Hacker News Reader

Comment by amelius | Hacker News Reader

We need some strong AI to strip away the ads, and extract only the main text, then format that text using the reader's preferences.

OCR 2.0, if you will.

Caveat: it may kill the advertisement industry.

View on HN

The FYWD browser:

https://ello.co/dredmorbius/post/ubkidr7yuc7azg9hdnl7lq

The premise: A web browser aimed at information presentation, largely textual. It will offer a standardised format, non-Javascript, uniformly presented readable Web page, not intended for apps, but a sane set of presentation defaults for a given site.

Site-offered CSS would be ignored (or at best strongly deprecated). Straight-bog HgsML 1, 2, / HTML5 sites would render best. User could opt for specific client-side stylesheets, e.g., night mode, large/small font, etc.

That doesn't require strong AI, firefox reader view achieves that without strong AI. It even preserves article images nicely.

Reader view isn't perfect yet, which is probably why (useful as it is) most people don't use it.

Also, as more readers use it, the ad industry will try to circumvent it, e.g, by tricks like turning part of the text into images, et cetera.

Eventually, you will probably be better off by OCR'ing the rendered webpage (i.e., as the user sees it), and using AI to only keep the relevant text and images.

Don't let perfect be the enemy of the good.

It's not perfect so people don't use it?

No, people don't use it because it has poor discover-ability and it isn't in other browsers, further reducing discover-ability.

> It's not perfect so people don't use it?

Unfortunately, I think this is the case. Look at DuckDuckGo. It works for 90% of searches, but for that last 10% you really need Google. As a result, people use Google for all their searches.

I think this is another case of poor discover-ability. If you "discover" and understand how the bang syntax works, you can easily use DDG for your 90% of searches and only dispatch to Google when you need to. (e.g. "!gi dogs").

DuckDuckGo is great, but unlike Google which has immense brand momentum and recognition, it lacks the marketing to drive users to the service.

Not sure if that's how it works. I guess users would start to use "!g" syntax on more and more of their searches (as a habit).

See also: Readability, a project first discussed on HN some 8 years ago: https://news.ycombinator.com/item?id=501970

Does it perform OCR? Because if not, then the content industry will circumvent it by rendering the text as images.

Content industry doesn't really care that much about few nerds scraping the content for themselves as long as it remains small enough minority. And if they do something, it'll most likely follow the tracks of other anti-adblock tech instead of doing something as stupid as prerendering text.

killing the ad industry means killing the current web unless there is another revenue stream

It's that so? My experience says otherwise. For example, hn will stay, and so will the linked blog. They don't show ads.

HN is a meta-site. Without ads, there will be dramatically less content to link to.

I marked all the stories that I think are not dependent on ads: https://sli.mg/ztY27D

18 out of the 30 stories on the first page do not depend on ads. That's good enough for me, even though many interesting things are lost.

How so? Without ads, the same money will be looking for ways to be spent.

The overall efficiency will go up as we'd save the overhead for ads, be less manipulated, can trust the information better ("is this an ad?") and a few shitty business practices will go away.

Less fluff content perhaps but I remember lots of people wrote great content before ad networks were easy to include on a site. I write content, I'm biased I like my own stuff, but I don't expect revenue and I know plenty of other people would continue to blog without revenue from ads.

That's a good thing.

Information as a public good, financed by an income / wealth indexed content tax, pro-rated to content authors, and/or basic income.

https://www.reddit.com/r/dredmorbius/comments/1uotb3/a_modes...

Rationale for information as a public good:

https://www.reddit.com/r/dredmorbius/comments/2vm2da/why_inf...

And from Hal Varian (UC Berkeley / Google):

http://people.ischool.berkeley.edu/~hal/Papers/japan/japan.h...

Why micropayments (even with repudiation) won't work:

https://www.reddit.com/r/dredmorbius/comments/4r683b/repudia...

The problem with advertising-financed publishing (from 1909):

https://archive.org/stream/commercialismjou00holtuoft#page/0...

To me, that sounds a whole lot like, "shooting a zombie in the head will stop it from getting up again".

We had this. It was amazing. It was called textmode.org. You pasted the URL after the slash and it would give you just the text, e.g., http://textmode.org/http://website-a-b-c.com

I died a little inside the day it went down.

Was it open source? If so it should be possible to run it locally. Might be an interesting project for someone.

But of course you can go a long way toward it just by switching to View|Page Style|No Style in Firefox.

Why stop there, implant it directly into the reader's brain.

Reply on news.ycombinator.com