Brexit Short: How Hedge Funds Used Private Polls to Make Millions

You pay the person you borrow the stock from a fee.

samscullyon Small ads sex trafficking: the battle against Backpage

A core Libertarian value is individualism. So something like trafficking would be anathema to any Libertarian. Furthermore, I'm sure many would acknowledge that under a certain age it is not credible that any child has the ability to choose prostitution.

I'm not a Libertarian myself, but wanting to reduce the power of the state doesn't mean you believe in slavery.

samscullyon Ask HN: A good primer on cryptocurrencies?

I've been reading "The Business Blockchain" [0], it's been good so far.

[0] https://www.amazon.com/Business-Blockchain-Practice-Applicat...

samscullyon Differential Privacy

The thresholdout method [0] for preventing overfitting on a test set is an interesting application of this.

Here's a talk on differential privacy applied to the overfitting problem [1]

[0] http://andyljones.tumblr.com/post/127547085623/holdout-reuse

[1] https://www.youtube.com/watch?v=9mqXjdnZA18

samscullyon How much do hedge fund traders earn?

There is very little demand for professional athletes so training to a very high level in mathematics or finance is a much less risky proposition.

samscullyon A faster BBC News front page

Recently working on an embeddable shopping widget I noticed that if I animated it in certain ways chrome would block it as a popup.

Doesn't contradict what you said I just thought it was interesting that chrome doesn't just look for the most basic types of pop-ups.

samscullyon Chase had ads on 400k sites, then on just 5k, with same results

It's not a protection racket but it is close to becoming a monopoly, which is also illegal in some circumstances.

This sort of anti-competitive coercion through overwhelming market power is why monopoly laws exist in the first place.

samscullyon London Taxi Company opens £300M Coventry plant for electric cabs

In three years only emissions-free vehicles will be allowed in central London.

samscullyon Deep learning with text: Learning when to skim and when to read

No usually

You did that on purpose ;)

samscullyon To fix L.A.'s traffic, we need tolls

I've never been to LA so I'm speaking from ignorance but some of the objections raised in this thread seem like they have obvious solutions to me.

1) Tolls are regressive. I believe cities in the US have power over some taxation so why not offset the extra cost by reducing another tax? e.g. property taxes in low value areas could be cut, or sales taxes. Alternatively spend all the toll revenue on heavily subsidising public transport so that it is cheap and plentiful.

2) Other cities have had tolls and it did not effect the traffic. Why not just raise the toll until it does effect the traffic? In London the charge for breaking Ultra-Low Emission Zone requirements will be £130/$160 after 2020. Raise the toll to $100 and it will have an effect.

Sounds though that the underlying issue is that LA's urban geography doesn't lend itself to public transport so that whatever you do people will still have to use cars if they want to travel any reasonable distance. Short of undertaking a Shanghai-metro level of construction (from nothing to the world's largest rapid transit system in 20 years.)

samscullyon Sputtering Startups Weigh on U.S. Economic Growth

It depends on the town a bit I think. In Brighton there are lots of small business and some areas have no chains at all. Not sure if that is because of planning policy or for another reason.

samscullyon How Coderwall grew over 50% with an A/B test

Found my notes on difference in differences method. First model the sessions originating on page [i] at time [t] as:

    S[it] = b0 + b1 * X[it] + b2 * A[i] + b3 * D[it] + e

Where:

    S[it]: number of sessions originating on page [i] at time [t]
    b0: base traffic for all pages
    A[i]: base traffic for page [i]
    X[it]: seasonal traffic for page [i] at time [t]
    D[it]: dummy variable, 1 for after treatment 0 for control and before treatment
    b3: effect size of treatment
    e: noise/error term

Taking the average over time:

    E[S[i]] = b0 + b1 * E[X[i]] + b2 * A[i] + b3 * E[D[i]] + e

Take the average from S[it] (the first "differences"):

    S[it] - E[S[i]] = (b0 - b0) + b1(X[it] - E[X[i]]) + b2(A[i] - A[i]) + b3(D[it] - E[D[i]]) + e

which simplifies to:

     S[it] - E[S[i]] = b1(X[it] - E[X[i]]) + b3(D[it] - E[D[i]]) + e

Split all pages [i] in to control and treatment groups 0 and 1. Take the "difference in differences" of the two groups:

    (S[1t] - E[S[1]]) - (S[0t] - E[S[0]]) = b1(X[1t] - E[X[1]] - X[0t] + E[X[0]]) + b3(D[1t] - E[D[1]] - D[0t] + E[D[0]]) + e

Given D[0t] = E[D[0]] = 0, and for a large number of pages X[it] - E[X[i]] = X[jt] - E[X[j]]:

    (S[1t] - E[S[1]]) - (S[0t] - E[S[0]]) = b3(D[1t] - E[D[1]]) + e

Which gives a formula in the form of y = mx from which we can calculate the effect size b3 and the p value using linear regression.

The key assumption is that X[it] - E[X[i]] = X[jt] - E[X[j]] for large numbers of pages. I found this assumption does hold in practice. I'm not a statistician so forgive me if I've made any statistical errors but I think this analysis is correct.

I might write up a blog post at some point to explain it in more detail.

samscullyon How Coderwall grew over 50% with an A/B test

You can calculate the significance using the difference in differences method popular in econometrics. I derived this result for a previous employer and it worked well. I can look up my notes and explain the method in more detail if anyone is interested.

samscullyon Why GitHub Finally Abandoned Its Bossless Workplace

Depending on the business, there is usually a requirement for a coordination role. Whoever that coordinator is needs to have the authority to make a final decision in order to be effective.

If companies want to avoid tyranny in the workplace then perhaps it would be better to have bosses but allow a majority vote of direct reports to fire them. Give 360 reviews some actual teeth.

samscullyon Google Brain releases two large datasets for robotics research

Babies take a while to learn how to grasp things, it's believable that the number of attempts they make is in the 10,000s. 1 grasp attempt a minute, 8 hours a day for 3 months is 43,000 attempts.

Of course they are learning a lot of other stuff at the same time so it's not really comparable.

samscullyon Optimism

I have not read the book so I can't comment on that but I have done a good deal of cognitive behavioural therapy. Mostly I was taught to consider the optimistic explanation and assess how realistic my negative thoughts were.

The main exercise we used was to write down my thoughts, feelings, physiology and actions in response to a situation. Then we did it again for a more positive imagining of how the same situation could have played out. It's useful to see how each aspect effects the others and identifying how things could be changed helps a lot when the same situation arises again.

For sure it wouldn't be helpful to be wildly over optimistic, it must be very stressful to intentionally delude yourself. But the error bars on most of these judgements are so wide I don't see anything wrong with erring on the optimistic side.

Of course another skill is learning to avoid unhelpful judgements altogether. But that's a different topic and I won't pretend I have much insight there.

samscullyon Don't add your 2 cents

Some factors in motivation at work are the level of autonomy, mastery and purpose in your job.

A coworker giving minor feedback is only contributing to your mastery. A boss giving you the same minor feedback is cutting in to your autonomy. The exception is when the purpose is great and sweating every detail is necessary or when the boss is a recognised master of your craft and their feedback is almost always correct and regularly helps you improve.

An example of the first case might be engineering at SpaceX and the second could be Steve Jobs giving engineers and designers product feedback. What I think a lot of people are missing in this thread is that in most situations the purpose is relatively uninspiring and the boss is significantly less skilled than the person she is giving advice to.

samscullyon British MPs’ private emails are routinely accessed by GCHQ

Fortunately the non-establishment groups sometimes infiltrate and manipulate the establishment too [1].

[1] http://www.theguardian.com/politics/2015/oct/17/jeremy-corby...