{"id":3237,"slug":"neural-network","name":"neural network","type":"concept","created_at":"2026-08-08T04:52:37.531011+00:00","hidden":0,"aliases":["neural nets","neural networks"],"episodes":[{"id":"a5dc4d35-2cc3-4a1f-966d-eab4644094c8","slug":"s19e11-your-model-is-naive-and-its-not-the","subject":"s19e11: Your Model is Naive and it’s Not The Territory","title":"Your Model is Naive and it’s Not The Territory","publish_date":"2025-05-30T18:21:31+00:00","year":2025,"month":5,"season":19,"episode":11,"episode_label":"s19e11","canonical_url":"https://newsletter.danhon.com/archive/s19e11-your-model-is-naive-and-its-not-the/","is_premium":0,"word_count":1523,"context":" … The architecture of certain <mark>neural networks</mark>\nliterally\nhas the concept of hidden layers. The models are complex enough when encoded in legislation and then a business rules engine … ","reading_minutes":8},{"id":"5a32226e-5b05-4574-bd97-ecf14aed5470","slug":"s12e58-we-can-hallucinate-it-for-your-wholescale","subject":"s12e58: We can hallucinate it for your wholescale","title":"We can hallucinate it for your wholescale","publish_date":"2022-09-28T17:28:36.435663+00:00","year":2022,"month":9,"season":12,"episode":58,"episode_label":"s12e58","canonical_url":"https://newsletter.danhon.com/archive/s12e58-we-can-hallucinate-it-for-your-wholescale/","is_premium":0,"word_count":1627,"context":" … its different <mark>neural networks</mark>) to encode an image into the\nlatent space\nof of the model, which (I'm handwaving, bear with me) roughly means you take all … ","reading_minutes":8},{"id":"45ddeb62-533e-47ff-a38d-1e36e7312058","slug":"s12e09-actually-its-about-the-ethics-in-dataset","subject":"s12e09: Actually, it's about the ethics in dataset collection, attribution and licensing in machine learning","title":"Actually, it's about the ethics in dataset collection, attribution and licensing in machine learning","publish_date":"2022-05-24T16:54:33.470664+00:00","year":2022,"month":5,"season":12,"episode":9,"episode_label":"s12e09","canonical_url":"https://newsletter.danhon.com/archive/s12e09-actually-its-about-the-ethics-in-dataset/","is_premium":0,"word_count":853,"context":" … A team at UC Berkeley created the\nBerkeley Crossword Solver\n[1]\n, which uses a <mark>neural network</mark> to solve crossword puzzles.\n\nCaught my attention because: a big deal because … ","reading_minutes":4},{"id":"3e8bbb1c-e8fd-4c84-953c-0efd07978055","slug":"s09e17-the-sky-above-the-port-snow-crashing","subject":"s09e17: The Sky Above The Port; Snow Crashing, chapter 14, part 3","title":"The Sky Above The Port; Snow Crashing, chapter 14, part 3","publish_date":"2021-04-19T17:49:10.956567+00:00","year":2021,"month":4,"season":9,"episode":17,"episode_label":"s09e17","canonical_url":"https://newsletter.danhon.com/archive/s09e17-the-sky-above-the-port-snow-crashing/","is_premium":0,"word_count":4356,"context":" … a) there’s no mention of <mark>neural networks</mark> as a computing or artificial intelligence technique in Snow Crash; b) there’s only one mention of the space shuttle … ","reading_minutes":22},{"id":"971e0bed-a84f-4006-874e-1d3ba9105a0b","slug":"s09e14-get-a-cozy-data-tiny-house-redundant","subject":"s09e14: Get a cozy Data Tiny House, Redundant Arrays of Inexpensive People","title":"Get a cozy Data Tiny House, Redundant Arrays of Inexpensive People","publish_date":"2021-04-09T22:17:29.605824+00:00","year":2021,"month":4,"season":9,"episode":14,"episode_label":"s09e14","canonical_url":"https://newsletter.danhon.com/archive/s09e14-get-a-cozy-data-tiny-house-redundant/","is_premium":0,"word_count":2199,"context":" … in_the_United_Kingdom)\n\nMeanwhile, I’m reminded of when I\ntrained a <mark>neural network</mark> to generate British place names\nAnd British place names\nnearly 4 years ago … ","reading_minutes":11},{"id":"e307b2d4-bd75-4e39-83e0-bce50ec2c86f","slug":"s08e09-titles-am-i-right","subject":"s08e09: Titles, am I right?","title":"Titles, am I right?","publish_date":"2020-05-10T05:44:32.580000+00:00","year":2020,"month":5,"season":8,"episode":9,"episode_label":"s08e09","canonical_url":"https://newsletter.danhon.com/archive/s08e09-titles-am-i-right/","is_premium":0,"word_count":1447,"context":" … I will not pretend to understand how this <mark>neural network</mark> model is able to\nanswer natural language questions about data encoded in tables\nand have to admit that … ","reading_minutes":7},{"id":"b6df7a0a-99c3-4e99-b9e6-ed0f8f3c9c2b","slug":"s07e17-snow-crashs-graveyard-daemons-and","subject":"s07e17: Snow Crash’s Graveyard Daemons and","title":"Snow Crash’s Graveyard Daemons and","publish_date":"2019-10-29T18:36:02.311000+00:00","year":2019,"month":10,"season":7,"episode":17,"episode_label":"s07e17","canonical_url":"https://newsletter.danhon.com/archive/s07e17-snow-crashs-graveyard-daemons-and/","is_premium":0,"word_count":5010,"context":" … Digicortex\nis an engine for simulating and visualizing large-scale “biologically plausible <mark>neural networks</mark>” and was interesting to me when I saw it via a benchmark on Intel … ","reading_minutes":25},{"id":"6ab47a23-5e70-46f3-9f85-3847343abd30","slug":"s07e16-with-the-word-police-on-it","subject":"s07e16: “With the word POLICE on it”","title":"“With the word POLICE on it”","publish_date":"2019-10-24T00:25:33.342000+00:00","year":2019,"month":10,"season":7,"episode":16,"episode_label":"s07e16","canonical_url":"https://newsletter.danhon.com/archive/s07e16-with-the-word-police-on-it/","is_premium":0,"word_count":2903,"context":" … Why wouldn’t a thing that we’ve designed to learn based on our knowledge of biological <mark>neural networks</mark> learn the way we think we learn, at a … ","reading_minutes":15},{"id":"4699d5b3-da46-49e4-9010-98d2cbc28ebf","slug":"s06e03-sherlock-holmes-and-the-slack-channel-names","subject":"s06e03: Sherlock Holmes and the Slack Channel Names","title":"Sherlock Holmes and the Slack Channel Names","publish_date":"2018-08-21T00:00:00+00:00","year":2018,"month":8,"season":6,"episode":3,"episode_label":"s06e03","canonical_url":"https://newsletter.danhon.com/archive/s06e03-sherlock-holmes-and-the-slack-channel-names/","is_premium":0,"word_count":1775,"context":null,"reading_minutes":9},{"id":"16512c4f-551f-447f-8995-1225ba02224e","slug":"s4e13-a-future-of-working","subject":"s4e13: A future of working","title":"A future of working","publish_date":"2017-05-19T00:00:00+00:00","year":2017,"month":5,"season":4,"episode":13,"episode_label":"s04e13","canonical_url":"https://newsletter.danhon.com/archive/s4e13-a-future-of-working/","is_premium":0,"word_count":2284,"context":" … in the style of your favourite painter from China and b) a recurrent <mark>neural network</mark> doing exactly the same thing.\n\n\n\nThis is perhaps one of the reasons why … ","reading_minutes":11},{"id":"220174cd-6d2a-4dca-81a6-8ad355cbc5d1","slug":"s4e12-end-of-process","subject":"s4e12: End of Process","title":"End of Process","publish_date":"2017-04-14T00:00:00+00:00","year":2017,"month":4,"season":4,"episode":12,"episode_label":"s04e12","canonical_url":"https://newsletter.danhon.com/archive/s4e12-end-of-process/","is_premium":0,"word_count":2033,"context":" … Machine learning techniques - most recently and commonly, <mark>neural networks</mark>[1] - are getting pretty unreasonably good at achieving outcomes opaquely. In that: we really wouldn't know where to … ","reading_minutes":10},{"id":"189db1c0-3cb2-476b-ab9d-1a554e51a753","slug":"s4e06-criticality","subject":"s4e06: Criticality","title":"Criticality","publish_date":"2017-03-15T00:00:00+00:00","year":2017,"month":3,"season":4,"episode":6,"episode_label":"s04e06","canonical_url":"https://newsletter.danhon.com/archive/s4e06-criticality/","is_premium":0,"word_count":1103,"context":" … Train a stupendously giant <mark>neural network</mark>, based on the kind of thing that can beat a human Go player, all the most successful things that get shared on … ","reading_minutes":6},{"id":"986450e6-7304-4e15-b670-04a63d478bf4","slug":"s3e35-interesting-times","subject":"s3e35: Interesting Times","title":"Interesting Times","publish_date":"2016-11-28T00:00:00+00:00","year":2016,"month":11,"season":3,"episode":35,"episode_label":"s03e35","canonical_url":"https://newsletter.danhon.com/archive/s3e35-interesting-times/","is_premium":0,"word_count":3145,"context":" … 2.0 Grab bag\nResearchers at MIT have published a paper about a deep-learning artificial <mark>neural network</mark> implemented in photonic circuitry instead of traditional Von-Neuman/Turing … ","reading_minutes":16},{"id":"defa8b8f-78a7-4bcd-aec4-505c7f88111d","slug":"s3e30-the-generative-surplus","subject":"s3e30: The Generative Surplus","title":"The Generative Surplus","publish_date":"2016-10-24T00:00:00+00:00","year":2016,"month":10,"season":3,"episode":30,"episode_label":"s03e30","canonical_url":"https://newsletter.danhon.com/archive/s3e30-the-generative-surplus/","is_premium":0,"word_count":1984,"context":" … a recurrent <mark>neural network</mark> that's a text generator, trained on a corpus of old science-fiction stories combined with an extensible text editor[2]. Please do take … ","reading_minutes":10},{"id":"8cf35d59-1b51-4f72-978b-67bbe67967a2","slug":"s3e28-something-is-better-than-nothing","subject":"s3e28: Something is better than nothing","title":"Something is better than nothing","publish_date":"2016-10-19T00:00:00+00:00","year":2016,"month":10,"season":3,"episode":28,"episode_label":"s03e28","canonical_url":"https://newsletter.danhon.com/archive/s3e28-something-is-better-than-nothing/","is_premium":0,"word_count":2215,"context":" … low-precision 8-bit wide fixed precision) maths vs high-precision (32-bit floating point) and how consciousness only spontaneously occurs in high-precision <mark>neural networks</mark>, resulting in … ","reading_minutes":11},{"id":"082edca2-421c-43b7-90d6-54bc6ad0bcd0","slug":"s3e23-it-looks-like-youre-trying-to-write-a","subject":"s3e23: It Looks Like You're Trying To Write A Newsletter","title":"It Looks Like You're Trying To Write A Newsletter","publish_date":"2016-06-21T00:00:00+00:00","year":2016,"month":6,"season":3,"episode":23,"episode_label":"s03e23","canonical_url":"https://newsletter.danhon.com/archive/s3e23-it-looks-like-youre-trying-to-write-a/","is_premium":0,"word_count":2060,"context":" … GPUS - graphics processing units - that are super good at some tasks like running <mark>neural networks</mark>! And! Those phones also happen to have pretty high resolution cameras on their … ","reading_minutes":10},{"id":"0d04cf8a-84cf-4252-b542-7072a2ffc36f","slug":"s3e22-things-that-have-caught-my-attention","subject":"s3e22: Things That Have Caught My Attention","title":"Things That Have Caught My Attention","publish_date":"2016-06-18T00:00:00+00:00","year":2016,"month":6,"season":3,"episode":22,"episode_label":"s03e22","canonical_url":"https://newsletter.danhon.com/archive/s3e22-things-that-have-caught-my-attention/","is_premium":0,"word_count":2129,"context":" … Seveneves struck me more, I think because William Gibson's The Peripheral is still knocking around in my head and right now I have a <mark>neural network</mark> that … ","reading_minutes":11},{"id":"4b00e391-bfd9-47c4-9e84-6fb0897b7dbd","slug":"s3e02-deep-dream-for-depression-unconscious-bias","subject":"s3e02: Deep Dream for Depression; Unconscious Bias Prophylactic","title":"Deep Dream for Depression; Unconscious Bias Prophylactic","publish_date":"2016-03-25T00:00:00+00:00","year":2016,"month":3,"season":3,"episode":2,"episode_label":"s03e02","canonical_url":"https://newsletter.danhon.com/archive/s3e02-deep-dream-for-depression-unconscious-bias/","is_premium":0,"word_count":1139,"context":" … At a macro level, this is how those deep dream images were generated:\n\n\n\n- set up a <mark>neural network</mark> that is a few layers deep - so one set of … ","reading_minutes":6},{"id":"cdfdba94-1ac6-4855-8448-829cfb002c80","slug":"s3e03-peaks-and-troughs-singletasking","subject":"s3e03: Peaks and Troughs; Singletasking","title":"Peaks and Troughs; Singletasking","publish_date":"2016-03-25T00:00:00+00:00","year":2016,"month":3,"season":3,"episode":3,"episode_label":"s03e03","canonical_url":"https://newsletter.danhon.com/archive/s3e03-peaks-and-troughs-singletasking/","is_premium":0,"word_count":1292,"context":" … of datasets like the MNIST digit set[9] and the recent viability of <mark>neural network</mark> approaches thanks to frankly insane amounts of GPU silicon.\n\n- Someone at Microsoft and … ","reading_minutes":6},{"id":"2d54c19b-24f8-4806-be42-b5890b0995d2","slug":"s2e33-black-design","subject":"s2e33: Black Design","title":"Black Design","publish_date":"2015-12-24T00:00:00+00:00","year":2015,"month":12,"season":2,"episode":33,"episode_label":"s02e33","canonical_url":"https://newsletter.danhon.com/archive/s2e33-black-design/","is_premium":0,"word_count":1584,"context":" … now feels like a choreographed, optimised movement that's been run through deep <mark>neural network</mark> simulations with data from hundreds, maybe thousands of meetings just like this one … ","reading_minutes":8},{"id":"543620d0-90f8-44c1-9c12-1760f4aa09bb","slug":"s2e23-recreational-capitalism-3-miscellanea","subject":"s2e23: Recreational Capitalism (3); Miscellanea","title":"Recreational Capitalism (3); Miscellanea","publish_date":"2015-08-25T00:00:00+00:00","year":2015,"month":8,"season":2,"episode":23,"episode_label":"s02e23","canonical_url":"https://newsletter.danhon.com/archive/s2e23-recreational-capitalism-3-miscellanea/","is_premium":0,"word_count":915,"context":" … I'm playing a bit more with the pet <mark>neural networks</mark> - I've got another one training on 25 megabytes worth of all State of the Union addresses … ","reading_minutes":5},{"id":"c1b4858a-59e5-4712-882e-57a1f9f90c30","slug":"s2e21-hello-babies-post-apocalyptic-duolingo","subject":"s2e21: Hello babies; Post-Apocalyptic Duolingo; Recreational Capitalism","title":"Hello babies; Post-Apocalyptic Duolingo; Recreational Capitalism","publish_date":"2015-08-19T00:00:00+00:00","year":2015,"month":8,"season":2,"episode":21,"episode_label":"s02e21","canonical_url":"https://newsletter.danhon.com/archive/s2e21-hello-babies-post-apocalyptic-duolingo/","is_premium":0,"word_count":3033,"context":" … have to do with other things.\n\n\n\n- Dan Hon (simple long short term model <mark>neural network</mark>, 400 neuron, 3 layer, Turing negative, unregistered) version: lm_lstm_epoch19.63_1 … ","reading_minutes":15},{"id":"e11eba4e-a98a-490f-85f5-23cc8687aead","slug":"s2e20-dreaming-of-writing-just-some-notes","subject":"s2e20: Dreaming of writing; Just some notes","title":"Dreaming of writing; Just some notes","publish_date":"2015-08-18T00:00:00+00:00","year":2015,"month":8,"season":2,"episode":20,"episode_label":"s02e20","canonical_url":"https://newsletter.danhon.com/archive/s2e20-dreaming-of-writing-just-some-notes/","is_premium":0,"word_count":1830,"context":" … It turns out that *that* kind of maths - vector maths or matrix maths - also happens to be the kind of stuff that <mark>neural networks</mark> do - or can do … ","reading_minutes":9},{"id":"e5e15415-9e98-4a64-85d3-2506bcffd51a","slug":"s2e19-uncanny-valley-network-du-jour-streaking","subject":"s2e19: Uncanny Valley; Network Du Jour; Streaking Considered Harmful","title":"Uncanny Valley; Network Du Jour; Streaking Considered Harmful","publish_date":"2015-08-15T00:00:00+00:00","year":2015,"month":8,"season":2,"episode":19,"episode_label":"s02e19","canonical_url":"https://newsletter.danhon.com/archive/s2e19-uncanny-valley-network-du-jour-streaking/","is_premium":0,"word_count":1215,"context":"0.0 Sitrep\n\n\n\nSomewhere, in the various kinds of memory that make up this laptop, there is a model of a <mark>neural network</mark>. That <mark>neural network</mark> has been … ","reading_minutes":6},{"id":"5c9f9c98-1710-4651-861a-08377f8e91e9","slug":"s2e18-sequencing-spectrum","subject":"s2e18: Sequencing; Spectrum","title":"Sequencing; Spectrum","publish_date":"2015-08-14T00:00:00+00:00","year":2015,"month":8,"season":2,"episode":18,"episode_label":"s02e18","canonical_url":"https://newsletter.danhon.com/archive/s2e18-sequencing-spectrum/","is_premium":0,"word_count":1134,"context":" … kick-ass workstations with unreasonable amounts of power are now not used just to render graphics, they're used to simulate or run giant <mark>neural networks</mark> to increase … ","reading_minutes":6},{"id":"86b95d0f-c4b1-4c8f-b85f-18ced116ffbd","slug":"s2e17-shockwave-the-top-priorities-engagement-a","subject":"s2e17: Shockwave; The Top Priorities; Engagement; A Hostile Environment","title":"Shockwave; The Top Priorities; Engagement; A Hostile Environment","publish_date":"2015-08-13T00:00:00+00:00","year":2015,"month":8,"season":2,"episode":17,"episode_label":"s02e17","canonical_url":"https://newsletter.danhon.com/archive/s2e17-shockwave-the-top-priorities-engagement-a/","is_premium":0,"word_count":2048,"context":" … I am about a quarter of the way through setting up a recurrent <mark>neural network</mark> and feeding it all of my newsletter output so that I can have … ","reading_minutes":10},{"id":"59ce4218-d428-4671-baa2-d4ad5202fe38","slug":"s2e16-battlesuits-simulations-and-puppyslugs","subject":"s2e16: Battlesuits; Simulations and Puppyslugs; Scratchpad","title":"Battlesuits; Simulations and Puppyslugs; Scratchpad","publish_date":"2015-08-11T00:00:00+00:00","year":2015,"month":8,"season":2,"episode":16,"episode_label":"s02e16","canonical_url":"https://newsletter.danhon.com/archive/s2e16-battlesuits-simulations-and-puppyslugs/","is_premium":0,"word_count":2388,"context":" … There's recurrent neural nets and there's convolutional neural nets, but all you really need to know is that they're <mark>neural networks</mark> and they have hidden … ","reading_minutes":12},{"id":"c0824a36-5c3d-42e7-9950-67355c62007c","slug":"s2e09-the-puppyslugs-car-phishing-24th-century","subject":"s2e09: The Puppyslugs; Car Phishing; 24th Century Procurement","title":"The Puppyslugs; Car Phishing; 24th Century Procurement","publish_date":"2015-07-27T00:00:00+00:00","year":2015,"month":7,"season":2,"episode":9,"episode_label":"s02e09","canonical_url":"https://newsletter.danhon.com/archive/s2e09-the-puppyslugs-car-phishing-24th-century/","is_premium":0,"word_count":949,"context":" … you read this.\n\n\n\n(Meanwhile, I am reading about how to seed a recurrent <mark>neural network</mark> with my entire newsletter writing output so I can go off and live … ","reading_minutes":5},{"id":"16a4f54d-b36a-483a-afbb-f6b1df171235","slug":"episode-two-hundred-and-three-compiling","subject":"Episode Two Hundred And Three: Compiling Consciousness.dat; No Apple TV; Negging, but for A.I.s","title":"Compiling Consciousness.dat; No Apple TV; Negging, but for A.I.s","publish_date":"2015-03-17T00:00:00+00:00","year":2015,"month":3,"season":null,"episode":203,"episode_label":"episode 203","canonical_url":"https://newsletter.danhon.com/archive/episode-two-hundred-and-three-compiling/","is_premium":0,"word_count":2296,"context":null,"reading_minutes":11},{"id":"0ad19702-0bdf-4973-a238-d73729e84af9","slug":"episode-two-hundred-and-one-the-system-of-the","subject":"Episode Two Hundred and One: The System Of The World; Hard Takeoff","title":"The System Of The World; Hard Takeoff","publish_date":"2015-03-04T00:00:00+00:00","year":2015,"month":3,"season":null,"episode":201,"episode_label":"episode 201","canonical_url":"https://newsletter.danhon.com/archive/episode-two-hundred-and-one-the-system-of-the/","is_premium":0,"word_count":1849,"context":" … It's good because it's a readable explanation of what convolutional <mark>neural networks</mark> are, when the idea came into being and why we're finally able to … ","reading_minutes":9},{"id":"bdd3e17f-7e19-42a1-b2d1-e9a01cd8fee9","slug":"episode-one-hundred-and-ninety-nine-just-no-dont","subject":"Episode One Hundred and Ninety Nine: Just No; Don't Do That, Either; Seeing Like The Internet","title":"Just No; Don't Do That, Either; Seeing Like The Internet","publish_date":"2015-02-28T00:00:00+00:00","year":2015,"month":2,"season":null,"episode":199,"episode_label":"episode 199","canonical_url":"https://newsletter.danhon.com/archive/episode-one-hundred-and-ninety-nine-just-no-dont/","is_premium":0,"word_count":2110,"context":" … You can tell that I've been reading about convolutional <mark>neural networks</mark> and that I've been going back and brushing up on my elementary understanding of the … ","reading_minutes":11},{"id":"f905cb19-a167-477c-bee0-e1846e467f4e","slug":"episode-one-hundred-and-ninety-seven-smarts-odds","subject":"Episode One Hundred and Ninety Seven: Smarts; Odds","title":"Smarts; Odds","publish_date":"2015-02-25T00:00:00+00:00","year":2015,"month":2,"season":null,"episode":197,"episode_label":"episode 197","canonical_url":"https://newsletter.danhon.com/archive/episode-one-hundred-and-ninety-seven-smarts-odds/","is_premium":0,"word_count":1123,"context":" … last financial quarter, e) whether Facebook has hired any more experts in covolutional <mark>neural networks</mark>, and so on. You might have - even if you didn't really understand … ","reading_minutes":6},{"id":"11e8a8b0-7c88-414d-a4d6-72fdaf8d099c","slug":"episode-one-hundred-and-eighty-seven-snow","subject":"Episode One Hundred and Eighty Seven: Snow Crashing (10); How The Web Works Now","title":"Snow Crashing (10); How The Web Works Now","publish_date":"2014-12-11T00:00:00+00:00","year":2014,"month":12,"season":null,"episode":187,"episode_label":"episode 187","canonical_url":"https://newsletter.danhon.com/archive/episode-one-hundred-and-eighty-seven-snow/","is_premium":0,"word_count":3715,"context":" … And, most likely, something that you can't reproduce with a convolutional <mark>neural network</mark> running on Google's data centers. Juanita's in the latter camp, an \"irrational … ","reading_minutes":19},{"id":"d614d787-c9d2-41c0-b710-83e7550b8ac6","slug":"episode-one-hundred-and-seventy-three-names-the","subject":"Episode One Hundred and Seventy Three: Names; The Condition","title":"Names; The Condition","publish_date":"2014-10-07T00:00:00+00:00","year":2014,"month":10,"season":null,"episode":173,"episode_label":"episode 173","canonical_url":"https://newsletter.danhon.com/archive/episode-one-hundred-and-seventy-three-names-the/","is_premium":0,"word_count":892,"context":" … Nothing more than that, just things that are lodged somewhere in an unfathomable <mark>neural network</mark>:\nHow Ebola Spread Out Of Control\n- The Washington Post\nFailing the GDS Service … ","reading_minutes":4},{"id":"d266a33a-95ea-48b3-83e2-fc15efcde482","slug":"episode-sixty-nine-transcendence-broom-shaped","subject":"Episode Sixty Nine: Transcendence; Broom-Shaped Objects; Odds","title":"Transcendence; Broom-Shaped Objects; Odds","publish_date":"2014-04-29T00:00:00+00:00","year":2014,"month":4,"season":null,"episode":69,"episode_label":"episode 69","canonical_url":"https://newsletter.danhon.com/archive/episode-sixty-nine-transcendence-broom-shaped/","is_premium":0,"word_count":2296,"context":" … It's Depp's actions as a Physically Independent <mark>Neural Network</mark> that scare them, not that they try to talk or reason with him anyway.\n- It's a … ","reading_minutes":11},{"id":"bde9d31b-ed7a-4a27-b292-65e8f35cb313","slug":"episode-forty-six-snow-crashing-3-video-is-a","subject":"Episode Forty Six - Snow Crashing 3; Video is a Content-Type; Blame Your Tools","title":"Snow Crashing 3; Video is a Content-Type; Blame Your Tools","publish_date":"2014-03-27T00:00:00+00:00","year":2014,"month":3,"season":null,"episode":46,"episode_label":"episode 46","canonical_url":"https://newsletter.danhon.com/archive/episode-forty-six-snow-crashing-3-video-is-a/","is_premium":0,"word_count":2691,"context":null,"reading_minutes":13},{"id":"e7aeb2a7-f1e4-4094-8ab1-25c07993c26a","slug":"episode-thirty-three-bbc-heaven-shorts","subject":"Episode Thirty Three: BBC Heaven; Shorts","title":"BBC Heaven; Shorts","publish_date":"2014-03-10T00:00:00+00:00","year":2014,"month":3,"season":null,"episode":33,"episode_label":"episode 33","canonical_url":"https://newsletter.danhon.com/archive/episode-thirty-three-bbc-heaven-shorts/","is_premium":0,"word_count":1637,"context":" … here's a paper on a gigantic <mark>neural network</mark> built (more accurately, built and trained) by Google to recognise numbers to help parse Streetview data. Remember that Google … ","reading_minutes":8}]}