{"id":46177,"date":"2026-10-29T11:41:44","date_gmt":"2026-10-29T16:41:44","guid":{"rendered":"https:\/\/huewhite.com\/umb\/?p=46177"},"modified":"2026-10-29T11:41:44","modified_gmt":"2026-10-29T16:41:44","slug":"word-of-the-day-1155","status":"publish","type":"post","link":"https:\/\/huewhite.com\/umb\/2026\/10\/29\/word-of-the-day-1155\/","title":{"rendered":"Word Of The Day"},"content":{"rendered":"<p><em>Open weights<\/em>:<\/p>\n<blockquote>\n<p id=\"mwCA\"><b id=\"mwCQ\">Open weights<\/b>\u00a0are the publicly released learned\u00a0<a id=\"mwCg\" class=\"mw-redirect\" title=\"Parameter (machine learning)\" href=\"https:\/\/en.wikipedia.org\/wiki\/Parameter_(machine_learning)\" rel=\"mw:WikiLink\">parameters<\/a>\u00a0of a trained\u00a0<a id=\"mwCw\" title=\"Artificial intelligence\" href=\"https:\/\/en.wikipedia.org\/wiki\/Artificial_intelligence\" rel=\"mw:WikiLink\">artificial intelligence<\/a>\u00a0model, principally its weights and biases. In an\u00a0<a id=\"mwDA\" class=\"mw-redirect\" title=\"Artificial neural network\" href=\"https:\/\/en.wikipedia.org\/wiki\/Artificial_neural_network\" rel=\"mw:WikiLink\">artificial neural network<\/a>, weights are numerical values that determine how strongly inputs contribute to the network&#8217;s calculations. Publishing these values allows others to download and use the model, while permission to modify,\u00a0<a id=\"mwDQ\" title=\"Fine-tuning (deep learning)\" href=\"https:\/\/en.wikipedia.org\/wiki\/Fine-tuning_(deep_learning)\" rel=\"mw:WikiLink\">fine-tune<\/a>\u00a0or redistribute it depends on its\u00a0<a id=\"mwDg\" title=\"Software license\" href=\"https:\/\/en.wikipedia.org\/wiki\/Software_license\" rel=\"mw:WikiLink\">license<\/a>. The term is commonly applied to\u00a0<a id=\"mwDw\" title=\"Large language model\" href=\"https:\/\/en.wikipedia.org\/wiki\/Large_language_model\" rel=\"mw:WikiLink\">large language models<\/a>,\u00a0<a id=\"mwEA\" title=\"Diffusion model\" href=\"https:\/\/en.wikipedia.org\/wiki\/Diffusion_model\" rel=\"mw:WikiLink\">diffusion models<\/a>\u00a0and other\u00a0<a id=\"mwEQ\" class=\"mw-redirect\" title=\"Generative artificial intelligence\" href=\"https:\/\/en.wikipedia.org\/wiki\/Generative_artificial_intelligence\" rel=\"mw:WikiLink\">generative artificial intelligence<\/a>.<sup id=\"cite_ref-OSI-weights_1-0\" class=\"mw-ref reference\" data-mw=\"{&quot;name&quot;:&quot;ref&quot;,&quot;attrs&quot;:{&quot;name&quot;:&quot;OSI-weights&quot;},&quot;body&quot;:{&quot;id&quot;:&quot;mw-reference-text-cite_note-OSI-weights-1&quot;}}\"><a id=\"mwEg\" href=\"https:\/\/en.wikipedia.org\/wiki\/Open_weights#cite_note-OSI-weights-1\"><span id=\"mwEw\" class=\"mw-reflink-text\"><span id=\"mwFA\" class=\"cite-bracket\">[<\/span>1<span id=\"mwFQ\" class=\"cite-bracket\">]<\/span><\/span><\/a><\/sup><sup id=\"cite_ref-2\" class=\"mw-ref reference\" data-mw=\"{&quot;name&quot;:&quot;ref&quot;,&quot;attrs&quot;:{},&quot;body&quot;:{&quot;id&quot;:&quot;mw-reference-text-cite_note-2&quot;}}\"><a id=\"mwFg\" href=\"https:\/\/en.wikipedia.org\/wiki\/Open_weights#cite_note-2\"><span id=\"mwFw\" class=\"mw-reflink-text\"><span id=\"mwGA\" class=\"cite-bracket\">[<\/span>2<span id=\"mwGQ\" class=\"cite-bracket\">]<\/span><\/span><\/a><\/sup>\u00a0Open weights is contrasted to\u00a0<a id=\"mwGg\" title=\"Open-source artificial intelligence\" href=\"https:\/\/en.wikipedia.org\/wiki\/Open-source_artificial_intelligence\" rel=\"mw:WikiLink\">open-source artificial intelligence<\/a>, which considers the public release of not only parameters but the model&#8217;s\u00a0<a id=\"mwGw\" title=\"Source code\" href=\"https:\/\/en.wikipedia.org\/wiki\/Source_code\" rel=\"mw:WikiLink\">source code<\/a>,\u00a0<a id=\"mwHA\" class=\"mw-redirect\" title=\"Training data\" href=\"https:\/\/en.wikipedia.org\/wiki\/Training_data\" rel=\"mw:WikiLink\">training data<\/a>,\u00a0<a id=\"mwHQ\" title=\"Language model benchmark\" href=\"https:\/\/en.wikipedia.org\/wiki\/Language_model_benchmark\" rel=\"mw:WikiLink\">evaluation<\/a>\u00a0results, intermediate checkpoints and\u00a0<a id=\"mwHg\" title=\"Software documentation\" href=\"https:\/\/en.wikipedia.org\/wiki\/Software_documentation\" rel=\"mw:WikiLink\">technical documentation<\/a>.<em> [<a href=\"https:\/\/en.wikipedia.org\/wiki\/Open_weights\" target=\"_blank\" rel=\"noopener\"><strong>Wikipedia<\/strong><\/a>]<\/em><\/p>\n<\/blockquote>\n<p>The page carries a warning that some of the text may be AI-generated, which is prohibited in <em><strong>Wikipedia<\/strong><\/em> articles, and not a practice I condone. Noted in &#8220;<a href=\"https:\/\/www.washingtonpost.com\/technology\/2026\/10\/26\/anthropic-discloses-80-billion-loss-company-prepares-go-public\/?linkNumber=9&amp;utm_campaign=wp_post_most&amp;utm_medium=email&amp;utm_source=newsletter&amp;utm_content=&amp;utm_term=\" target=\"_blank\" rel=\"noopener\"><em>Anthropic reveals $80 billion loss, surging revenue as it prepares to go public<\/em><\/a>,&#8221; Gerrit De Vynck and Miriam Waldvogel, <em><strong>WaPo<\/strong><\/em>:<\/p>\n<blockquote><p>But not everyone is convinced. Some financial analysts have questioned the companies\u2019 abilities to ever become profitable, given the huge costs involved in training AI and the close competition from open-weights AI models, which are nearly as capable as the high-end AI from Anthropic and OpenAI but are much cheaper to operate.<\/p><\/blockquote>\n<p>The analog to open source.<\/p>\n<p>Or is it more appropriate to suggest the <em>digital<\/em> to open source?<\/p>\n<p>Never mind me, I&#8217;m a bit short of sleep. Nudge nudge, wink wink.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Open weights: Open weights\u00a0are the publicly released learned\u00a0parameters\u00a0of a trained\u00a0artificial intelligence\u00a0model, principally its weights and biases. In an\u00a0artificial neural network, weights are numerical values that determine how strongly inputs contribute to the network&#8217;s calculations. Publishing these values allows others to download and use the model, while permission to modify,\u00a0fine-tune\u00a0or redistribute \u2026 <a class=\"continue-reading-link\" href=\"https:\/\/huewhite.com\/umb\/2026\/10\/29\/word-of-the-day-1155\/\"> Continue reading <span class=\"meta-nav\">&rarr; <\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"nf_dc_page":"","_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_post_was_ever_published":false},"categories":[1],"tags":[],"class_list":["post-46177","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/posts\/46177","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/comments?post=46177"}],"version-history":[{"count":1,"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/posts\/46177\/revisions"}],"predecessor-version":[{"id":46178,"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/posts\/46177\/revisions\/46178"}],"wp:attachment":[{"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/media?parent=46177"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/categories?post=46177"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/huewhite.com\/umb\/wp-json\/wp\/v2\/tags?post=46177"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}