An attempt at articulating thoughts and annoyances of the state of WWW. Decentralised Open Indexes for Discovery (DOID): A conceptual and technical framework for resource discovery on the WWW using decentralized, open, machine-readable indexes as the building block, free of eroding quality and gatekeeping by BigSearch™ and BigPlatform™, whose goals are not quality, but revenue. https://lnkd.in/gz-HWA59
Quite an interesting and relatable take. As a researcher and writer, I've myself been using LLMs first then using search engines to query very specific phrases that can help me find resources that a short keyword wouldn't return. But with the solution proposed in your piece, or with any solution that aims to scale to a large audience, I think, monetization/revenue is a critical aspect (exceptions exist). Wikipedia for example is not monetized, but it relies on donations. I doubt we can build a decentralized internet just using that model. So, the biggest challenge with DOID would be to scale to the masses and/or compete with platforms that keep choosing revenue over quality.
Anubhav agarwal
2mo
Hii iam ANUBHAV AGARWAL 17year old want to build a bio cng plant but first i need to build cattlefeed buisness which is in kota please reply i just want meet you for regaerding cattle feed start up which is based on napier grass so lets find out what is napier grass it is a grass which can be cut multiple time in year it have regular yield about 200 ton per acre but some farmer acheive 400 to 1000 ton per acre so the conclusion is bcoz of high yeild i want to provide affordable cattlefeed to dairyfarm please drop mail [email protected] or drop message on linkedin I need help from you and help is not INVESTMENT REGARDING
Bhagyesh Pathak
2mo
Wow! Kailash, like you said, for the past 1.5 years, LLM agents are my go-to question-answer tools. WWW is almost dead except the BigSearch and BigPlatform decides to reverse, but that's not how systems work. Systems repair themselves by collapsing, so we are going to stare at this cesspool for a bit longer. I wish there were more people with a spirit such as yours. Who say "here's the poblem, this is what I'm going to try to address it. Wanna join?" Thanks for sharing the idea.
Nabheet Madan
2mo
Wow. You have rightly pointed out, we have so much noise that it gets tough to discover the quality content. Somehow even the places where we use to find quality content like stackoverflow etc. are also being facing this challenge where quality is goong down due to noise. I think for me right now it LLM accompanied by documentation of that product remains the goto resource to filter out the noise. Quality content discovery is a big TASK.
Noorul Ali
2mo
https://alinoorul.github.io/matrixnet has Botnet, my planned browser to get JSON content directly from user computers (https://alinoorul.github.io/personaldatabank) and render it in a text-only website with https://www.github.com/matrixdex/devpy
Dhaval Panchal
2mo
Very thoughtful idea. This is an alternate approach then the crawling algorithms(which use url/refer as a starting point) for search engines. DOID can be clubbed to create blockchain supported platform which answers more like prompts then search engines.
Sumit Kishore
2mo
Kailash Nadh In my opinion, search will transition from “query-based” to “prompt-based” actions. Creating a user-friendly UI for DOID could be a great step forward. Good luck, and keep us updated!
Anil Chaudhary
2mo
Hmm, Can't this be solved using people's own way of bookmarking specific content and then have a RAG based search on it ? Every user will have what they like and value with full memory
Jeeban Bhagat
2mo
The idea of DOID really stands out. I’m curious how you see this balancing usability with decentralization—it’s a shift that could transform how we access information.
Software Engineer
2moI am amazed at how 'good' (?) the NotebookLM podcast conversation sounds, but also found it sad in some ways. Also, while I was reading through, RSS feeds came to mind, it is XML (not ideal), there are 'awesome lists' for RSS feeds, specially engineering related, maybe thats a good start for people wanting to have access to curated content. The only thing missing from this is though a good search interface.