Bionic Reading – Convert Text into Better Way to Read Faster | Hacker News Reader

Bionic Reading – Convert Text into Better Way to Read Faster

github.com

71 pointsandsoitis61 comments

https://news.ycombinator.com/item?id=30787290

Since it isn't mentioned this is inspired/based on https://bionic-reading.com/ (discussed in https://news.ycombinator.com/item?id=30787290). Note, although it has been granted patent only in France, the Bionic Reading name is registered trademark. A web extension was made soon after the link was posted and can be found in https://github.com/ahrm/chrome-fastread (Firefox vers: https://github.com/akay/firefox-fastread).

Oh god.

I know you're (presumably) not the author of the website, but if you're going to produce a method based on the theory of faster/easier reading, it helps to not have a landing page that has up front in big massive fonts a sentence that my mind parses as:

"Did you know that we humans store learned words and so just a few letters your brain reads faster are enough to recognise whole words. than your eye?"

How can one take it seriously when the very first example on the webpage fails on reading cognition!

When I first opened the submitted link I thought bionic reading was about overlapping sentences. They should just remove this section and instead use the first pic that showcases what they're doing.

If you want to try it out on a PDF, I have implemented this in sioyek PDF reader:

https://github.com/ahrm/sioyek

You need to enable experimental features in `prefs_user.config` by adding `enable_experimental_features 1` and then enable this mode using `toggle_fastread` command within sioyek.

I also experimented with more sophisticated algorithms, you can read about the experiment here:

https://ahrm.github.io/jekyll/update/2022/04/14/using-langug...

Interesting. There's probably still more that could be done.

I'm probably weird but since I was a kid through high school (1960s-1970s) I was a subject of various research studies and advanced learning techniques. I don't exactly know why I ended up in these but it's probably "right place, right time".

A lot of these involved "speed reading" as well as "speed learning". What this code does definitely broaches some of what I was taught and what was at one time more widely known. There are other aspects that are missing that relate to typography and dynamic reading.

Kind of similar to this bookmarklet:

javascript:(function(){var q="p, title",e=document.querySelectorAll(q),o=[],str1="",str0="",str="",j=0,pivotchar=0,finvar="",ans="",d=window.open("","_blank"); for(var i in e){var t=e[i].textContent; if(t){o = o + "\n" + t;}} str0 = o; str1=str0.replace(/\n/g, " "); str=str1.split(" "); for(j=0;j<str.length;j++) { pivotchar=Math.floor((str[j].length)/3)+1; finvar = "" + str[j].substring(0,pivotchar) + "" + "" + str[j].substring(pivotchar,str[j].length) + "" + " "; if(str[j].substring(str[j].length-1,str[j].length)=='.') { finvar=finvar+" * "; } ans=ans+finvar; } d.document.write("<html>"+ans+"</html>");})()

https://www.locserendipity.com/Hyper.html

This bookmarklet is a little jewel.

I saw this pop up on Twitter a couple of days ago and made a JS version with a variable font. https://codepen.io/onion2k/pen/qBxmVpR it's a fun idea, but I don't really find it helps me read things.

It seems like bionic reading highlights the first syllable of each word, while your version only highlights the first letter. It does make a difference, since the first syllable helps guess the word.

Edit: Other comment pointed to this bookmarklet, which does that: https://www.locserendipity.com/Hyper.html

I tried the samples on the bionic reading website, I don’t feel like this makes me read faster. Am I the only one?

Yeah it makes reading significantly harder for me. I recognize what they're trying to get my eyes to do but they already do that anyways and bolding half of everything on the page makes it harder for me to pick out the bits that matter. Bolding any part of "is", "in", "a", etc. is a hilarious choice to me.

Nope. This does nothing for me.

Not just you. I don't get the point of this either. I'm guessing it has something to do with your existing reading method? My eyes naturally jump to the first letters and tallest/round characters when determining the shape of a word.

There is an accelerated reader I played with a few years ago, based on flashing the words for you with some letters colored in red. That increased reading speed mildly, but the lack of ability to quickly backtrack defeated any gains.

No, you're not. I can't even read it. I can read colour-coded (i.e. syntax highlighted) text perfectly well though. This just completely shorts out my reading - might as well be hieroglyphics to me.

It does for me. The bold "anchors" make my eyes wanders less. English in not my native language, maybe that helps.

How the hell did they manage to patent this? At least in Germany it is still "pending". I had the strong impression that you cannot patent software or algorithms (at least in Germany).

From what little I know, in the US you can only file patents for "method and apparatus". I could be wrong but I think this is a very common restriction in patent systems around the world.

Suppose they file this patent for an ebook reader ("apparatus") with this specific feature implemented in a similar way ("method"). You would be in clear violation of the patent if you were to build and distribute a competing ebook reader with a substantially similar feature. But protection drops off rapidly the further you go from replicating both method and apparatus. I.e. if you only work on a piece of software (absent the ebook reader apparatus) I think you should be fine (but IANAL).

Now suppose you were to publish an Android ebook reader app with this feature, and publish it via Google Play. Supposedly this would result in a combined software and apparatus in violation of the patent. I'm not clear on how this is usually regarded (you'd have to at least worry about patent trolls I assume) but I doubt the patent owner has a legal leg to stand on (which won't necessarily stop them from trying).

This simply doesn't work for me. Maybe it's because I'm Autistic, but I can't even read text highlighted in this way. A friend posted up a side-by side example, and I couldn't read the first line in the time it took me to read the whole, plain passage. My eye simply won't follow the line - maybe I am getting distracted by the odd formatting. As a coder - used to syntax highlighting - this was very surprising, but for me, this is absolutely terrible.

I wonder if the effectiveness of bionic reading would be improved by using a gradient of weights for each character - instead of just bolding the first half of the word, as this extension does.

For example, Open Sans has six levels of weight: light, regular, medium, semi-bold, bold, and extra bold:

https://fonts.google.com/specimen/Open+Sans

there’s already beeline that does this

https://www.beelinereader.com/

They ask for money. I am not sure about their claims to make me read faster. I couldn't proceed beyond installing extension and then uninstalling it.

BeeLine creator here! Some of our tools are paid (with a free trial), and others are completely free. [1]

You can also use BeeLine for free via Blackboard's File Transformer, [2] on the Bookshare platform, [3] and in various other education platforms.

1: https://apps.apple.com/us/app/beeline-extension/id1571623734

2: https://ally.ac/covid19/

3: http://www.bookshare.org

Saw this on reddit days ago, still waiting for someone to show any kind of scientific evidence for it's usefulness.

Agree. Would love to see some text even explaining what this is or why it's supposed to help. I just see a git repo with no explanation...

As mentioned in one of the comment this is linked to a certain methodology of teaching reading.

One where the reader is more focused on the overall shape of the word rather than the individual letters (from memory I think it is the "global reading method"). I think it can have great effect on the speed of reading, but downside effect is less accurate spelling. Also when encountering "Exotic" words reading speed can slow to a crawl as to try to memorize the shape of the word.

I have not found study linking the reading speed and the speed of learning, I would theorize that one has no effect on the other.

But recent technologies like auto-correction and word completion have partially compensated these mistakes and recent trend of journalism to "paraphrasing" news cable somewhat amplified its effect.

Thanks so much for your comment! I learned a thing! I (apparently) read that way (global reading method), I don't know any other way to read, and I wasn't sure it was anything unusual. My parents didn't do anything special, it's just how I read.

I also have atrocious spelling, always have, and can't read things with exotic words. (when reading sci-fi I skip all proper nouns).

This also explains why "bionic" is doing nothing for me -- it's meant to nudge one to where I already am.

I did some research and Wikipedia pointed to the whole language method (https://en.m.wikipedia.org/wiki/Whole_language) as opposed to the phonics method (https://en.m.wikipedia.org/wiki/Phonics) which focus more on how words are composed from syllables.

As a non expert, my analysis would be reduced to :

- whole word would focus of growing the list of words you would be able to recognize. - phonics focus on the link between speaking and reading, which I guess could build up self confidence for being more eloquent, and in my opinion would help with learning foreign languages.

This definitely does not have the desired effect for me. The uneven contrast made me read slower since I found myself re-scanning the same words again to confirm their spelling. It's only slightly less distracting than reading rANdoMlY CaPitAliSeD tEXt.

Me too, I find it absolutely terrible. I read it at about 10% of the speed I usually read. In the sample I saw, I could read the plain version in a few seconds, but I just couldn't make any headway with the highlighted version. I'm used to syntax highlighting, which doesn't slow me down at all (might make in fact make reading code faster) but this just throws me for a loop.

Algorithm of this version seems brutally simple.

const mid Math.floor(preElem.length / 2); > {preElem.slice(0, mid)}{preElem.slice(mid)}

I thought the original was more complex a nuanced than this.

I read through the sample and my eyes/brain were able to move over the lines incredibly quickly and recognize the words. I notice it only worked when my intent and goal was to move my eyes quickly over the lines. And, I noticed my eyes could do little fast forward skips as they passed over the words, which is not the norm for me. I don't think this technology would do anything for my retention though, I only skimmed the words. Maybe with practice I could read faster and retain the information as well at speed.

I've tried it a few times. Does not work for me.

I do not perceive any speed difference, there could be one but it would be marginal.

I am a fairly quick reader, I trained to read fast and extract content in high school.

I do - it makes me read at about 1/10th the speed I normally do. Hate it.

It is obvious that the improvement to this, is to highlight just the key words of a paragraph and not all words. Requires a way to figure out which words are the important words. Use some AI tool that understands a language.

It's too much clutter to highlight all words. No need for the connective words, etc.

Many have been doing this manually for a long time.

This is patent encumbered, so if you use it for a product, you’re setting yourself up for trouble down the line.

The project page should be updated to mention the patent situation IMO. The license is MIT, but that doesn’t mean you can use this without paying the patent holder whatever has asks.

This reminds me a little bit of the free Spreeder app (https://www.spreeder.com/app.php). I used it quite a bit at one point but forgot about it and that was that.

Im not totally sold that this is useful, but Im curious. I'll see if I can convert an epub and spend some time reading with their weird font thing on my remarkable. For science!

I trace with my finger while reading. It's a super simple way to improve focus and speed.

how does someone acqiure a patent for stylistic typography that reaches across typefaces?

this is really cool. I wonder if it's possible to create a bionic font such that you could get system wide support, well for the most part.

Patents for things like this make me feel like committing civil disobedience. Bolding the first few letters is something that needs patent protection? Give me a break. That's not a thing, man.

What if we underline the first character(s)? Can we get specs from those who grant software patents? What is included, what is not?

_Patented_

no reason for this to have a patent at all

the author could've trademarked the name and sell it as "the original"

personally the only app i know that has the technology is Reeder RSS reader, although i've never found it convenient

Patents shouldn't be a thing full stop.

Why spend my forcefully taken tax money protecting imaginary rights of a few inventors?

I understand the problem of industrial spies and creating something whose recipe is hard to find and easy to make, but I don't see why we should solve it societally for the large corporations which happen to have this issue.

Let them spend money suing their customers who violated some ToS and reverse engineered the formula, let them worry about vetting or doing suing their own employees. And if it's too expensive to pursue, consider it a natural tax of the market.

If they're creating something that will give value over a long period of time, they're already making more than the rest of the population who's paying for their patent bs and their copyright enforcement.

I agree that the current patent system is broken, but there are a lot of situations where I think protecting an invention for some time is warranted.

Let's say you invest 10 years of your life trying to develop a new cure or drug, but some other company reverse engineers the formula and sells it for cheaper, preventing you from even recoup your investment costs.

This might not be a problem for large corporations, but nobody else would be incentivized to try to create or develop something cutting edge, which might have a high risk of failure or take a long time to see through.

I think once you've actually sussed out the expenses of patents and copyrights they outweigh the benefits. The system as it stands is a product of the system as it was designed. It's real cute and nice at the outset, but lending monopolies to individuals (corporations or otherwise) means concentration of wealth, means outsized political power, means longer more inequitous parents which ultimately gives rise to inefficient rentiers. This defeats the competitive elements necessary for a free market.

Like have you read about the pharma companies that pay out generic mfgs to not manufacture patent-lapsed medication? This allows the patent-holder to buoy their prices and continue the monopolistic racket for medicines without competition. And pharma is held to a different standard.

If we take your example, we'll assume the competition's process is more efficient, well they can't patent that so the inventor can fund and build a factory on par with the competition. Consumer benefits. As it stands one party could design the drug, the other the process, and they could compete inefficiently with royalty bargaining or some such, ultimately leaving little benefit to the consumer. More realistically a conglomerate buys out everything related and sets the price for maximum ROI - supported by the system which has been driven to the inevitable conclusion of artificial monopolies leveraging their outsized wealth in the political arena.

Mentioning the propensity for larger established companies to buy out nascent companies and strangle their products which is enabled by patent law and has lost the collective a considerable amount of goods.

Except they're never used that way in practise. The small startup or inventor doesn't have enough resources to litigate if someone violates their patent so it only serves as a tax on inventors in exchange for not beingnsued by a patent troll for using their own ideas.

It’s not about protecting their rights, it’s about encouraging corporations to release the processes of their invention. The point is that we provide legally enforced protection in exchange for it entering the public domain after a few years. The alternative is corporate secrets that never become public (unless stolen), like the formula for Coca Cola.

Patents generally don't these days either, the patents are so vague and general that you can't reproduce the device using the patent or even know which of the 50 off-handedly mentioned methods is used to implement a general idea like 'synchronizing some clocks to measure a thing but using radio waves or the internet'

A sane system would require registration andnsource and only last a handful of years.

Try these bookmarklets:

With a peach background color and a pivot character one-third of the way through the word (plus one):

javascript:void function(){javascript:(function(){var a=Math.floor,b=document.querySelectorAll("p, title"),c=[],e="",f="",g="",h=0,k=0,l="",m="",n=window.open("","_blank");for(var d in b){var i=b[d].textContent;i%26%26(c=c+"\n"+i)}for(f=c,e=f.replace(/\n/g," "),g=e.split(" "),h=0;h<g.length;h++)k=a(g[h].length/3)+1,l=""+g[h].substring(0,k)+""+g[h].substring(k,g[h].length)+" ","."==g[h].substring(g[h].length-1,g[h].length)%26%26(l+=" * "),m+=l;n.document.write("<html>"+m+"</html>")})()}();

With a peach background color and no change to bolding for comparison:

javascript:void function(){javascript:(function(){var a=Math.floor,b=document.querySelectorAll("p, title"),c=[],e="",f="",g="",h=0,k=0,l="",m="",n=window.open("","_blank");for(var d in b){var i=b[d].textContent;i%26%26(c=c+"\n"+i)}for(f=c,e=f.replace(/\n/g," "),g=e.split(" "),h=0;h<g.length;h++)k=a(g[h].length/3)+1,l=""+g[h].substring(0,k)+""+g[h].substring(k,g[h].length)+" ","."==g[h].substring(g[h].length-1,g[h].length)%26%26(l+=" * "),m+=l;n.document.write("<html>"+m+"</html>")})()}();

It's always fun for me to see flashes in the pan come and go.

I saw this first on reddit a few days ago and went "hmm, ok, whatever..."

And then on HN today.

Now I'm saying "Oh, someone's putting real effort into this thing. Shame, really, I wish they'd move on to something more productive or at the least talk to some experts about their goals. Sigh."

And it'll be gone in a week, or a month at most. Sorry.

"Convert Text into Better Way to Read Faster"

But, based on what? I don't see any pointers to papers or other academic work or collaboration.

It makes it harder to read imo. I’d rather the beginning and ending of each word be bold and the middle faded.

Not seeing the end of the words clearly makes my brain only read the bold part.

Since we only need the first and last of a word for our brain to figure out, why not utilize a style that reflects this knowledge?

I agree with you - I only see the bold part. In some cases that makes it feel faster- but if you don't “guess” the word correctly from just the bold part, then my reading stutters.

From comments on HN, Reddit and Twitter, this approach clearly does help many people read faster and with less visual strain - just not everyone. Others say they find it harder to read. I looked at a couple of the iPhone apps that integrate the original Bionic Reading tech and they make it an optional feature that can be toggled on by the user based on preference.

It would be very interesting to research what makes it more effective for some people than others. It implies some difference in how their brains or eyes are processing text.

Well, no. It clearly provides self-reported testament that it helps.

But if you believe that, then 9/10 women found shampoo/moisturiser to improve their shine/visibly fade wrinkles!

/that is to say, self-reported testaments and improvements have historically been the chosen method of snake-oil because placebo and self-dellusion are so powerful. There does not appear to be robust evidence that it works at all yet.

The lack of existing research to explain (or contradict) people’s reported experience is what makes projects like this so interesting, and worth devoting resources to researching further.

> But, based on what? I don't see any pointers to papers or other academic work or collaboration.

Do you only experience things once they have been thoroughly peer reviewed?

Why not quickly try it out for yourself and decide.

> Do you only experience things once they have been thoroughly peer reviewed?

What a naive and leading question.

I judge strong assertions on the evidence provided. This is what taking "Caveat emptor" to heart looks like.

> Why not quickly try it out for yourself and decide.

I did. It doesn't make any difference for me.

That's the same argument that I've heard from religious members of my family. "Just read this book and apply what it says, your life will definitely improve."

Sure.

This is so simple that anybody can have a go at implementing it in a few minutes. Definitely more productive than a dismissive comment here.

Somehow this has been granted a patent... the way I see it, this needs to have 800 free/open implementations just to prove a point to the patent office that somehow let this pass as "innovative" (_whatever_, like you say).

Aside from various js snippets posted here, a firefox and chrome plugin I also found an emacs mode https://github.com/melpa/melpa/pull/7990

vim _definitely_ needs one too now please

Reply on news.ycombinator.com