{"id":64723,"date":"2025-10-07T09:04:08","date_gmt":"2025-10-07T09:04:08","guid":{"rendered":"https:\/\/www.newsbeep.com\/ie\/64723\/"},"modified":"2025-10-07T09:04:08","modified_gmt":"2025-10-07T09:04:08","slug":"what-if-life-is-just-another-kind-of-computer","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/ie\/64723\/","title":{"rendered":"What If Life Is Just Another Kind of Computer?"},"content":{"rendered":"

\"\" <\/a>Credit: ZME Science\/Midjourney. <\/p>\n

In 1994, a strange, pixelated machine came to life on a computer screen. It read a string of instructions, copied them, and built a clone of itself \u2014 just as the Hungarian-American Polymath John von Neumann had predicted half a century earlier. It was a striking demonstration of a profound idea: that life, at its core, might be computational.<\/p>\n

Although this is seldom fully appreciated, von Neumann was one of the first to establish a deep link between life and computation. Reproduction, like computation, he showed, could be carried out by machines following coded instructions. In his model, based on Alan Turing\u2019s Universal Machine, self-replicating systems read and execute instructions much like DNA does: \u201cif\u00a0the next instruction is the codon CGA,\u00a0then\u00a0add an arginine to the protein under construction.\u201d It\u2019s not a metaphor to call DNA a \u201cprogram\u201d \u2014 that is literally the case.<\/p>\n

Of course, there are meaningful differences between biological computing and the kind of digital computing done by a personal computer or your smartphone. DNA is subtle and multilayered, including phenomena like epigenetics and gene proximity effects. Cellular DNA is nowhere near the whole story, either. Our bodies contain (and continually swap) countless bacteria and viruses, each running their own code.<\/p>\n

It\u2019s not a metaphor to call DNA a \u201cprogram\u201d \u2014 that is literally the case.<\/p>\n

Biological computing is \u201cmassively parallel<\/a>,\u201d decentralized, and noisy. Your cells have somewhere in the neighborhood of 300\u00a0quintillion\u00a0ribosomes, all working at the same time. Each of these exquisitely complex floating protein factories is, in effect, a tiny computer \u2014 albeit a stochastic one, meaning not entirely predictable. The movements of hinged components, the capture and release of smaller molecules, and the manipulation of chemical bonds are all individually random, reversible, and inexact, driven this way and that by constant thermal buffeting. Only a statistical asymmetry favors one direction over another, with clever origami moves tending to \u201clock in\u201d certain steps such that a next step becomes likely to happen.<\/p>\n

This differs greatly from the operation of \u201clogic gates\u201d in a computer, basic components that process binary inputs into outputs using fixed rules. They are irreversible and engineered to be 99.99 percent reliable and reproducible.<\/p>\n

Biological computing is computing, nonetheless. And its use of randomness is a feature, not a bug. In fact, many classic algorithms in computer science also require randomness (albeit for different reasons), which may explain why Turing insisted that the Ferranti Mark I, an early computer he helped to design in 1951, include a random number instruction. Randomness is thus a small but important conceptual extension to the original Turing Machine, though any computer can simulate it by calculating deterministic but random-looking or \u201cpseudorandom\u201d numbers.<\/p>\n

Parallelism, too, is increasingly fundamental to computing today. Modern AI, for instance, depends on both massive parallelism\u00a0and\u00a0randomness \u2014 as in the parallelized \u201cstochastic gradient descent\u201d (SGD) algorithm, used for training most of today\u2019s neural nets, the \u201ctemperature\u201d setting used in chatbots to introduce a degree of randomness into their output, and the parallelism of Graphics Processing Units (GPUs), which power most AI in data centers.<\/p>\n

Traditional digital computing, which relies on the centralized, sequential execution of instructions, was a product of technological constraints. The first computers needed to carry out long calculations using as few parts as possible. Originally, those parts were flaky, expensive vacuum tubes, which had a tendency to burn out and needed frequent replacement by hand. The natural design, then, was a minimal \u201cCentral Processing Unit\u201d (CPU) operating on sequences of bits ferried back and forth from an external memory. This has come to be known as the \u201cvon Neumann architecture.\u201d<\/p>\n

Turing and von Neumann were both aware that computing could be done by other means, though. Turing, near the end of his life, explored how biological patterns like leopard spots could arise from simple chemical rules, in a field he called morphogenesis. Turing\u2019s model of morphogenesis was a biologically inspired form of massively parallel, distributed computation. So was his earlier concept of an \u201cunorganized machine,\u201d a randomly connected neural net modeled after an infant\u2019s brain.<\/p>\n

These were visions of what computing without a central processor could look like \u2014 and what it\u00a0does\u00a0look like, in living systems.<\/p>\n

Von Neumann also began exploring massively parallel approaches to computation as far back as the 1940s. In discussions with Polish mathematician Stanis\u0142aw Ulam at Los Alamos, he conceived the idea of \u201ccellular automata,\u201d pixel-like grids of simple computational units, all obeying the same rule, and all altering their states simultaneously by communicating only with their immediate neighbors. With characteristic bravura, von Neumann went so far as to design, on paper, the key components of a\u00a0self-reproducing\u00a0cellular automaton, including a horizontal \u201ctape\u201d of cells containing instructions and blocks of cellular \u201ccircuitry\u201d for reading, copying, and executing them.<\/p>\n

Designing a cellular automaton is far harder than ordinary programming, because every cell or \u201cpixel\u201d is simultaneously altering its own state and its environment. Add randomness and subtle feedback effects, as in biology, and it becomes even harder to reason about, \u201cprogram,\u201d or \u201cdebug.\u201d<\/p>\n

With characteristic bravura, von Neumann went so far as to design, on paper, the key components of a\u00a0self-reproducing\u00a0cellular automaton.<\/p>\n

Nonetheless, Turing and von Neumann grasped something fundamental: Computation doesn\u2019t require a central processor, logic gates, binary arithmetic, or sequential programs. There are infinite ways to compute, and, crucially, they are all equivalent. This insight is one of the greatest accomplishments of theoretical computer science.<\/p>\n

This \u201cplatform independence\u201d or \u201cmultiple realizability\u201d means that any computer can emulate any other one. If the computers are of different designs, though, the emulation may be glacially slow. For that reason, von Neumann\u2019s self-reproducing cellular automaton has never been physically built \u2014 though that would be fun to see!<\/p>\n

That demonstration in 1994 \u2014\u00a0the first successful emulation of von Neumann\u2019s self-reproducing automation<\/a>\u00a0\u2014 couldn\u2019t have happened much earlier. A serial computer requires serious processing power to loop through the automaton\u2019s 6,329 cells over the 63\u00a0billion\u00a0time steps required for the automaton to complete its reproductive cycle. Onscreen, it worked as advertised: a pixelated two-dimensional Rube Goldberg machine, squatting astride a 145,315-cell\u2013long instruction tape trailing off to the right, pumping information out of the tape and reaching out with a \u201cwriting arm\u201d to slowly print a working clone of itself just above and to the right of the original.<\/p>\n

It\u2019s similarly inefficient for a serial computer to emulate a parallel neural network, heir to Turing\u2019s \u201cunorganized machine.\u201d Consequently, running big neural nets like those in Transformer-based chatbots has only recently become practical, thanks to ongoing progress in the miniaturization, speed, and parallelism of digital computers.<\/p>\n

In 2020, my colleague\u00a0Alex Mordvintsev<\/a>\u00a0combined modern neural nets, Turing\u2019s morphogenesis, and von Neumann\u2019s cellular automata into the \u201cneural cellular automaton\u201d (NCA), replacing the simple per-pixel rule of a classic cellular automaton with a neural net. This net, capable of sensing and affecting a few values representing local morphogen concentrations, can be trained to \u201cgrow\u201d any desired pattern or image, not just zebra stripes or leopard spots.<\/p>\n

Real cells don\u2019t literally have neural nets inside them, but they do run highly evolved, nonlinear, and purposive \u201cprograms\u201d to decide on the actions they will take in the world, given external stimulus and an internal state. NCAs offer a general way to model the range of possible behaviors of cells whose actions don\u2019t involve movement, but only changes of state (here, represented as color) and the absorption or release of chemicals.<\/p>\n

The first NCA Alex showed me was of a lizard emoji,\u00a0which could regenerate not only its tail, but also its limbs and head<\/a>! It was a powerful demonstration of how complex multicellular life can \u201cthink locally\u201d yet \u201cact globally,\u201d even when each cell (or pixel) is running the same program \u2014 just as each of your cells is running the same DNA. Simulations like these show how computation can produce lifelike behavior across scales. Building on von Neumann\u2019s designs and extending into modern neural cellular automata, they offer a glimpse into the computational underpinnings of living systems.<\/p>\n

This article is adapted from Blaise Ag\u00fcera y Arcas\u2019s book \u201cWhat Is Intelligence?<\/a>\u201d An open access edition of the book is\u00a0available here<\/a>. The article originally appeared on The MIT Press Reader<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"Credit: ZME Science\/Midjourney. In 1994, a strange, pixelated machine came to life on a computer screen. It read…\n","protected":false},"author":2,"featured_media":64724,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[21],"tags":[18292,4312,241,619,61,60,44236,326,80],"class_list":{"0":"post-64723","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-computing","8":"tag-alan-turing","9":"tag-computers","10":"tag-computing","11":"tag-dna","12":"tag-ie","13":"tag-ireland","14":"tag-john-von-neumann","15":"tag-life","16":"tag-technology"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/64723","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/comments?post=64723"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/posts\/64723\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media\/64724"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/media?parent=64723"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/categories?post=64723"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/ie\/wp-json\/wp\/v2\/tags?post=64723"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}