Show HN: 100M Books – Open a new tab, discover a new book
100millionbooks.org
100millionbooks.org
For a start, people don't want to be challenged. I'd love to know how much of the bubbling is self imposed. E.g. Ignoring everything else until Google, etc, gets the hint.
That and the fact that everyone seems to feel time poor. Sorting through what's valuable and what's not is a burden. I've tried handling the firehose new Hacker News, or even worse the RSS feeds for all the sites that typically posted. Doesn't last long.
Having said that there was an interesting paper on using Machine Learning to summarise the other day. I wonder if that could help with the snippets problem.
As the internet and it's use has trended away from user driven discovery and toward curated feed driven discovery, it's become harder to pull yourself away from the feed of just-okay content and the content is in a very narrow band of safe vs interesting. It's also almost never productive, just mildly interesting.
My mind is bored and I know that's not from lack of available subjects. It's just that the pipes and feeds I set up have become stale. I don't believe more pipes are the answer, I believe a shift in paradigm is needed for me. Cut the pipes and retreat back to a curiosity driven discovery approach. HN doesn't know what I'm curious about at any given moment, nor can it know what discoveries would drive my development and progress forward, so it doesn't suffice.
I slowly let feeds and aggregators replace my curiosity and now I am out of practice. I'm not saying I need to replace HN and others, but certainly to be mindful that I shouldn't forget to find things for myself and limit how much of my mindshare I give to the aggregators.
All this is grand exaggeration and I am much more functional than that makes it sound, but it's a real problem and I think many others are in the same boat.
Regarding value: this is a complex subject. Within bubbles, I think value outside bubbles is often unfairly minimized. See American politics, for example. Most Democrats think ALL Republicans are stupid science-hating climate-change deniers, while most Republicans think ALL Democrats are history-deaf socialists. Neither mindset is true, but you'd never know it if you remain trapped in the other bubble.
But on a personal level, everyone's different. I have an internal alarm that goes off when I feel like I've been thinking a certain way for too long.
It's going to take a long time to do all 100 million.
I personally prefer (and think people will benefit much more) from the latter.
FEEDBACK EDIT: unless, I dunno, am I wrong? Would you rather see a random book cover with the generic Amazon/Goodreads description? You'd get much more variety that way.
- What is a snippet? 'Use your judgement' isn't a scalable option.
- What are the measures against abuse? How is validation occurring? (ie: how can you say a snippet is from a book the submitter claims to be)
- What books qualify? I've read some deeply technical books that contain interesting ideas that go beyond the subject matter.
- Google Books search usually works well.
- Anything goes. I don't underestimate a layman's ability to understand (or at least somehow benefit) from highly technical work [1]. Obviously, I don't want technical snippets to be over-represented, but I keep track of snippets by subject, time period, and other metrics. I'll open up access to these metrics when it makes sense to do so.
[1] ever seen Sugata Mitra's TED talk? https://www.ted.com/talks/sugata_mitra_shows_how_kids_teach_...
I'm not sure how well GEB has aged though since it seemed very tied to a particular point in AI history. Has anyone read it recently?
There's an excellent MIT course for it. All the Lectures on Youtube as well.
https://ocw.mit.edu/high-school/humanities-and-social-scienc...
If you want heavy-but-popular, try Foucault's Pendulum or House of Leaves …
My favourite bits were a dialogue on the six part ricercar and a piece of music that can shake the player to pieces, it's full of playful but profound little ideas, definitely worth reading.
The dialogues (Achilles & Turtle) between the "essay" chapters are insanely enjoyable if you like wordplay and divertissement writing. "Crab Canon" is amongst the pages of literature I can re-read ad infinitum.
It gets somewhat harder to follow as it proceeds, and the lack of a "plot" makes it easy to abandon it, but I'd still recommend it.
For the gist of it, I recommend I Am a Strange Loop.
Some Kindle highlights are publicly available on
...but that site feels abandoned and is hit-or-miss for most books.
So I made this to solve that. To each their own!
Regarding distraction...it's a personal thing. I've been running the extension on my own browser for a few days and I usually just ignore it when I open a tab with purpose.
(Parenthetically, it reminds me of the advice Charles Olson wrote to a younger writer: "Best thing to do is to dig one thing or place or man until you yourself know abt that than is possible to any other man. It doesn’t matter whether it’s Barbed Wire or Pemmican or Paterson or Iowa. But exhaust it. Saturate it. Beat it. And then U KNOW everything else very fast: one saturation job (it might take 14 years)".)
And a good chunk of the existing collection is already sourced from others who've used my previous book apps.
The process is still young so it's imperfect, but improving. Do you have any suggestions how I can do better?
Speaking of humility, it must've taken a good lack of humility on your part to make such assumptions without knowing the details of my process or the background of this effort.
Publish your book list, your selection process, and your metrics, to enable independent evaluation. I looked for such information and didn't find it; if it's there, surface it clearly. If you're presenting the balanced, apolar perspective you claim to seek, rather than some flavor of "alternative facts", this should not pose an issue, I think.
As for the rest, I wouldn't feel special about it. In these times, I see no reason why anyone claiming political motives, and apparently declining any meaningful transparency in sources and methods, merits any kind of credibility. Your claims are your claims, and prove nothing beyond that you've made them. You have an opportunity to substantiate them. Perhaps you will do so.
I noted elsewhere in the thread that I plan to open access to that information when it makes sense. This is a weekend project and I really didn't expect this much attention this soon!
But you made me realize just how important this issue really is. So I just added a Transparency section (link on very bottom of main page) with the titles of all books currently in the system, along with a link to a public Google Form showing all suggestions and how they were handled. I was using Typeform before but results were private.
It's not as thorough as what you had in mind, but I'll put that in place over time.
You might argue that this is against the idea of your site (you seem attached to the serendipity idea) but, if you are crowd-sourcing books and quotes, you'll end up getting something very similar to this with a lot more work. Look at the books and quotes you already have -- they are all in the top few percent of Goodreads, and the quotes have already been selected.
The point is not that you don't want to use algorithms; the point is that you want an algorithm that allows for serendipitous exploration.
All very easy to do! Am happy to do myself if you put the site up on github.
As long as quality remains high and variety is verifiably strong, it doesn't matter how the quotes are obtained.
This was more of a proof-of-concept, and I already had a database of books from a previous project, so it was a convenient starting point.
Small suggestion: Slow down the image carousel on the homepage it moves to quickly to process what it's showing you. Also it would be neat to able integrate this into other projects, for example I like https://momentumdash.com/ (unaffiliated user) and I'd rather have this than a quote of the day or what not.
I saw somewhere that official support should be coming in Firefox 57. I plan to implement as soon as it's available.
And a perfect use-case for Service Workers... even suitable for a Service Worker tutorial.
- I love it, it is making me so hungry to read more and the snippets are great.
- I hate it, because now my reading list is going to go from "I'll only ever finish this in my retirement" to "there is no point to keeping a reading list as its now 100,000,000 books"
I really enjoy it, the only issue is that it is a bit of a distraction when I am about to do something.
1) Pre-fetch the data such that the page can be immediately rendered when opening a new tab. Seeing a 1+ second loading animation for something done as often as opening a new tab introduces unnecessary friction.
2) Retain a history such that the user can see which books they have had shown to them so they don't suddenly lose a suggestion by closing a tab.
1) agreed, will implement this in a future release.
2) books cache for ~10 seconds for this reason, but yes i see why history would help. it's a popular request so i'll also add this in a future release.
My hunch is that people don't want to go out of their way to discover stuff. Hence the extension...we're always opening new tabs!
The only issues I have is that opening a new tab now makes me forget why I opened a new tab for, since I immediately see new content.
I also don't like the 1s latency and the little circle animations until the book info shows up. Is there a way of removing that?
Set as your home page: https://en.wikipedia.org/wiki/Special:Random
See the icons from my pages and see which pages I visited ?
The #2 feature request was to show Chrome's Top Sites list somewhere on the page (you know, those sites Chrome shows by default on the New Tab page). So I implemented it, and that required a new permission called 'topSites' which sounded innocent, so I did it.
I guess I should've looked into the details, because I had no idea it would ask users to read all sites they've ever visited...all I needed to see were the top 8 (in order to show links, not harvest the data).
But still that's a no-no. I can see why it'd sound shady, and I'd be wary of granting such permissions myself. The people who wanted that top sites list will have to get by some other way.
In hindsight, I should've done more research before requesting the permission : / Lesson learned.
https://chrome.google.com/webstore/detail/juicy-drops%E2%9D%...
*We were pleasantly suprised that people used it as a marketing research helper.
Otherwise great idea.
I am big fan of momentum homepage.
It showed me the same Bruce Lee book three times within 5 minutes.
Snippets cache for ~10 seconds. Also keep in mind the library of snippets isn't huge yet, and selection is totally random, so there will be repetition.
Something like repicking a book when it has been shown t seconds ago would mitigate this problem.