1 00:00:01,760 --> 00:00:04,480 Speaker 1: Hi there, I'm your host, Lauren volg Obam, a researcher 2 00:00:04,519 --> 00:00:10,719 Speaker 1: and writer here at House to Works. Every week, I'm 3 00:00:10,720 --> 00:00:13,240 Speaker 1: bringing you three stories from our team about the weird 4 00:00:13,280 --> 00:00:17,599 Speaker 1: and wondrous developments we've seen in science, technology, and culture Today. 5 00:00:17,800 --> 00:00:21,200 Speaker 1: Wildlife is thriving in Chernobyl thirty years after the world's 6 00:00:21,280 --> 00:00:26,840 Speaker 1: largest nuclear accident and unrelated, tiny freshwater predators called hydra 7 00:00:27,000 --> 00:00:30,639 Speaker 1: don't have mouths, and now we know how they pull 8 00:00:30,720 --> 00:00:34,120 Speaker 1: open their own skin to swallow their prey. But First, 9 00:00:34,520 --> 00:00:37,240 Speaker 1: Senior writer Jonathan Strickland brings us a story about a 10 00:00:37,280 --> 00:00:39,320 Speaker 1: group of students who built a database of Game of 11 00:00:39,360 --> 00:00:43,199 Speaker 1: Thrones characters, over two thousand of them. The students then 12 00:00:43,280 --> 00:00:46,960 Speaker 1: created an algorithm to answer the ultimate fan question, who 13 00:00:47,040 --> 00:00:53,479 Speaker 1: will live and who will die? That was the question 14 00:00:53,520 --> 00:00:56,800 Speaker 1: a computer science class at the Technical University of Munich 15 00:00:56,880 --> 00:00:59,880 Speaker 1: had on their minds, so they wrote a computer pro 16 00:01:00,080 --> 00:01:03,480 Speaker 1: graham to answer it. Yes, a computer algorithm is making 17 00:01:03,520 --> 00:01:09,639 Speaker 1: predictions about who will be the next character to get stabbed, burned, shot, crushed, poisoned, eaten, 18 00:01:09,800 --> 00:01:13,240 Speaker 1: or otherwise eliminated. The class mind the book series for 19 00:01:13,280 --> 00:01:16,880 Speaker 1: as much data about characters as possible, relying heavily on 20 00:01:16,920 --> 00:01:20,920 Speaker 1: a Game of Thrones themed WICKI. They identified two thousand, 21 00:01:21,200 --> 00:01:25,920 Speaker 1: twenty eight characters and recorded data such as gender, age, 22 00:01:26,240 --> 00:01:30,040 Speaker 1: social standing, the house they belong to, their relationship status, 23 00:01:30,080 --> 00:01:33,720 Speaker 1: and more. In the process, the team made some observations. 24 00:01:33,840 --> 00:01:36,840 Speaker 1: For example, men outnumber women in the series by a 25 00:01:36,880 --> 00:01:39,800 Speaker 1: factor of two to one. Male characters are also more 26 00:01:39,880 --> 00:01:42,360 Speaker 1: likely to be members of the nobility and to meet 27 00:01:42,400 --> 00:01:45,280 Speaker 1: a violent end. The prime age for getting your life 28 00:01:45,319 --> 00:01:48,160 Speaker 1: cut short is between twenty one and forty. If you 29 00:01:48,200 --> 00:01:51,880 Speaker 1: make it to seventy, congratulations, you're probably just going to 30 00:01:51,960 --> 00:01:55,120 Speaker 1: die of old age. Social status is a wash. You 31 00:01:55,200 --> 00:01:57,600 Speaker 1: might be a king or a little street sweeper, but 32 00:01:57,720 --> 00:02:00,720 Speaker 1: sooner or later you'll dance with the reaper. The students 33 00:02:00,800 --> 00:02:03,920 Speaker 1: created a machine learning algorithm to study all these points 34 00:02:03,960 --> 00:02:07,120 Speaker 1: of data and note which characters have already died. The 35 00:02:07,160 --> 00:02:10,480 Speaker 1: algorithm then relied on this information to make predictions about 36 00:02:10,560 --> 00:02:13,959 Speaker 1: future deaths. And that brings us to the moment you've 37 00:02:13,960 --> 00:02:17,720 Speaker 1: all been waiting for. Who is living on borrowed time. 38 00:02:18,000 --> 00:02:21,040 Speaker 1: According to the algorithm, the character most likely to depart 39 00:02:21,160 --> 00:02:27,000 Speaker 1: next is tom and Barathian at nine seven percent. Bummer 40 00:02:27,919 --> 00:02:31,320 Speaker 1: Tommin is the young lad currently occupying the iron throne, 41 00:02:31,400 --> 00:02:34,840 Speaker 1: so he's definitely a prime target. But the kid didn't 42 00:02:34,880 --> 00:02:38,720 Speaker 1: do nothing to nobody Right behind tommin our Stannus barathian 43 00:02:38,800 --> 00:02:43,520 Speaker 1: At and the Mother of Dragons herself, the nearest targarian at. 44 00:02:44,840 --> 00:02:46,800 Speaker 1: As for who appears to be safe, that would be 45 00:02:46,880 --> 00:02:50,440 Speaker 1: Sansa Stark, with a tiny three percent chance of death. 46 00:02:50,560 --> 00:02:53,520 Speaker 1: Then again, the algorithm identified John Snow as a top 47 00:02:53,600 --> 00:02:56,960 Speaker 1: survivor with only an eleven percent chance of dying. But 48 00:02:57,160 --> 00:03:00,760 Speaker 1: remember it's a computer algorithm, not George are Our Martin 49 00:03:00,880 --> 00:03:04,280 Speaker 1: telling us these things. The program isn't full proof. In fact, 50 00:03:04,320 --> 00:03:08,400 Speaker 1: it has a forty precision rating for predicting dead characters. 51 00:03:08,440 --> 00:03:11,360 Speaker 1: That's a coin flip, people, though I guess those characters 52 00:03:11,400 --> 00:03:14,240 Speaker 1: could still die before the series is over. The algorithm 53 00:03:14,280 --> 00:03:16,880 Speaker 1: does do better at predicting who will live, with eighty 54 00:03:16,960 --> 00:03:25,600 Speaker 1: five percent precision. Now, Senior writer Robert Lamb explores a 55 00:03:25,680 --> 00:03:29,639 Speaker 1: study into the fascinating feeding mechanism of Hydra, the predator 56 00:03:29,800 --> 00:03:37,120 Speaker 1: without a mouth. The hydro genus gets its name from 57 00:03:37,120 --> 00:03:41,440 Speaker 1: the mythological monster that multi headed serpent with the frustrating 58 00:03:41,480 --> 00:03:44,480 Speaker 1: power to grow back multiple heads for each one you 59 00:03:44,560 --> 00:03:48,480 Speaker 1: lop off and Indeed, the real life hydra boast fascinating 60 00:03:48,520 --> 00:03:52,920 Speaker 1: regenerative powers as well, including the grimlins likability to reproduce 61 00:03:53,040 --> 00:03:56,400 Speaker 1: via a sexual budding. And then there's that whole wound mouth. 62 00:03:56,640 --> 00:04:00,440 Speaker 1: After incapacitating its prey with tentacled poison ball RBS, the 63 00:04:00,480 --> 00:04:03,680 Speaker 1: creature tears a page from the John Carpenter playbook and 64 00:04:03,920 --> 00:04:07,440 Speaker 1: rips its body open into a hungry, gaping mass, sometimes 65 00:04:07,520 --> 00:04:11,080 Speaker 1: wider than the creature's actual body, like some fever dream 66 00:04:11,160 --> 00:04:14,680 Speaker 1: horror glimpsed in a Flemish hell painted. The hydra gobbles 67 00:04:14,760 --> 00:04:17,520 Speaker 1: up its prey, and the wound mouth heals itself over 68 00:04:17,560 --> 00:04:21,200 Speaker 1: the doomed victim, reverting to a continuous sheet of tissue. Now, 69 00:04:21,200 --> 00:04:24,360 Speaker 1: the basic process here and the chemical triggers involved are 70 00:04:24,400 --> 00:04:26,960 Speaker 1: not new to science, but nobody ever figured out how 71 00:04:27,000 --> 00:04:30,640 Speaker 1: the wound mouth worked until now. In a study published 72 00:04:30,800 --> 00:04:34,200 Speaker 1: this month in the Biophysical Journal, researchers from the University 73 00:04:34,240 --> 00:04:38,719 Speaker 1: of California stare deep into the feeding hole of hydro vulgaris. 74 00:04:38,720 --> 00:04:43,680 Speaker 1: They engineered transgenic hydras with fluorescent proteins in their endodermal 75 00:04:43,760 --> 00:04:47,960 Speaker 1: and ectodermal cell layers. This created glowing skin layers to 76 00:04:48,040 --> 00:04:51,320 Speaker 1: illuminate the mouth opening mechanics. As it turns out, the 77 00:04:51,360 --> 00:04:54,520 Speaker 1: cells don't move around, they actually change their shape in 78 00:04:54,600 --> 00:04:58,479 Speaker 1: order to birth these wondrous mouths. Cell nuclei even appear 79 00:04:58,560 --> 00:05:02,240 Speaker 1: to deform in the process. Radially oriented fibers in the 80 00:05:02,240 --> 00:05:05,520 Speaker 1: tissue contract to stretch the cells apart, similar to the 81 00:05:05,560 --> 00:05:08,919 Speaker 1: muscular behavior in the iris of the human eye. More 82 00:05:09,000 --> 00:05:12,520 Speaker 1: revelations away the researches, as future studies revealed the precise 83 00:05:12,680 --> 00:05:15,640 Speaker 1: inner workings of the entire mouthing process of the hydra, 84 00:05:15,960 --> 00:05:20,080 Speaker 1: as well as perhaps the evolutionary reason for such a 85 00:05:20,160 --> 00:05:30,599 Speaker 1: unique feeding mechanism. Finally, this week, I wrote about the 86 00:05:30,640 --> 00:05:33,159 Speaker 1: diverse ecosystem that sprung up around the site of the 87 00:05:33,240 --> 00:05:37,279 Speaker 1: Chernobyl nuclear accident, and whether studying these animals could help 88 00:05:37,279 --> 00:05:45,920 Speaker 1: researchers discover how radiation affects us all. In April, and 89 00:05:46,000 --> 00:05:48,919 Speaker 1: explosion at the Chernobyl Nuclear power plant resulted in the 90 00:05:48,960 --> 00:05:52,599 Speaker 1: worst nuclear accident in history. The resulting fire lasted ten 91 00:05:52,720 --> 00:05:55,800 Speaker 1: days and at least an untold amount of radioactive materials 92 00:05:55,800 --> 00:05:59,840 Speaker 1: into the atmosphere to gradually fall out over the surrounding countryside, 93 00:06:00,040 --> 00:06:02,599 Speaker 1: an area stretching about a thousand square miles around the 94 00:06:02,600 --> 00:06:06,880 Speaker 1: disaster site is still designated the Chernobyl Exclusion Zone, unfit 95 00:06:06,920 --> 00:06:09,560 Speaker 1: for human habitation, but research over the years has shown 96 00:06:09,560 --> 00:06:12,560 Speaker 1: that in our absence, wildlife is thriving. A study from 97 00:06:12,920 --> 00:06:17,279 Speaker 1: found substantial animal tracks and as April, a University of 98 00:06:17,279 --> 00:06:21,039 Speaker 1: Georgia team has documented fourteen species of mammals using remote 99 00:06:21,040 --> 00:06:24,480 Speaker 1: station cameras, mostly wolves and raccoon dogs a ka tanuki. 100 00:06:24,560 --> 00:06:26,600 Speaker 1: The cameras were set up for a week each at 101 00:06:26,680 --> 00:06:29,200 Speaker 1: ninety four sites and bated with a fatty acid sent 102 00:06:29,279 --> 00:06:32,359 Speaker 1: to attract animals, particularly carnivores. The team was focused on 103 00:06:32,400 --> 00:06:34,680 Speaker 1: the top of the food chain because these critters received 104 00:06:34,680 --> 00:06:37,919 Speaker 1: the most radiation exposure directly from the air, soil, and water, 105 00:06:38,080 --> 00:06:41,560 Speaker 1: as all animals would, but also from the contamination accumulated 106 00:06:41,600 --> 00:06:44,720 Speaker 1: in their prey. The international science community is still trying 107 00:06:44,760 --> 00:06:47,320 Speaker 1: to figure out how much damage all the radiation from 108 00:06:47,400 --> 00:06:51,720 Speaker 1: Chernobyl has caused and is still causing. And okay, radiation 109 00:06:51,800 --> 00:06:55,320 Speaker 1: is all around us all of us. Light, heat, radio waves, 110 00:06:55,320 --> 00:06:59,520 Speaker 1: and microwaves are non ionizing radiation. They carry enough energy 111 00:06:59,560 --> 00:07:01,919 Speaker 1: to excite atoms, but not to break them apart. But 112 00:07:02,120 --> 00:07:05,440 Speaker 1: X rays, gamma rays, and admissions from radioactive materials are 113 00:07:05,560 --> 00:07:09,720 Speaker 1: ionizing radiation. They can bust electrons right out of their atoms. 114 00:07:09,840 --> 00:07:13,280 Speaker 1: On a cellular level, that's what scientists refer to as bad. 115 00:07:13,480 --> 00:07:16,400 Speaker 1: Ionizing radiation can break apart the genes that tell your 116 00:07:16,440 --> 00:07:19,120 Speaker 1: cells and systems how to function, causing all kinds of 117 00:07:19,120 --> 00:07:22,720 Speaker 1: health problems. But there's no consensus on how much damage 118 00:07:22,720 --> 00:07:25,360 Speaker 1: different levels of exposure can cause. In the long run, 119 00:07:25,600 --> 00:07:27,520 Speaker 1: there just isn't enough data for us to draw from. 120 00:07:27,560 --> 00:07:30,119 Speaker 1: All we can do is watch the populations affected, human 121 00:07:30,160 --> 00:07:32,720 Speaker 1: and wildlife alike and wait. The team from u g 122 00:07:32,840 --> 00:07:37,360 Speaker 1: A says that higher levels of contamination didn't suppress wildlife populations, 123 00:07:37,520 --> 00:07:40,600 Speaker 1: rather the animals they observed when wherever food and water 124 00:07:40,640 --> 00:07:43,400 Speaker 1: could be found. Regardless, they're hoping that further studies will 125 00:07:43,440 --> 00:07:50,760 Speaker 1: measure the animals health and survival rates. And that's all 126 00:07:50,800 --> 00:07:53,640 Speaker 1: for this week. Thanks so much for tuning in. Subscribe 127 00:07:53,680 --> 00:07:56,600 Speaker 1: now for more of the latest and strangest science news, 128 00:07:56,800 --> 00:07:59,200 Speaker 1: and send us your ideas for new stories, plus any 129 00:07:59,240 --> 00:08:02,360 Speaker 1: feedback or hey reading recommendations that you might have. Shoot 130 00:08:02,440 --> 00:08:05,000 Speaker 1: us an email at now Podcast at how stuff works 131 00:08:05,320 --> 00:08:15,080 Speaker 1: dot com.