{"id":5816,"date":"2025-01-14T07:00:00","date_gmt":"2025-01-14T06:00:00","guid":{"rendered":"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/"},"modified":"2025-01-14T07:00:00","modified_gmt":"2025-01-14T06:00:00","slug":"kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence","status":"publish","type":"post","link":"https:\/\/viva.racunalniske-novice.com\/en\/how-data-scarcity-threatens-the-future-of-artificial-intelligence\/","title":{"rendered":"How a lack of data threatens the future of artificial intelligence"},"content":{"rendered":"<p class=\"wp-block-paragraph\">The world of artificial intelligence is facing a shortage of its most valuable raw material \u2013 data. This has sparked discussions about an increasingly popular alternative: synthetic or even \u201cfake\u201d data. For years, companies like OpenAI and Google have been mining data from the internet to train the large language models (LLMs) that power their AI solutions. These models have digested vast amounts of human-generated content, from research papers and novels to YouTube videos.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Now, that data is slowly running out, its quantity becoming increasingly limited. Some major players in the field, such as OpenAI CEO Sam Altman, believe that self-learning models will be able to use synthetic data, which would provide a cheap and almost endless source of data.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">However, researchers warn of risks. Synthetic data could reduce the quality of models, as they can be \u201cpoisoned\u201d with their own errors. <a href=\"https:\/\/arxiv.org\/pdf\/2305.17493\" target=\"_blank\" rel=\"noreferrer noopener\">Research by Oxford and Cambridge universities<\/a> showed that feeding models exclusively with synthetic data leads to poor results and \u201cnonsense.\u201d In their opinion, a balanced use of synthetic and real data is key.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">More and more companies are creating synthetic data<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The lack of data is leading companies to look for alternatives, such as synthetic data generated by AI systems based on real data. Tech companies, including OpenAI and Google, are already paying millions to access data from platforms like Reddit and various media outlets, as websites increasingly restrict the free use of their content. Still, resources are limited.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Nvidia, Tencent, and startups Gretel and SynthLabs are developing tools to create synthetic data, which is often cleaner and more specific than human-generated data. Meta, with its Llama 3.1 model, has used synthetic data to improve skills such as programming and mathematical problem-solving. Synthetic data also offers the potential to reduce the bias inherent in real-world data, although researchers warn that ensuring accuracy and impartiality remains a major challenge.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u201cHabsburg\u201d artificial intelligence<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Although synthetic data brings advantages, it also poses serious risks. <a href=\"https:\/\/scontent-lhr8-1.xx.fbcdn.net\/v\/t39.2365-6\/453304228_1160109801904614_7143520450792086005_n.pdf?_nc_cat=108&amp;ccb=1-7&amp;_nc_sid=3c67a6&amp;_nc_ohc=PC3CtquZIecQ7kNvgEd56UN&amp;_nc_ht=scontent-lhr8-1.xx&amp;oh=00_AYCZLndlqJrzHln7YJPZgA20dTYBRdoZWwrQxxzEPpDRPQ&amp;oe=66B815C7\" target=\"_blank\" rel=\"noreferrer noopener\">Meta Research on the Llama 3.1 Model <\/a>showed that training a model on its own synthetic data can actually degrade its performance. Similarly, <a href=\"https:\/\/affiliate.insider.com\/?h=be468ee8ce72a81899f985f5f8550abdd05c3e150925e5b87d82d02e5a0c7498&amp;platform=browser&amp;postID=66acfb7ee3c9582388c4ffd0&amp;postSlug=ai-synthetic-data-industry-debate-over-fake-2024-8&amp;site=bi&amp;u=https%3A%2F%2Fwww.nature.com%2Farticles%2Fs41586-024-07566-y\" target=\"_blank\" rel=\"noreferrer noopener\">study in the journal Nature<\/a> warned that the uncontrolled use of synthetic data leads to \u201cmodel collapse,\u201d which the researchers compared to genetic degeneration and symbolically called the phenomenon \u201cHabsburg artificial intelligence.\u201d A term coined by researcher Jathan Sadowski.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The main question remains: how much synthetic data is too much? Some experts suggest using hybrid data, where synthetic data is combined with real data to prevent model degradation. Companies like Scale AI are exploring this approach, and their CEO Alexandr Wang believes that the hybrid approach is \u201cthe real future.\u201d<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Finding new solutions<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">In January, Google DeepMind unveiled AlphaGeometry, a system that solves geometric problems at an extremely high level using a \u201cneuro-symbolic\u201d approach. It combines the advantages of data-intensive deep learning and rule-based reasoning. The model was trained entirely on synthetic data and is seen as a potential step towards artificial general intelligence.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The field of neuro-symbolic is still young, but it could offer a promising direction for the future of AI development. Under pressure to monetize, companies like OpenAI, Google, and Microsoft will try all possible solutions to overcome the data crisis. <\/p>","protected":false},"excerpt":{"rendered":"<p>The world of artificial intelligence is facing a shortage of its most valuable raw material \u2013 data. This has sparked discussions about an increasingly popular alternative: synthetic or even \u201cfake\u201d data. For years, companies like OpenAI and Google have been mining data from the internet to train the large language models (LLMs) that power their AI solutions. These models have digested huge [\u2026]<\/p>","protected":false},"author":2,"featured_media":0,"comment_status":"","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[66,4],"tags":[192],"class_list":["post-5816","post","type-post","status-publish","format-standard","hentry","category-programi","category-racunalnistvo-telefonija","tag-umetna-inteligenca"],"acf":{"subtitle":"Umetna inteligenca se soo\u010da s pomanjkanjem klju\u010dnih podatkov, kar vodi k uporabi sinteti\u010dnih re\u0161itev. So \"la\u017eni\" podatki lahko prihodnost umetne inteligence ali tveganje za kakovost modelov in njihove zmogljivosti?","heading":"","summary":"Umetna inteligenca se soo\u010da s pomanjkanjem klju\u010dnih podatkov, kar vodi k uporabi sinteti\u010dnih re\u0161itev. So \"la\u017eni\" podatki lahko prihodnost umetne inteligence ali tveganje za kakovost modelov in njihove zmogljivosti?","thumbnail_small":"https:\/\/racunalniske-novice.com\/wp-content\/uploads\/2025\/01\/ali-shah-lakhani-sp1BZ1atp7M-unsplash-560x315.jpg","thumbnail_large":"https:\/\/racunalniske-novice.com\/wp-content\/uploads\/2025\/01\/ali-shah-lakhani-sp1BZ1atp7M-unsplash-768x1024.jpg","thumbnail_caption":"","gallery":"","video_gallery":null,"author":"","links":null,"sources":[{"title":"Business Insider","url":"https:\/\/www.businessinsider.com\/ai-synthetic-data-industry-debate-over-fake-2024-8?utm_source=chatgpt.com"},{"title":"Unsplash","url":"https:\/\/unsplash.com\/photos\/assorted-source-codes-sp1BZ1atp7M"}],"skip_language":[]},"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v22.8 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence - Ra\u010dunalni\u0161ke novice<\/title>\n<meta name=\"description\" content=\"Umetna inteligenca se soo\u010da s pomanjkanjem klju\u010dnih podatkov, kar vodi k uporabi sinteti\u010dnih re\u0161itev. So &quot;la\u017eni&quot; podatki lahko prihodnost umetne inteligence ali tveganje za kakovost modelov in njihove\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/posts\/5816\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence - Ra\u010dunalni\u0161ke novice\" \/>\n<meta property=\"og:description\" content=\"Svet umetne inteligence se soo\u010da s pomanjkanjem svoje najdragocenej\u0161e surovine \u2013 podatkov. To je spro\u017eilo razprave o vse bolj priljubljeni alternativi: sinteti\u010dnih ali celo &#8220;la\u017enih&#8221; podatkih. Dolga leta so podjetja, kot sta OpenAI in Google, za u\u010denje velikih jezikovnih modelov (LLM-jev), ki poganjajo njihove UI re\u0161itve, pridobivala podatke z interneta. Ti modeli so prebavili ogromne [&hellip;]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/viva.racunalniske-novice.com\/en\/how-data-scarcity-threatens-the-future-of-artificial-intelligence\/\" \/>\n<meta property=\"og:site_name\" content=\"Ra\u010dunalni\u0161ke novice\" \/>\n<meta property=\"article:published_time\" content=\"2025-01-14T06:00:00+00:00\" \/>\n<meta name=\"author\" content=\"sinusiks\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"sinusiks\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"3 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/\",\"url\":\"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/\",\"name\":\"Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence - Ra\u010dunalni\u0161ke novice\",\"isPartOf\":{\"@id\":\"https:\/\/viva.racunalniske-novice.com\/en\/#website\"},\"datePublished\":\"2025-01-14T06:00:00+00:00\",\"dateModified\":\"2025-01-14T06:00:00+00:00\",\"author\":{\"@id\":\"https:\/\/viva.racunalniske-novice.com\/en\/#\/schema\/person\/afb62e36efa34516d50249517e4cdbb4\"},\"breadcrumb\":{\"@id\":\"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/viva.racunalniske-novice.com\/en\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/viva.racunalniske-novice.com\/en\/#website\",\"url\":\"https:\/\/viva.racunalniske-novice.com\/en\/\",\"name\":\"Ra\u010dunalni\u0161ke novice\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/viva.racunalniske-novice.com\/en\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"en-US\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/viva.racunalniske-novice.com\/en\/#\/schema\/person\/afb62e36efa34516d50249517e4cdbb4\",\"name\":\"sinusiks\",\"sameAs\":[\"https:\/\/ml.racunalniske-novice.com\"],\"url\":\"https:\/\/viva.racunalniske-novice.com\/en\/author\/sinusiks\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence - Ra\u010dunalni\u0161ke novice","description":"Umetna inteligenca se soo\u010da s pomanjkanjem klju\u010dnih podatkov, kar vodi k uporabi sinteti\u010dnih re\u0161itev. So \"la\u017eni\" podatki lahko prihodnost umetne inteligence ali tveganje za kakovost modelov in njihove","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/posts\/5816","og_locale":"en_US","og_type":"article","og_title":"Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence - Ra\u010dunalni\u0161ke novice","og_description":"Svet umetne inteligence se soo\u010da s pomanjkanjem svoje najdragocenej\u0161e surovine \u2013 podatkov. To je spro\u017eilo razprave o vse bolj priljubljeni alternativi: sinteti\u010dnih ali celo &#8220;la\u017enih&#8221; podatkih. Dolga leta so podjetja, kot sta OpenAI in Google, za u\u010denje velikih jezikovnih modelov (LLM-jev), ki poganjajo njihove UI re\u0161itve, pridobivala podatke z interneta. Ti modeli so prebavili ogromne [&hellip;]","og_url":"https:\/\/viva.racunalniske-novice.com\/en\/how-data-scarcity-threatens-the-future-of-artificial-intelligence\/","og_site_name":"Ra\u010dunalni\u0161ke novice","article_published_time":"2025-01-14T06:00:00+00:00","author":"sinusiks","twitter_card":"summary_large_image","twitter_misc":{"Written by":"sinusiks","Est. reading time":"3 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/","url":"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/","name":"Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence - Ra\u010dunalni\u0161ke novice","isPartOf":{"@id":"https:\/\/viva.racunalniske-novice.com\/en\/#website"},"datePublished":"2025-01-14T06:00:00+00:00","dateModified":"2025-01-14T06:00:00+00:00","author":{"@id":"https:\/\/viva.racunalniske-novice.com\/en\/#\/schema\/person\/afb62e36efa34516d50249517e4cdbb4"},"breadcrumb":{"@id":"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/viva.racunalniske-novice.com\/kako-pomanjkanje-podatkov-ogroza-prihodnost-umetne-inteligence\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/viva.racunalniske-novice.com\/en\/"},{"@type":"ListItem","position":2,"name":"Kako pomanjkanje podatkov ogro\u017ea prihodnost umetne inteligence"}]},{"@type":"WebSite","@id":"https:\/\/viva.racunalniske-novice.com\/en\/#website","url":"https:\/\/viva.racunalniske-novice.com\/en\/","name":"Ra\u010dunalni\u0161ke novice","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/viva.racunalniske-novice.com\/en\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"en-US"},{"@type":"Person","@id":"https:\/\/viva.racunalniske-novice.com\/en\/#\/schema\/person\/afb62e36efa34516d50249517e4cdbb4","name":"sinusiks","sameAs":["https:\/\/ml.racunalniske-novice.com"],"url":"https:\/\/viva.racunalniske-novice.com\/en\/author\/sinusiks\/"}]}},"_links":{"self":[{"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/posts\/5816","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/comments?post=5816"}],"version-history":[{"count":0,"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/posts\/5816\/revisions"}],"wp:attachment":[{"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/media?parent=5816"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/categories?post=5816"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/viva.racunalniske-novice.com\/en\/wp-json\/wp\/v2\/tags?post=5816"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}