Hacking Search: Kagi's Revolt Against the Ad-Tech Machine

Vladimir Prelovac, founder of the paid search engine Kagi, begins with his path into computing and his doubts about ad-funded search. He cites Google's own early warnings about advertising conflicts and a 2018 query that showed a sponsored and an organic link to the same site. The core argument is that when users pay directly, incentives shift toward their interests and privacy follows as a byproduct.

The middle covers user-facing features such as blocking or promoting domains, lenses that search subsets of the web, custom URL rewrite rules, and on-demand AI answers. Behind the scenes, a crawler measures how many ads and trackers a page carries as a rough signal of low quality, while results blend several external indexes with Kagi's own and then apply personalization. A small web initiative boosts personal blogs, and a privacy pass built on cryptographic tokens lets users query without the service linking them to an account.

The closing questions touched on international coverage, competition, the cost of paying for upstream indexes, accessibility, and planned profit sharing with publishers whose content feeds AI answers. The main takeaway is that a small team can deliver useful search features without owning a full web index, and that the business model shapes privacy and product choices.