{"id":85771,"date":"2025-06-13T12:58:15","date_gmt":"2025-06-13T12:58:15","guid":{"rendered":"https:\/\/hanstimmerman.me\/?p=85771"},"modified":"2025-06-30T08:03:57","modified_gmt":"2025-06-30T08:03:57","slug":"dna-de-oervorm-van-data-opslag","status":"publish","type":"post","link":"https:\/\/hanstimmerman.me\/en\/dna-de-oervorm-van-data-opslag\/","title":{"rendered":"DNA: De Oervorm van Data-opslag"},"content":{"rendered":"<p style=\"text-align: right;\"><em><span style=\"color: #000000;\">\ud83c\uddec\ud83c\udde7English version: scroll down<\/span><\/em><\/p>\n<h5><span style=\"color: #000000;\">Een <b>zettabyte<\/b> is een <b>triljard bytes<\/b> \u2014 een 1 met 21 nullen. Omgerekend: \u00e9\u00e9n <b>miljoen petabytes<\/b>, ofwel <b>\u00e9\u00e9n biljoen gigabytes<\/b>. In 2012 overschreden we voor het eerst de grens van \u00e9\u00e9n zettabyte aan gegenereerde data. In 2016 werd \u00e9\u00e9n zettabyte via internet verstuurd. En nu, in 2025? Dan produceren we naar verwachting zo\u2019n <b>180 zettabyte<\/b> aan data wereldwijd \u2014 een hoeveelheid die ongeveer elke twee jaar verdubbelt.<\/span><\/h5>\n<p><strong><span style=\"color: #000000;\">Vooralsnog <span style=\"caret-color: #000000;\">exponenti\u00eble<\/span>\u00a0groei<\/span><\/strong><\/p>\n<p><span style=\"color: #000000;\">AI versterkt die groei exponentieel. En daarmee rijzen ook urgente vragen: <i>Waar laten we al die data? Hoeveel moet er bewaard blijven? En hoe lang?<\/i> Vragen die voor storage-engineers geen abstracte kwesties zijn, maar harde, fysieke uitdagingen. Want de cloud is geen wolk, maar een verzameling datacenters die gebouwd, gekoeld, gevoed en betaald moeten worden. We produceren inmiddels meer data dan we ooit kunnen opslaan. Veel daarvan bestaat slechts kortstondig in sensoren of streaming: gemaakt, gebruikt en weer verdwenen. Er is simpelweg niet genoeg fysieke opslagcapaciteit beschikbaar om het allemaal te bewaren.<\/span><\/p>\n<p><span style=\"color: #000000;\">Terwijl we systemen bouwen die schaken winnen, foto\u2019s genereren en agenda\u2019s beheren, blijft \u00e9\u00e9n verschil opvallend: zelfs de meest geavanceerde AI krabbelt pas aan het oppervlak van menselijke intelligentie. En op het gebied van data-opslag is het verschil misschien nog schrijnender. Want niets, echt niets, komt ook maar in de buurt van wat de natuur al miljarden jaren feilloos doet: DNA-gebaseerde opslag.<\/span><\/p>\n<p><span style=\"color: #000000;\"><b>De Ultieme Compacte Opslag: DNA<\/b><\/span><\/p>\n<p><span style=\"color: #000000;\">Elke menselijke cel bevat zo\u2019n 3 miljard baseparen aan genetische code. Dat staat gelijk aan tientallen gigabytes aan informatie \u2014 gecomprimeerd in een volume kleiner dan een stofkorrel. Zet dat af tegen onze meest geavanceerde opslagmedia en de conclusie is duidelijk: <b>DNA is nog steeds de meest compacte, duurzame en betrouwbare opslagtechnologie die we kennen.<\/b><b><\/b><\/span><\/p>\n<p><span style=\"color: #000000;\">In een <a style=\"color: #000000;\" href=\"https:\/\/www.newyorker.com\/culture\/the-lede\/could-we-store-our-data-in-dna?utm_source=chatgpt.com\">recent artikel<\/a> in <i>The New Yorker<\/i> \u2013 \u201cThe Race to Save the World\u2019s DNA\u201d \u2013 wordt beschreven hoe onderzoekers wereldwijd experimenteren met het opslaan van digitale bestanden in synthetisch DNA. Bestanden worden vertaald naar de letters A, C, G en T, en vervolgens moleculair gesynthetiseerd. De opslagcapaciteit? Tot <b>215 petabyte per gram DNA<\/b>. De houdbaarheid? <b>Tientallen tot honderdduizenden jaren<\/b> \u2014 mits goed bewaard.<\/span><\/p>\n<p><span style=\"color: #000000;\"><b>Epigenetica: Opslag \u00e9n Intelligentie<\/b><\/span><\/p>\n<p><span style=\"color: #000000;\">Maar de natuur doet meer dan alleen opslaan. <b>DNA werkt ook als processor.<\/b> Dankzij <b>epigenetica<\/b> \u2014 chemische schakelaars zoals methylgroepen \u2014 bepaalt een cel dynamisch welke genen worden \u2018gelezen\u2019 en wanneer. Geen centrale processor. Geen besturingssysteem. Geen cloud. Het is alsof elke cel haar eigen slimme, zelfregulerende dataserver is.<\/span><\/p>\n<p><span style=\"color: #000000;\">We zien hier een fascinerende analogie: zoals AI nog ver verwijderd is van menselijke intu\u00eftie, creativiteit en contextbegrip, zo blijven onze opslagtechnologie\u00ebn ver achter bij de elegantie en effici\u00ebntie van biologische opslag. Wij bouwen datacenters van duizenden vierkante meters en honderden megawatt. De natuur doet hetzelfde in een <b>eicel van 0,1 millimeter<\/b>.<\/span><\/p>\n<p><span style=\"color: #000000;\"><b>De Toekomst: DNA als Archiefdrager<\/b><\/span><\/p>\n<p><span style=\"color: #000000;\">Onderzoekers bouwen inmiddels prototypes van DNA-gebaseerde archieven. Microsoft Research slaagde erin om video\u2019s, tekst en muziek te coderen in synthetisch DNA. Start-ups ontwikkelen DNA-kopieerapparaten als alternatief voor magneetband. In biohackinglabs worden zelfs logische bewerkingen met DNA getest: <b>DNA-computing<\/b> in plaats van elektronica.<\/span><\/p>\n<p><span style=\"color: #000000;\">Toch staan we pas aan het begin. Synthese is duur. Lezen vereist complexe sequencing. Opslagsnelheden zijn traag. Maar de richting is duidelijk: een technologische ontdekkingstocht waarin de natuur al miljarden jaren voorloopt.<\/span><\/p>\n<p><span style=\"color: #000000;\"><b>Slimme Data: Van Dode Informatie naar Levend Archief<\/b><\/span><\/p>\n<p><span style=\"color: #000000;\">Veel van onze opgeslagen data is \u2018dom\u2019: platte teksten, statische beelden, simpele datasets. Metadata geeft wel wat extra context, maar data-objecten kunnen nog niet zelfstandig waarde cre\u00ebren. Zelfs met AI-agents ontbreekt het vaak aan begrip van <b>relaties, relevantie en context<\/b>.<\/span><\/p>\n<p><span style=\"color: #000000;\">Maar stel dat we metadata verrijken met betekenis: context, semantiek, doelgerichtheid. Dan kunnen informatie-objecten \u2014 met hulp van AI-agents en microservices \u2014 zelfstandig waarde opbouwen door verwante informatie te verzamelen en te combineren. <b>Van passieve archieven naar actieve kennisnetwerken.<\/b><b><\/b><\/span><\/p>\n<p><span style=\"color: #000000;\"><b>Een Lessenboek van de Natuur<\/b><\/span><\/p>\n<p><span style=\"color: #000000;\">Zoals onze hersenen herinneringen koppelen en associatief denken \u2014 zo zouden ook onze data-architecturen kunnen werken. Informatie die niet alleen vindbaar is, maar zichzelf organiseert en verbindt. Dan wordt opslag geen eindstation, maar een bron voor creativiteit, begrip en innovatie.<\/span><\/p>\n<p><span style=\"color: #000000;\">De natuur laat ons zien hoe het kan. <b>Veerkrachtig, energiezuinig, zelforganiserend.<\/b> Misschien is het tijd dat we \u2014 in plaats van de natuur te imiteren \u2014 eerst leren begrijpen hoe zij al miljarden jaren met data omgaat. En pas daarna, met gepaste nederigheid, proberen te bouwen wat zij allang beheerst.<\/span><\/p>\n<p>Photo by <a href=\"https:\/\/unsplash.com\/@sangharsh_l?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash\">Sangharsh Lohakare<\/a> on <a href=\"https:\/\/unsplash.com\/photos\/a-close-up-of-a-structure-of-a-structure-Iy7QyzOs1bo?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash\">Unsplash<\/a><\/p>\n<p style=\"text-align: center;\">&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211; \u00a0Translated by ChatGPT \u00a0&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8211;<\/p>\n<h2 data-start=\"276\" data-end=\"319\"><strong data-start=\"279\" data-end=\"319\">DNA: The Primal Form of Data Storage<\/strong><\/h2>\n<p data-start=\"321\" data-end=\"736\">A <strong data-start=\"323\" data-end=\"336\">zettabyte<\/strong> is a <strong data-start=\"342\" data-end=\"364\">trillion gigabytes<\/strong> \u2014 a 1 followed by 21 zeros. That\u2019s one <strong data-start=\"404\" data-end=\"425\">million petabytes<\/strong>, or a <strong data-start=\"432\" data-end=\"453\">billion terabytes<\/strong>. In 2012, we crossed the threshold of generating one zettabyte of data globally. In 2016, one zettabyte was transmitted across the internet. And now, in 2025? We&#8217;re expected to generate around <strong data-start=\"647\" data-end=\"665\">180 zettabytes<\/strong> of data this year \u2014 a volume that roughly <strong data-start=\"708\" data-end=\"735\">doubles every two years<\/strong>.<\/p>\n<h2 data-start=\"743\" data-end=\"776\"><strong data-start=\"746\" data-end=\"776\">Still an Exponential Surge<\/strong><\/h2>\n<p data-start=\"778\" data-end=\"1210\">AI is accelerating this growth exponentially. And with that, some fundamental questions arise: <em data-start=\"873\" data-end=\"952\">Where will we store all that data? How much should be kept? And for how long?<\/em> These are not abstract dilemmas for storage engineers \u2014 they\u2019re physical challenges with real-world consequences. The cloud, after all, is not a vaporous entity, but a collection of physical data centers that must be <strong data-start=\"1170\" data-end=\"1209\">built, cooled, powered and paid for<\/strong>.<\/p>\n<p data-start=\"1212\" data-end=\"1432\">We\u2019re now generating more data than we could ever store. Much of it exists only fleetingly \u2014 created by sensors, streamed, used, and gone. There simply isn\u2019t enough physical storage capacity available to preserve it all.<\/p>\n<p data-start=\"1434\" data-end=\"1835\">And while we build systems that win chess games, generate images, and manage calendars, one striking gap remains: even the most advanced AI only scratches the surface of human intelligence. And in the domain of data storage, the gap is perhaps even more stark. Because <strong data-start=\"1703\" data-end=\"1834\">nothing \u2014 truly nothing \u2014 comes close to what nature has already been doing flawlessly for billions of years: DNA-based storage<\/strong>.<\/p>\n<h2 data-start=\"1842\" data-end=\"1889\"><strong data-start=\"1845\" data-end=\"1889\">The Ultimate Compact Storage Medium: DNA<\/strong><\/h2>\n<p data-start=\"1891\" data-end=\"2240\">Every human cell contains about <strong data-start=\"1923\" data-end=\"1947\">3 billion base pairs<\/strong> of genetic code. That\u2019s equivalent to <strong data-start=\"1986\" data-end=\"2009\">dozens of gigabytes<\/strong> of information \u2014 compressed into a volume smaller than a speck of dust. Compare that to our most advanced storage media, and the conclusion is clear: <strong data-start=\"2160\" data-end=\"2240\">DNA is still the most compact, durable, and reliable storage system we know.<\/strong><\/p>\n<p data-start=\"2242\" data-end=\"2734\">In a recent <em data-start=\"2254\" data-end=\"2266\">New Yorker<\/em> article \u2014 <a class=\"cursor-pointer\" target=\"_new\" rel=\"noopener\" data-start=\"2277\" data-end=\"2392\">\u201cThe Race to Save the World\u2019s DNA\u201d<\/a> \u2014 researchers describe how they\u2019re experimenting with storing digital files in synthetic DNA. Files are translated into the letters A, C, G, and T, and then molecularly synthesized. The capacity? Up to <strong data-start=\"2595\" data-end=\"2621\">215 petabytes per gram<\/strong> of DNA. The durability? <strong data-start=\"2646\" data-end=\"2701\">Tens of thousands to hundreds of thousands of years<\/strong>, provided it\u2019s stored correctly.<\/p>\n<h2 data-start=\"2741\" data-end=\"2787\"><strong data-start=\"2744\" data-end=\"2787\">Epigenetics: Storage <em data-start=\"2767\" data-end=\"2772\">and<\/em> Intelligence<\/strong><\/h2>\n<p data-start=\"2789\" data-end=\"3122\">But nature does more than store data. <strong data-start=\"2827\" data-end=\"2859\">DNA also acts as a processor<\/strong>. Thanks to <strong data-start=\"2871\" data-end=\"2886\">epigenetics<\/strong> \u2014 chemical switches like methyl groups \u2014 each cell dynamically determines <strong data-start=\"2961\" data-end=\"2993\">which genes to read and when<\/strong>. No central processor. No operating system. No cloud. It\u2019s as if every cell is its own intelligent, self-regulating data server.<\/p>\n<p data-start=\"3124\" data-end=\"3504\">Here we find a fascinating analogy: just as AI is still far from achieving human intuition, creativity, and contextual awareness, our storage technologies remain rudimentary compared to the <strong data-start=\"3314\" data-end=\"3363\">elegance and efficiency of biological storage<\/strong>. We build data centers the size of aircraft hangars, consuming hundreds of megawatts. Nature does the same in a <strong data-start=\"3476\" data-end=\"3503\">0.1-millimeter egg cell<\/strong>.<\/p>\n<h2 data-start=\"3511\" data-end=\"3552\"><strong data-start=\"3514\" data-end=\"3552\">The Future: DNA as Archival Medium<\/strong><\/h2>\n<p data-start=\"3554\" data-end=\"3892\">Researchers are already building prototypes of DNA-based archives. Microsoft Research has successfully encoded video, text, and music into synthetic DNA. Start-ups are developing DNA copiers as alternatives to magnetic tape. And in biohacking labs, logical operations using DNA are being tested \u2014 <strong data-start=\"3851\" data-end=\"3868\">DNA computing<\/strong> instead of electronics.<\/p>\n<p data-start=\"3894\" data-end=\"4125\">But this is still the beginning. DNA synthesis is expensive. Reading requires complex sequencing. Storage speeds are slow. Yet the direction is clear: a technological exploration where nature has been leading for billions of years.<\/p>\n<h2 data-start=\"4132\" data-end=\"4191\"><strong data-start=\"4135\" data-end=\"4191\">Smart Data: From Dead Information to Living Archives<\/strong><\/h2>\n<p data-start=\"4193\" data-end=\"4462\">Much of our stored data is \u2018dead\u2019: flat text, static images, simple datasets. Metadata adds some context, but data objects <strong data-start=\"4316\" data-end=\"4356\">cannot yet create value on their own<\/strong>. Even with AI agents, there\u2019s often a lack of understanding of <strong data-start=\"4420\" data-end=\"4461\">relationships, relevance, and context<\/strong>.<\/p>\n<p data-start=\"4464\" data-end=\"4740\">But imagine enriching metadata with meaning: context, semantics, purpose. Information objects could then \u2014 supported by AI agents and microservices \u2014 autonomously <strong data-start=\"4627\" data-end=\"4642\">build value<\/strong> by gathering and linking related content. <strong data-start=\"4685\" data-end=\"4740\">From passive archives to active knowledge networks.<\/strong><\/p>\n<h2 data-start=\"4747\" data-end=\"4779\"><strong data-start=\"4750\" data-end=\"4779\">A Lesson Book from Nature<\/strong><\/h2>\n<p data-start=\"4781\" data-end=\"5058\">Just as our brains link memories and think associatively, so too could our data architectures function. Information not just indexed, but <strong data-start=\"4919\" data-end=\"4958\">self-organizing and self-connecting<\/strong>. Storage would no longer be a destination, but a <strong data-start=\"5008\" data-end=\"5057\">source of creativity, insight, and innovation<\/strong>.<\/p>\n<p data-start=\"5060\" data-end=\"5382\">Nature already shows us how it can be done \u2014 resilient, energy-efficient, and self-organizing. Perhaps it\u2019s time we stop merely trying to imitate nature, and start truly understanding <strong data-start=\"5244\" data-end=\"5294\">how she has handled data for billions of years<\/strong>. Only then, with humility, can we begin to build what she already masters effortlessly.<\/p>","protected":false},"excerpt":{"rendered":"<p>Just as our brains link memories and think associatively, so too could our data architectures function. Information not just indexed, but self-organizing and self-connecting. Storage would no longer be a destination, but a source of creativity, insight, and innovation.<\/p>\n<p>Nature already shows us how it can be done \u2014 resilient, energy-efficient, and self-organizing. Perhaps it\u2019s time we stop merely trying to imitate nature, and start truly understanding how she has handled data for billions of years. Only then, with humility, can we begin to build what she already masters effortlessly.<\/p>","protected":false},"author":3,"featured_media":85775,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[380,70,78],"tags":[],"class_list":["post-85771","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-archiving","category-data-ownership","category-storage"],"jetpack_featured_media_url":"https:\/\/i0.wp.com\/hanstimmerman.me\/wp-content\/uploads\/2025\/06\/sangharsh-lohakare-Iy7QyzOs1bo-unsplash-scaled-e1749818984807.jpg?fit=2550%2C1028&ssl=1","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/posts\/85771","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/comments?post=85771"}],"version-history":[{"count":6,"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/posts\/85771\/revisions"}],"predecessor-version":[{"id":85826,"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/posts\/85771\/revisions\/85826"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/media\/85775"}],"wp:attachment":[{"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/media?parent=85771"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/categories?post=85771"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/hanstimmerman.me\/en\/wp-json\/wp\/v2\/tags?post=85771"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}