Ask HN: What will the web look like in 5 years with AI bots becoming widespread? | Hacker News Reader

Ask HN: What will the web look like in 5 years with AI bots becoming widespread?

2 pointsbwb12 comments

Hi all,

What will the web look like if these LLM bots become users' primary interface (rather than traditional websites)?

Do the majority of websites disappear?

Will monetized MCP servers become a new version of websites?

Are mobile apps even more critical now, as the app store is the last place for discovery?

Curious to hear your predictions, worries, or hopeful takes!

I run a book discovery website (https://shepherd.com) where people share books they love, and we try to unearth great books people haven't heard about. We've seen our traffic from Google drop massively as Google replaces search results with its AI chatbot. We get direct traffic from our fans, but without Google introducing us to new people, I'm worried there won't be any discovery. As a result... I'm working to launch a mobile app with a full reader app.

Any advice for me?

View on HN

The underlying commercialization of the Web is too lucrative to be gone in 5 years, it's just a question of what watering holes the non-dead congregate at, and how the predators target them.

At this point, I'd expect a further move towards walled gardens to reduce competition (harder to scrape, copy, etc.) and increase the barrier for entry to new players: it's going to be even more pay to play than it is now.

I don't expect the Web to die, websites will stay a thing, but how they are discovered and used will undoubtedly shift. E.g., relying on tricks to get search traffic is a waste of time if nobody uses search.

Hopeful take: other protocols take more prominence. Disruptions and a fracture or two might help the Internet stay useful.

Well said, I also hope this is a good thing that wipes a lot of AI slop off the web. But I also see friends destroyed by it and many websites shut down.

Any "out there" ideas on where discovery shifts?

i.e., I built something cool, I do a Show HN, but how do other people find it in the future?

Seems like search will disappear completely. I used ChatGPT to find a coffee table that was exactly 58-inches long (and iirc with other characteristics), google search and amazon/wayfair internal search had completely failed at this task. I wonder can you replicate your site with AI -- ie. prompt like "I'm a person like this A,B,C looking for a book like this X,Y,Z"...

Yeah, I like it for that.

Do you want to be recommended a book by AI?

We use AI internally as a tool, and I am working on some things around "book twins" to help people find books they will love for sure. I am going to look at possibility of a prompt type search, right now we do filters.

It will become much more walled garden, I think we will see a rejection of the algorithms and we will get content more from communities than ad's tracking you and serving you up content. I think the web is going to look alot like it was before Facebook with Forums etc but these will be slack channels/discord groups.

The rest of the web will just be an uncharted landscape, I'm looking forward to Stumbleupon making a comeback.

I miss StumbleUpon :). Did you see that Digg is also about to make a comeback? I joined their group and it's been fun to see how they are thinking through it -> https://reboot.digg.com/

I am disgusted with book categories on the website. No math books, programming books, philosophy books, etc but there are so many categories of not worth to read. I did not even expect the year is important metrique for somebody.

Eh? Do you mean my website?

Those are all over the website:

Math as a discovery point -> https://shepherd.com/search/shelf/693

Math as a bookshelf -> https://shepherd.com/bookshelf/math?order=most_recommended

Tons of great interviews with authors around math as well:

30 year professor of mathematics -> https://shepherd.com/best-books/books-that-make-maths-intere...

Another that goes deep into mathematical concepts -> https://shepherd.com/best-books/for-mathematical-inspiration

Another examples of math and science and where they mesh -> https://shepherd.com/best-books/narrative-merit-in-mathemati...

And same goes for programming and philosophy, here are those discovery points: https://shepherd.com/search/shelf/876 https://shepherd.com/search/shelf/2568

Plus specific concepts like: - machine learning - https://shepherd.com/search/shelf/2055 - AI (mix nonfiction / fiction) - https://shepherd.com/search/shelf/77 - Agile - https://shepherd.com/search/shelf/5117 - Plato - https://shepherd.com/search/shelf/1087

Hope that helps, where did you go that it was confusing? Let me know and I'll fix it :)

That math books are mostly not a real math books. When I want to have a math book I go to torrents and download hundreds of them. They do not have so rich annotation as books on your website, but too rich annotations means you have only 10 or less books per page. The biggest flaw of your math book collection is that it was collected by not a mathematician. When I am looking a math book I typically looking for a book with some tasks after each topic.

Same about philosophy - those are not a philosophy books. All attemts to mark some math or philosophy books as "best" do not work here, because I do not interested about what most people appreciated. I am interested in all books per field. For example, I am interested in Stoicism, but you have only "stoicism from a psychiatrist and philosopher" (what was the stupid reason to prohibit me from selecting and copying this line) - but I hate psychiatrists as scientific freaks.

Another rebuke is that your webpage takes too much memory, I would like to see not so rich and colorful interface which should take less than 50MB per page.

Dude, don't steal books :(, that hurts authors.

Yep, we are probably not made for you. I get it. Our UX is designed for a different group of readers :). I hope you can find a great place to find books other than us.

Will authors be pleased if I stop reading their books? How about dead ones? The most upsetting part of your website for me is focusing on new (recently published) books, while I use to appreciate older books more.

We don’t focus on newly published books, why did you think that?

We focus on the most loved books published in any year with everything we built. We have filters for the readers who do want to see specific years in some sections, but that is it. Books are fairly timeless.

I think authors really appreciate it when you pay them for said book. The value is immense compared to the small cost, and otherwise there will not be more authors. Libraries are great too! I know if you are outside an English speaking country, books in English can be expensive, especially technical books, so i do understand. But hopefully one day you can pay that back.

Reply on news.ycombinator.com