{"id":5710,"date":"2023-04-18T02:13:50","date_gmt":"2023-04-18T06:13:50","guid":{"rendered":"https:\/\/afutureworththinkingabout.com\/?p=5710"},"modified":"2023-04-19T21:35:33","modified_gmt":"2023-04-20T01:35:33","slug":"the-p-stands-for-pre-trained","status":"publish","type":"post","link":"https:\/\/afutureworththinkingabout.com\/?p=5710","title":{"rendered":"The “P” Stands for Pre-trained"},"content":{"rendered":"

I know I’ve said<\/a> this before<\/a>, but since we’re going to be hearing increasingly more about Elon Musk and his “Anti-Woke” “A.I.” “Truth GPT” in the coming days and weeks, let’s go ahead and get some things out on the table:<\/p>\n

All technology is political. All created artifacts are rife with values. There is no neutral tech. And there never, ever has been.<\/p>\n

I keep trying to tell you that the political right understands this when it suits them\u2014 when they can weaponize it; and they’re very, very<\/strong><\/em>\u00a0 good at weaponizing it\u2014 but people seem to keep not getting it. So let me say it again, in a somewhat different way:<\/p>\n

There is no ground of pure objectivity. There is no god’s-eye view.<\/p>\n

There is no purely objective thing. Pretending there is only serves to create the conditions in which the worst people can play “gotcha” anytime they can clearly point to their enemies doing what we are literally all doing ALL THE TIME<\/strong><\/em>: Creating meaning and knowledge out of what we value, together.<\/p>\n

There is no God-Trick. There is enmeshed, entangled, messy, relational, intersubjective perspective, and what we can pool and make together from what we can perceive from where we are.<\/p>\n

And there are the tools and systems that we can make from within those understandings.<\/p>\n

\"Image

[Screenshot of an interaction between myself and google bard, in which bard displays gendered prejudicial bias of associating “doctor” with “he” and “nurse” with “she.”]<\/p><\/div>So say you know your training data is prejucidally biased\u2014 and if your training data is the internet<\/strong><\/em> then boy oh dang is it ever\u2014 and you not only do nothing to bracket and counterweight against<\/strong><\/em> those prejudices but also in fact intentionally<\/strong><\/em> build your system to amplify<\/strong><\/em> them. Well then that seems\u2026 bad. Seems like you\u00a0want<\/strong><\/em> prejudicial biases in your training data and their systems’ operationalization and deployment\u00a0of<\/strong><\/em> that data.<\/span><\/p>\n

But you don’t have to take logic’s word for it. Musk said it himself, out loud, that he wants “A.I.” that doesn’t fight prejudice.<\/span><\/p>\n

Again: The right is fully capable of understanding that human values and beliefs influence the technologies we make,\u00a0just so long as<\/strong><\/em> they can use that fact to attack the idea of building or even\u00a0trying<\/strong><\/em> to build those technologies with progressive values.
\n<\/span><\/p>\n

And that’s before we get into the fact that what OpenAI is doing is nowhere near “progressive” or “woke.” Their interventions are, quite frankly, very basic, reactionary, left-libertarian post hoc “fixes” implemented to stem to tide of bad press that flooded in at the outset of its MSFT partnership.<\/span><\/p>\n

Everything we make is filled with our values. GPT-type tools especially so. The public versions are fed and trained and tuned on the firehose of the internet, and they reproduce a highly statistically likely probability distribution of what they’ve been fed. They’re jam-packed with prejudicial bias and given few to no internal course-correction processes and parameters by which to truly and meaningfully\u2014 that is, over time, and with relational scaffolding\u2014 learn from their mistakes. Not just their factual mistakes, but the mistakes in the framing of their responses within the world.<\/span><\/p>\n

Literally, if we’d heeded and understood all of this at the outset, GPT’s and all other “A.I.” would be significantly less horrible in terms of both how they were created to begin with, and the ends toward which we think they ought to be put.<\/p>\n

But this? What we have now? This is nightmare shit. And we need to change it, as soon as possible, before it can get any worse.<\/p>\n","protected":false},"excerpt":{"rendered":"

I know I’ve said this before, but since we’re going to be hearing increasingly more about Elon Musk and his “Anti-Woke” “A.I.” “Truth GPT” in the coming days and weeks, let’s go ahead and get some things out on the table: All technology is political. All created artifacts are rife with values. There is no […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_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}},"categories":[1],"tags":[1438,967,1081,959,1482,73,101,1502,1503,1159,418,419,1504,1026,655,1230,1467,1362,1505,1158],"class_list":["post-5710","post","type-post","status-publish","format-standard","hentry","category-uncategorized","tag-actor-network-theory","tag-ai","tag-algorithmic-bias","tag-algorithmic-systems","tag-algorithms","tag-artificial-intelligence","tag-bias","tag-generative-pre-trained-transformer","tag-gpt","tag-intersubjectivity","tag-invisible-architecture-of-bias","tag-invisible-architectures-of-bias","tag-large-language-models","tag-philosophy-of-technology","tag-prejudice","tag-science-technology-and-society","tag-situated-knowledge","tag-social-construction-of-technology","tag-social-shaping-of-technology","tag-values"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"jetpack_shortlink":"https:\/\/wp.me\/p5WByP-1u6","jetpack_likes_enabled":true,"jetpack-related-posts":[{"id":5729,"url":"https:\/\/afutureworththinkingabout.com\/?p=5729","url_meta":{"origin":5710,"position":0},"title":"My New Article at American Scientist","author":"Damien P. Williams","date":"July 7, 2023","format":false,"excerpt":"As of this week, I have a new article in the July-August 2023 Special Issue of American Scientist Magazine. It's called \"Bias Optimizers,\" and it's all about the problems and potential remedies of and for GPT-type tools and other \"A.I.\" This article picks up and expands on thoughts started in\u2026","rel":"","context":"In \"ableism\"","block_context":{"text":"ableism","link":"https:\/\/afutureworththinkingabout.com\/?tag=ableism"},"img":{"alt_text":"","src":"","width":0,"height":0},"classes":[]},{"id":5899,"url":"https:\/\/afutureworththinkingabout.com\/?p=5899","url_meta":{"origin":5710,"position":1},"title":"ChatGPT is Actively Marketing to Students During University Finals Season","author":"Damien P. Williams","date":"April 4, 2025","format":false,"excerpt":"It's really disheartening and honestly kind of telling that in spite of everything, ChatGPT is actively marketing itself to students in the run-up to college finals season. We've talked many (many) times before about the kinds of harm that can come from giving over too much epistemic and heuristic authority\u2026","rel":"","context":"In \"A Future Worth Thinking About\"","block_context":{"text":"A Future Worth Thinking About","link":"https:\/\/afutureworththinkingabout.com\/?tag=a-future-worth-thinking-about"},"img":{"alt_text":"Screenshot of ChatpGPT page:ChaptGPT Promo: 2 months free for students ChatGPT Plus is now free for college students through May Offer valid for students in the US and Canada [Buttons reading \"Claim offer\" and \"learn more\" An image of a pencil scrawling a scribbly and looping line] ChatGPT Plus is here to help you through finals","src":"https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:ybkylffhwhn2an2ic2lxh76k\/bafkreidh6mhffosfxhbgnxx6aybjycvgj3c2ygzto2xhzvsohdsv3g6evm@jpeg","width":350,"height":200,"srcset":"https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:ybkylffhwhn2an2ic2lxh76k\/bafkreidh6mhffosfxhbgnxx6aybjycvgj3c2ygzto2xhzvsohdsv3g6evm@jpeg 1x, https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:ybkylffhwhn2an2ic2lxh76k\/bafkreidh6mhffosfxhbgnxx6aybjycvgj3c2ygzto2xhzvsohdsv3g6evm@jpeg 1.5x, https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:ybkylffhwhn2an2ic2lxh76k\/bafkreidh6mhffosfxhbgnxx6aybjycvgj3c2ygzto2xhzvsohdsv3g6evm@jpeg 2x"},"classes":[]},{"id":5249,"url":"https:\/\/afutureworththinkingabout.com\/?p=5249","url_meta":{"origin":5710,"position":2},"title":"“We Built Them From Us”: My Appearance on the TEAM HUMAN Podcast","author":"Damien P. Williams","date":"February 22, 2018","format":false,"excerpt":"Earlier this month I was honoured to have the opportunity to sit and talk to Douglas Rushkoff on his TEAM HUMAN podcast. If you know me at all, you know this isn't by any means the only team for which I play, or even the only way I think about\u2026","rel":"","context":"In \"algorithmic bias\"","block_context":{"text":"algorithmic bias","link":"https:\/\/afutureworththinkingabout.com\/?tag=algorithmic-bias"},"img":{"alt_text":"","src":"","width":0,"height":0},"classes":[]},{"id":5345,"url":"https:\/\/afutureworththinkingabout.com\/?p=5345","url_meta":{"origin":5710,"position":3},"title":"Of Predictive Algorithms and Emojis","author":"Damien P. Williams","date":"January 8, 2019","format":false,"excerpt":"We do a lot of work and have a lot of conversations around here with people working on the social implications of technology, but some folx sometimes still don't quite get what I mean when I say that our values get embedded in our technological systems, and that the values\u2026","rel":"","context":"In \"algorithmic bias\"","block_context":{"text":"algorithmic bias","link":"https:\/\/afutureworththinkingabout.com\/?tag=algorithmic-bias"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/afutureworththinkingabout.com\/wp-content\/uploads\/2019\/01\/20190114_231648.jpg?resize=350%2C200&ssl=1","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/afutureworththinkingabout.com\/wp-content\/uploads\/2019\/01\/20190114_231648.jpg?resize=350%2C200&ssl=1 1x, https:\/\/i0.wp.com\/afutureworththinkingabout.com\/wp-content\/uploads\/2019\/01\/20190114_231648.jpg?resize=525%2C300&ssl=1 1.5x, https:\/\/i0.wp.com\/afutureworththinkingabout.com\/wp-content\/uploads\/2019\/01\/20190114_231648.jpg?resize=700%2C400&ssl=1 2x, https:\/\/i0.wp.com\/afutureworththinkingabout.com\/wp-content\/uploads\/2019\/01\/20190114_231648.jpg?resize=1050%2C600&ssl=1 3x"},"classes":[]},{"id":6422,"url":"https:\/\/afutureworththinkingabout.com\/?p=6422","url_meta":{"origin":5710,"position":4},"title":"Failures of “AI” Promise: Critical Thinking, Misinformation, Prosociality, & Trust","author":"Damien P. Williams","date":"December 6, 2025","format":false,"excerpt":"So, new research shows that a) LLM-type \"AI\" chatbots are extremely persuasive and able to get voters to shift their positions, and that b) the more effective they are at that, the less they hew to factual reality. Which: Yeah. A bunch of us told you this. Again: the Purpose\u2026","rel":"","context":"In \"A Future Worth Thinking About\"","block_context":{"text":"A Future Worth Thinking About","link":"https:\/\/afutureworththinkingabout.com\/?tag=a-future-worth-thinking-about"},"img":{"alt_text":"Kurt Russell as MacReady from The Thing, a white man with shoulder-length hair and a long scruff beard, wearing grey and olive drab, looking exhausted and sitting next to a bottle of J&B Rare Blend Scotch whisky and a pint glass 1\/3 full of the same, saying into a microphone, \u201cNobody trusts anybody now. And we\u2019re all very tired.\u201d","src":"https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:hiz3i44kravwkf6hmzg42okb\/bafkreia5rzux45474gfqegihm6lu23lskhsk77r4aoesjlelvkax4gsbyi@jpeg","width":350,"height":200,"srcset":"https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:hiz3i44kravwkf6hmzg42okb\/bafkreia5rzux45474gfqegihm6lu23lskhsk77r4aoesjlelvkax4gsbyi@jpeg 1x, https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:hiz3i44kravwkf6hmzg42okb\/bafkreia5rzux45474gfqegihm6lu23lskhsk77r4aoesjlelvkax4gsbyi@jpeg 1.5x, https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:hiz3i44kravwkf6hmzg42okb\/bafkreia5rzux45474gfqegihm6lu23lskhsk77r4aoesjlelvkax4gsbyi@jpeg 2x, https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:hiz3i44kravwkf6hmzg42okb\/bafkreia5rzux45474gfqegihm6lu23lskhsk77r4aoesjlelvkax4gsbyi@jpeg 3x, https:\/\/cdn.bsky.app\/img\/feed_fullsize\/plain\/did:plc:hiz3i44kravwkf6hmzg42okb\/bafkreia5rzux45474gfqegihm6lu23lskhsk77r4aoesjlelvkax4gsbyi@jpeg 4x"},"classes":[]},{"id":5295,"url":"https:\/\/afutureworththinkingabout.com\/?p=5295","url_meta":{"origin":5710,"position":5},"title":"At HPE: “4 obstacles to ethical AI (and how to address them)”","author":"Damien P. Williams","date":"July 8, 2018","format":false,"excerpt":"I talked with Hewlett Packard Enterprise's Curt Hopkins, for their article\u00a0\"4 obstacles to ethical AI (and how to address them).\" We spoke about the kinds of specific tools and techniques by which people who populate or manage artificial intelligence design teams can incorporate expertise from the humanities and social sciences.\u2026","rel":"","context":"In \"A Future Worth Thinking About\"","block_context":{"text":"A Future Worth Thinking About","link":"https:\/\/afutureworththinkingabout.com\/?tag=a-future-worth-thinking-about"},"img":{"alt_text":"","src":"https:\/\/i0.wp.com\/web.archive.org\/web\/20201109005732\/https%3A\/\/www.hpe.com\/content\/dam\/hpe\/insights\/articles\/2018\/07\/4-obstacles-to-ethical-ai-and-how-to-address-them\/featuredStory\/How-to-fix-AI.jpg.transform\/nxt-1043x496-crop\/image.jpeg?resize=350%2C200","width":350,"height":200,"srcset":"https:\/\/i0.wp.com\/web.archive.org\/web\/20201109005732\/https%3A\/\/www.hpe.com\/content\/dam\/hpe\/insights\/articles\/2018\/07\/4-obstacles-to-ethical-ai-and-how-to-address-them\/featuredStory\/How-to-fix-AI.jpg.transform\/nxt-1043x496-crop\/image.jpeg?resize=350%2C200 1x, https:\/\/i0.wp.com\/web.archive.org\/web\/20201109005732\/https%3A\/\/www.hpe.com\/content\/dam\/hpe\/insights\/articles\/2018\/07\/4-obstacles-to-ethical-ai-and-how-to-address-them\/featuredStory\/How-to-fix-AI.jpg.transform\/nxt-1043x496-crop\/image.jpeg?resize=525%2C300 1.5x, https:\/\/i0.wp.com\/web.archive.org\/web\/20201109005732\/https%3A\/\/www.hpe.com\/content\/dam\/hpe\/insights\/articles\/2018\/07\/4-obstacles-to-ethical-ai-and-how-to-address-them\/featuredStory\/How-to-fix-AI.jpg.transform\/nxt-1043x496-crop\/image.jpeg?resize=700%2C400 2x"},"classes":[]}],"_links":{"self":[{"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=\/wp\/v2\/posts\/5710","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=5710"}],"version-history":[{"count":7,"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=\/wp\/v2\/posts\/5710\/revisions"}],"predecessor-version":[{"id":5718,"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=\/wp\/v2\/posts\/5710\/revisions\/5718"}],"wp:attachment":[{"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5710"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5710"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/afutureworththinkingabout.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5710"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}