29 March 2022

Kagi

by jensen

I used to love Google. When I was in high school, to google was already in use as a transitive verb though not yet officially recognized as it would be a few years later, they had already sent AltaVista screaming into the arms of Yahoo, and having a Gmail account - which were invite only at the time - was a badge of honor among the computer kids. Their code of conduct and motto was “Don’t be evil”. It resonated with me and many other young computer geeks. As they grew, there were signs of problems,but there was still a sense Google was on humanity’s side. They were so charming.

My friends and I read articles about how Google was Hoovering-up talent, the ad money was feeding all kinds of interesting side projects, and the work world was abuzz about “20% time” and Google’s many office amenities. It seemed like maybe here was a company that might do something good for the world. Some new money to shake up the old. We were naive. Coasting on the optimistic fumes of the early to mid 90s internet. As time went on, many of us saw the ropes and nets of capitalism closing in around us; the RIAA and MPAA had gone from minor nuisance to actual threat, Amazon was starting to look like the titan we know now, and Microsoft was being Microsoft. Despite all that, I and those around me believed the invaders couldn’t stop the free information utopia the internet was surely destined to bring. What I failed to realize was that evil corporations were not the only threat to the internet. I had not yet experienced enough of life to fully comprehend that people ruin everything. The western world’s gen pop fell upon the internet as a swarm of locusts. We mocked AOL, but AOL was a dam for nincompoops. Did I expect the people grazing in the pastures of America Online once exposed to the internet I knew would suddenly join a Linux IRC channel and stay up all night talking shell scripts and cypherpunks? I don’t think so, but I thought at the very least many of the ideals I picked up on the early internet would change them a little and not so much the other way round. Google, even as it was obviously more corporate than ever, still felt born in those ideals. I guess new or old, money begets a desire for more money.

By 2015 Google had restructured under a conglomerate, Alphabet Inc. Alphabet’s motto is the more malleable “Do the right thing.” In 2018, “Don’t be evil” disappeared from Google’s code of conduct preface, the only reminder left impotently in the last line “And remember…don’t be evil, and if you see something that you think isn’t right – speak up!” It didn’t matter, though. By then it was only ever mentioned with heavy italics, layered in sarcasm or irony. Google’s reputation had become that of an ever encroaching data glutton and peeping Tom. I myself had felt a bit sick about just how many of my eggs were in Google’s basket. Email, search, youtube, blogger, Keep, my Android phone. I was in too deep. I started to de-Google my life.

I moved to a different email provider. I use Jekyll for this blog, and I started using DuckDuckGo. I liked it. It was “good enough” and when it wasn’t the bangs made it easy to search Google or a slew of other sources. It got me thinking about privacy as a product feature. I realized if I don’t want to be a product, it’s going to take some effort and probably some money. My email had already been moved to a paid provider. This blog is using a github pages setup for hosting, but that doesn’t bug me. I can easily move this wherever I want if that relationship sours. Jekyll is lovely that way. I’ve replaced Google Keep with Obsidian kept in sync across devices via Syncthing. At this point, I’m feeling pretty good. I just wished I wasn’t using DDG bangs to go to Google so often. Something about their native search wasn’t doing it for me, I was really just using DDG as a proxy. I found myself wishing I could just pay Google for search and maps with no ads and no tracking. I saw someone express a similar desire in a discussion thread around privacy, and they were recommended Kagi.

Kagi is a search engine currently free for beta, but will eventually launch as a paid service. It has bangs like DDG, but I haven’t used them much. I’ve found its search results not only “fine” but quite good. For my needs (lots of powershell, active directory, sysadmin-ish stuff for work and video games, books, and articles personally) I find the results as good or better than Google’s. I like their “lenses” feature which allows you to customize and refine search results. It made finding the above Wired article rather simple - I knew I’d read something about Google, moral struggle, before 2004, and on or in Wired. I saved the “lens” and now I have an easy filter for Wired articles I can modify and use to refine my results whenever I need. Scoff if you like, but I find myself searching for articles I forgot to write down frequently. I was surprised to learn that Kagi search functions through a series of API calls to other sources (including Google). It then uses its own logic to select and sort the best results. It seems a little like cheating, but I’d argue it’s a clever way to provide search without the expensive task of indexing the web. Kagi does struggle with local results, it’s hard to beat Google’s immediate results that include directions, phone number, website, reviews. But, there’s no reason Kagi couldn’t get there, and what they have so far is something I would in fact pay for. Their search is good, and they are upfront that the price for those search results is money and not me.

tags: search