{"id":9862,"date":"2026-11-14T10:38:14","date_gmt":"2026-11-14T17:38:14","guid":{"rendered":"https:\/\/shermandorn.com\/?p=9862"},"modified":"2026-11-14T10:39:12","modified_gmt":"2026-11-14T17:39:12","slug":"zotero-beats-llms-for-me-at-least-and-why","status":"publish","type":"post","link":"https:\/\/shermandorn.com\/?p=9862","title":{"rendered":"Why Zotero beats LLMs, for me"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">I&#8217;ve avoided writing much about generative AI because the underlying technology has been shifting around too rapidly in the past 4 years for an historian to do much other than wave hands at the history of (education) technology and suggest people apply some perspective. But I have a real thought now!<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">I&#8217;ve come to rely on a critical academic infrastructure tool for my research in the past ten years. It&#8217;s transformed my productivity as a scholar, and while I just noodled on it during its early development, I suspected I&#8217;d use it heavily at some point. That&#8217;s Zotero, not ChatGPT or Claude or Muse or Gemini or Flibbijibbit 3.2 or whatever is the latest flavor of &#8220;it&#8221; LLMs. I think the reason is important to understanding technology as part of culture. (And yes, I agree with Farrell et al. that <a href=\"https:\/\/doi.org\/10.1126\/science.adt9819\">LLMs are cultural and social technologies<\/a>.)<\/p>\n\n\n\n<!--more-->\n\n\n\n<p class=\"wp-block-paragraph\">I know the team that originally developed <a href=\"https:\/\/www.zotero.org\/\">Zotero<\/a>, the ubiquitous, free and open-source citation package from the <a href=\"https:\/\/rrchnm.org\/\">Roy Rozensweig Center for History and New Media<\/a>. Lovely people, all of them, and I remember the original <a href=\"https:\/\/web.archive.org\/web\/20070104202651\/http:\/\/www.zotero.org\/blog\/zotero-10-public-beta-launch\/\">public beta launch<\/a> in October 2006. Like many of the Center&#8217;s projects, it was clever, the project team thought carefully about building a community so it didn&#8217;t die on the vine, and it addressed a real need: an open-source citation management system designed with the needs of humanities scholars in mind. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If you&#8217;ve never used a citation manager, Zotero is like a Kindle app except much, much better, and for anything you can find in an electronic form. Or even a non-electronic form, as I&#8217;ll explain below. And what you have is always, permanently available. And you can take notes on it, and tag any reading or set of readings, put them in folders, share them with friends and colleagues, and then &#8212; and this is the original use case &#8212; if you need to create a bibliography or add footnotes to a document, it&#8217;s just a matter of a few clicks to do that. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">While Zotero took years to overtake the commercial citation managers then ubiquitous in science disciplines &#8212; the incumbents had a huge advantage in the lock-ins they had built with both users and academic libraries &#8212; it was obvious how useful it was, especially as the team worked hard to expand its functionality. The team also figured out how to sustain its development: they charged reasonable rates to hold a synchronized copy of your personal library on a server owned by the Zotero team (now the property of the non-profit <a href=\"https:\/\/digitalscholar.org\/\">Digital Scholar<\/a>). <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">As noted above, I used Zotero casually for the first 12 years of its existence. The citation management wasn&#8217;t that important to me, since I had become accustomed to adding all my citations while writing, occasionally copying them from older manuscripts. Instead, it was valuable to me in accumulating articles and reports to read, especially as Zotero added a browser bookmark where I could find a document online or through my library&#8217;s electronic holdings and click once to download the document that would be attached to the basic entry (with the relevant citation information and other metadata). But I&#8217;d been saving PDFs to laptop hard drives for years, so it wasn&#8217;t as if I absolutely needed Zotero for my work. It was convenient for the types of documents it was convenient for (circular reasoning intended). <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">I began to use Zotero heavily in 2018 when I started my current major project on the history of educational broadcasting policy in the U.S. Eight years ago, Zotero was one of the two major citation managers the ASU Library supported, and I let Wooyeong Kim choose which one we&#8217;d use for the project. Now at the Academy of Korean Studies, Wooyeong was a first-semester PhD student I advised, he was my research assistant, and I wanted him to develop good habits and do so with a tool he chose. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">It was a good choice! We spent the first few months adding secondary sources to a shared library, and then at our first archival visit, we quickly developed a workflow: when looking at the primary source, add the entry with all the metadata. For anything that was remotely important, scan the document with a phone app (Scanner Pro for me), upload the resulting PDF to a temporary folder (a Dropbox folder for me), and then import the PDF file into Zotero. ASU is part of the &#8220;eduroam&#8221; network of scholarly institutions that allow you to join another network member&#8217;s wifi, and I have almost always been able to connect to an eduroam wifi network in an archive&#8230; and thus grab the files as they transferred to my Dropbox account. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Today, after exhausting the small grant the Spencer Foundation awarded me for archival travel, that shared library now holds several thousand entries, most of which are from archives and most of which have the document scans attached. A few hours of teaching a grad student to use Zotero, and then a little more time figuring out how to use it for archival research, and I could conduct an archival project at a scale I absolutely could not have for any of my other projects. Zotero absolutely transformed my scholarly work. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">I have used LLMs in a few ways at work thus far &#8212; four in total, I think, or one a year:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Demonstrating for doctoral students how one could use an LLM to help generate synonyms for first-choice literature search terms. This was part of a discussion of why one needs synonyms for search terms, and side-by-side with two other ways to generate synonyms sets (looking at landmark studies, and asking fellow scholars and also practitioners for relevant terms). <\/li>\n\n\n\n<li>(Briefly) testing ChatGPT&#8217;s capacity in 2024 to match publication titles, abstracts, and full texts to a set of inclusion\/exclusion criteria for a lit review. This was planned as a side project for a lit review I was helping some students with, used it to request access to ASU&#8217;s then-pilot contract with OpenAI, and then shelved after only a few hours of work &#8212; this was the summer my wife died, and every one of my low-priority projects died that year while some fairly important ones were on hiatus.<\/li>\n\n\n\n<li>Using LLMs to check my understanding of some R code a colleague had written, and to draft some R code for specific purposes where I could not find an existing package.  <\/li>\n\n\n\n<li>A pilot text-tagging project, using a fairly small set of existing tags for some texts. <\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">I have also kicked the tires on a few LLM projects of colleagues, but that&#8217;s not active use in the sense of learning how to use a tool or platform. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">There are ethical reasons why my use of LLM has been limited &#8212; these ethical considerations include the disproportionate resource consumption of some types of LLM use, such as textual manipulation (even tagging\/coding with a prior set of categories) that require far more tokens than others, and the brazen theft of intellectual property that&#8217;s inherent in frontier model training. I see my travel to conferences and archives as equivalent in both areas: when I visited Portland this month, that travel contributes to global carbon emissions, and the extraction of crude oil from the ground involves damage to people as part of the extraction of fossil fuels and through environmental degradation involved in all mining and other extraction. So I am selective where and when I use an LLM, as I am selective about conference attendance. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">But my ethical concerns about LLM use are not why I contrast my growing use of Zotero with my relatively limited use of LLMs. Like an LLM, Zotero is a cultural and social technology. Like an LLM, Zotero has evangelists who are happy to tell you about the wide range of uses for it. Unlike an LLM, no one is claiming that Zotero is a general-purpose technology. But that&#8217;s what makes it more useful for me: I could match it to my specific need. No one is claiming that Zotero will cure cancer, solve anthropogenic climate change, replace white-collar workers, or destroy the world. It&#8217;s easy to be clear-eyed about the role of Zotero. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">It&#8217;s much harder to be clear-eyed about LLMs, because the claims seem just plausible enough that one cannot dismiss them out of hand. Because of the overhyped nature of public discourse about them, I think LLMs are likely to be useful in a selection of specific contexts, and a serious problem when sold to people who don&#8217;t need them at all. But that&#8217;s an informed suspicion, and I could be wrong. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">More importantly, when people talk about the uses of LLMs, look at the existing and growing use of other technologies at the same time. Which has had a larger impact on the world this decade, the application of mRNA technology in vaccines, or LLMs? The spread of inexpensive solar and grid-scale battery technology, or LLMs? The military adaptation and use of drones, or LLMs? <\/p>\n","protected":false},"excerpt":{"rendered":"<p>I&#8217;ve avoided writing much about generative AI because the underlying technology has been shifting around too rapidly in the past 4 years for an historian to do much other than wave hands at the history of (education) technology and suggest people apply some perspective. But I have a real thought now! I&#8217;ve come to rely &hellip; <a href=\"https:\/\/shermandorn.com\/?p=9862\" class=\"more-link\">Continue reading<span class=\"screen-reader-text\"> &#8220;Why Zotero beats LLMs, for me&#8221;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_hide_featured_image":false,"_jetpack_featured_image_duplicate_dismissed":0,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"jetpack_post_was_ever_published":false},"categories":[12,4,18],"tags":[],"class_list":["post-9862","post","type-post","status-publish","format-standard","hentry","category-higher-education","category-research","category-the-academic-life"],"jetpack_publicize_connections":[],"jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/pag0MB-2z4","jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/shermandorn.com\/index.php?rest_route=\/wp\/v2\/posts\/9862","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/shermandorn.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/shermandorn.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/shermandorn.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/shermandorn.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=9862"}],"version-history":[{"count":16,"href":"https:\/\/shermandorn.com\/index.php?rest_route=\/wp\/v2\/posts\/9862\/revisions"}],"predecessor-version":[{"id":9878,"href":"https:\/\/shermandorn.com\/index.php?rest_route=\/wp\/v2\/posts\/9862\/revisions\/9878"}],"wp:attachment":[{"href":"https:\/\/shermandorn.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=9862"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/shermandorn.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=9862"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/shermandorn.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=9862"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}