WEBVTT - Tech News: Time Crystals 0:00:04.400 --> 0:00:07.800 Welcome to tech Stuff, a production from I Heart Radio. 0:00:12.039 --> 0:00:15.120 Hey there, and welcome to tech Stuff. I'm your host, 0:00:15.320 --> 0:00:18.440 Jonathan Strickland. I'm an executive producer with I Heart Radio 0:00:18.600 --> 0:00:20.959 and a love of all things tech. And this is 0:00:21.000 --> 0:00:25.320 the tech news for Thursday, August five, twenty twenty one. 0:00:26.160 --> 0:00:30.400 We've got a lot of it, including some really crazy 0:00:30.480 --> 0:00:34.680 quantum physics news that I am going to do my 0:00:34.760 --> 0:00:39.199 best to try and impart unto you, dear listener. But 0:00:39.320 --> 0:00:43.959 let's get started with some Facebook stories, because I just 0:00:44.159 --> 0:00:47.880 can't seem to go an episode without talking about Facebook. 0:00:48.400 --> 0:00:51.320 So one of these stories is likely not going to 0:00:51.360 --> 0:00:53.760 be a surprise to any of you listeners at all. 0:00:54.120 --> 0:00:58.520 And that's the fact that many law enforcement agencies are 0:00:58.760 --> 0:01:02.320 crawling on social networking sites to collect images of people 0:01:02.640 --> 0:01:06.440 and use those images for the purposes of facial recognition systems. 0:01:07.319 --> 0:01:11.440 The m Live site, which covers news in Michigan here 0:01:11.480 --> 0:01:14.679 in the United States, published an article this morning about this, 0:01:15.040 --> 0:01:18.880 quoting Eric Goldman of the High Tech Law Institute at 0:01:18.920 --> 0:01:22.880 Santa Clara University, and Goldman said that if you have 0:01:23.400 --> 0:01:27.800 a public social networking accounts, such as a Facebook account 0:01:27.880 --> 0:01:31.520 that posts publicly, it stands to reason that a lot 0:01:31.520 --> 0:01:36.120 of different people and organizations have grabbed your photos at 0:01:36.160 --> 0:01:40.600 some point because they are all publicly available. Publicly available 0:01:40.640 --> 0:01:44.440 information is you know, public, so anyone can see it. 0:01:44.880 --> 0:01:47.840 And we've seen numerous groups use software to crawl through 0:01:48.120 --> 0:01:51.480 enormous platforms to pull down mountains of data for the 0:01:51.520 --> 0:01:55.600 purposes of identification and analysis, and those are acts that 0:01:55.640 --> 0:01:59.760 platforms frequently resist for a variety of reasons, like there's 0:02:00.480 --> 0:02:03.640 an AI platform that has done this that has gotten 0:02:03.640 --> 0:02:06.320 in trouble for it in the past because of that. 0:02:07.120 --> 0:02:09.360 I should also add that while I say if you 0:02:09.400 --> 0:02:12.560 post publicly, that's not even the full picture obviously, right, 0:02:12.600 --> 0:02:15.960 Like let's say you don't even have a Facebook profile, 0:02:16.960 --> 0:02:20.079 you could be in photos that your friends take, right, 0:02:20.480 --> 0:02:25.000 and those photos might be on their their profiles, and 0:02:25.040 --> 0:02:28.839 they might be public which means that those photos can 0:02:28.880 --> 0:02:34.280 be captured and used for facial recognition purposes. So you 0:02:34.280 --> 0:02:36.440 don't even have to be involved in this. You have 0:02:36.880 --> 0:02:41.040 no say at that point if someone is posting your 0:02:41.040 --> 0:02:44.760 photo publicly, and it's fair game for a lot of 0:02:44.760 --> 0:02:48.079 these organizations, at least as far as they're concerned. Now, 0:02:48.120 --> 0:02:52.160 we've also seen that many facial recognition systems have an 0:02:52.160 --> 0:02:54.720 element of bias in them, with a tendency to be 0:02:54.800 --> 0:02:58.440 less accurate for people of color. And that might not 0:02:58.600 --> 0:03:01.520 sound you know, terrible, like, oh, it's less accurate, but 0:03:01.760 --> 0:03:04.720 then you figure that a less accurate system means that 0:03:04.880 --> 0:03:09.680 sometimes a person could be inaccurately identified as, say, a 0:03:09.840 --> 0:03:12.880 person of interest in a legal case, and then you 0:03:12.919 --> 0:03:15.679 can really see how it does become a problem. Such 0:03:15.720 --> 0:03:19.880 a misidentification can lead in a best case scenario, it 0:03:19.880 --> 0:03:24.440 can lead to an inconvenience. That's the best case scenario. 0:03:24.880 --> 0:03:29.600 Worst case scenario ends up being incredibly harmful and disruptive. 0:03:30.120 --> 0:03:33.520 I mean that could even include wrongful arrests which have 0:03:33.639 --> 0:03:39.240 happened as a result of facial recognition software mistakenly identifying 0:03:39.320 --> 0:03:43.320 someone as someone else. And the fact that this disproportionately 0:03:43.400 --> 0:03:47.000 harms specific communities, like communities of color. That's a real 0:03:47.080 --> 0:03:50.839 issue now, as the Peace and m Life points out, 0:03:51.240 --> 0:03:54.200 there have been documented cases of police action based off 0:03:54.280 --> 0:03:59.480 incorrect facial recognition analysis already in Michigan. There's a growing 0:03:59.560 --> 0:04:02.960 resist to the use of this technology. There's been a 0:04:02.960 --> 0:04:06.360 lot more pressure on local and federal governments to outlaw 0:04:06.360 --> 0:04:09.600 it's used there's also been pressure on companies to not 0:04:10.080 --> 0:04:14.720 make that technology available to law enforcement agencies, and this 0:04:14.760 --> 0:04:17.640 ties in with several other stories that are unfolding around 0:04:17.640 --> 0:04:21.160 the world, many of which tap into this issue of 0:04:21.520 --> 0:04:25.159 wanting to lean on technology in order to solve a 0:04:25.240 --> 0:04:29.719 particularly difficult social problem. But the issue is that in 0:04:29.760 --> 0:04:35.440 some cases the technology isn't necessarily reliable enough to depend upon, 0:04:35.960 --> 0:04:38.279 So I may have to do a full episode dedicated 0:04:38.320 --> 0:04:40.800 to that concept because there are a lot of different 0:04:40.800 --> 0:04:45.279 factors and play here. Right, there's a genuine desire to 0:04:45.560 --> 0:04:50.120 right a social wrong or to prevent a particular social 0:04:50.160 --> 0:04:54.599 problem from happening. There's a perception that the technology could 0:04:54.640 --> 0:04:57.400 be the solution to this, and then there's the reality 0:04:57.440 --> 0:05:01.000 of technology that may not be reliable enough for us 0:05:01.040 --> 0:05:04.560 to really leverage it that way, or else we have 0:05:04.760 --> 0:05:09.040 these exceptions pop up that then affect innocent people in 0:05:09.200 --> 0:05:13.640 harmful ways. Moving on, The Guardian published an article about 0:05:13.760 --> 0:05:17.200 how the fossil fuel industry was able to leverage Facebook 0:05:17.279 --> 0:05:21.080 to push climate misinformation. The article cites a group called 0:05:21.200 --> 0:05:25.760 influence Map, which is based in London, which uncovered this issue. 0:05:26.040 --> 0:05:29.279 Influence Maps said that there was an increase in advertising 0:05:29.320 --> 0:05:33.640 by fossil fuel companies like ex On Mobile in during 0:05:33.680 --> 0:05:38.400 the election season, with the campaign specifically creating a misinformation 0:05:38.480 --> 0:05:43.000 narrative to shape the conversation around climate change and fossil 0:05:43.040 --> 0:05:46.599 fuel policies that were meant to address that problem. Further 0:05:47.279 --> 0:05:51.960 influence MAP alleges that this campaign violated Facebook's own policies 0:05:51.960 --> 0:05:55.960 with regard to false advertising, and yet Facebook did not 0:05:56.080 --> 0:05:59.880 remove or label these posts as being misleading, which allowed 0:06:00.200 --> 0:06:03.920 those views to really take hold within certain communities and Facebook. 0:06:04.400 --> 0:06:07.400 The organization also found something that a lot of people 0:06:07.440 --> 0:06:11.400 have observed over the years that the messaging around conservation 0:06:11.960 --> 0:06:16.080 and you know, having an impact on climate change really 0:06:16.320 --> 0:06:20.200 was shifting the focus to the individual rather than to 0:06:20.560 --> 0:06:24.400 companies and industries. In other words, putting the burden of 0:06:24.440 --> 0:06:29.200 addressing climate change solely on individual people rather than both 0:06:29.320 --> 0:06:33.280 on people and on corporations. So this is kind of 0:06:33.279 --> 0:06:35.479 the message of turn off the lights after you leave 0:06:35.520 --> 0:06:38.760 a room, like or only you can prevent forest fires 0:06:38.839 --> 0:06:43.320 kind of approach, rather than place regulations on or otherwise 0:06:43.360 --> 0:06:46.640 tax companies that are consuming vast amounts of electricity, right 0:06:47.000 --> 0:06:50.240 Like that would be the other message. Uh, we frequently 0:06:50.360 --> 0:06:54.760 see this kind of personalized approach to messaging, and yeah, 0:06:54.800 --> 0:06:56.599 we all do have a part to play here. But 0:06:56.680 --> 0:07:00.919 when that messaging completely ignores how industry contribut mutes to 0:07:01.240 --> 0:07:05.160 a large extent, to climate change issues, then it becomes 0:07:05.200 --> 0:07:11.200 misinformation because it's purposefully omitting important details. Right. If you 0:07:11.280 --> 0:07:15.760 don't say, hey, the transportation industry ends up being a 0:07:15.800 --> 0:07:21.640 massive contributor to carbon emissions and you're instead saying turn 0:07:21.680 --> 0:07:24.160 off the lights and don't run stuff when you're not 0:07:24.320 --> 0:07:29.880 using it, then you allow the major contributing factor to 0:07:30.000 --> 0:07:33.520 go on unhindered, and it then the problem worsens as 0:07:33.560 --> 0:07:37.160 a result. So Facebook has received some pretty critical opposition 0:07:37.200 --> 0:07:39.960 to this trend already and will likely face even more 0:07:40.000 --> 0:07:42.840 scrutiny in the future. And it should. I mean, it's 0:07:42.880 --> 0:07:47.320 it's cashing those checks, right, The advertising checks are coming 0:07:47.320 --> 0:07:50.760 in and Facebook is cashing them, So it does definitely 0:07:50.800 --> 0:07:54.760 play a part. Oh and the CNN Business story about 0:07:54.920 --> 0:07:58.720 influence maps look into Facebook reveals that these fossil fuel 0:07:58.800 --> 0:08:02.720 ads showed up more than four hundred thirty one million 0:08:02.840 --> 0:08:06.640 times for US users, in which is not a bad 0:08:06.680 --> 0:08:10.840 return on investment for nine point six million dollars that 0:08:11.000 --> 0:08:14.840 was the given amount for that those collective ad campaigns 0:08:15.280 --> 0:08:18.920 now think about this a super Bowl ad. A commercial 0:08:19.000 --> 0:08:22.440 during the Super Bowl costs around five million dollars for 0:08:22.480 --> 0:08:25.400 a thirty second spot, and the last Super Bowl had 0:08:25.440 --> 0:08:28.800 around ninety s point four million people watching it. So 0:08:29.040 --> 0:08:33.160 for a little less than twice the cash, you can 0:08:33.200 --> 0:08:37.040 get four times the number of views by going through Facebook, 0:08:37.040 --> 0:08:40.520 which is not a bad deal right. Meanwhile, the petroleum 0:08:40.559 --> 0:08:44.120 industry companies have spoken out against the Influence Map study, 0:08:44.400 --> 0:08:46.960 claiming that these companies are spending a great amount of 0:08:46.960 --> 0:08:52.760 money on limiting carbon emissions and using carbon sequestration strategies 0:08:52.760 --> 0:08:55.560 and such, which you know, could be totally true, but 0:08:55.760 --> 0:08:58.760 doesn't change the fact that these companies actively try to 0:08:58.960 --> 0:09:03.679 undermine climate action. Heck, and x On lobbyists essentially spilled 0:09:03.679 --> 0:09:07.280 the beans on how they do that on video, which 0:09:07.360 --> 0:09:10.880 then prompted the CEO of x On to apologize for it. 0:09:11.320 --> 0:09:16.360 I'm sure they were sorry that they were caught. And 0:09:16.520 --> 0:09:19.400 we are not yet done with Facebook. We've got one 0:09:19.400 --> 0:09:23.040 more Facebook story here. Bloomberg reported yesterday that Facebook has 0:09:23.080 --> 0:09:26.720 disabled the personal Facebook accounts of some researchers from New 0:09:26.800 --> 0:09:32.079 York University who have been studying political ads on that platform. 0:09:32.080 --> 0:09:35.520 Facebook says this group has been using tools like automated 0:09:35.559 --> 0:09:40.000 tools to scrape Facebook of data, which is going against 0:09:40.040 --> 0:09:42.800 Facebook's policy. This is kind of alluding to what I 0:09:42.840 --> 0:09:47.400 was mentioning earlier about facial recognition technologies. You know, Facebook 0:09:47.720 --> 0:09:51.160 isn't supposed to allow that to happen. Now, the research 0:09:51.200 --> 0:09:55.360 team saw their project pretty much hamstrung by Facebook. The 0:09:55.400 --> 0:09:59.280 company deactivated various apps and pages that were associated with 0:09:59.320 --> 0:10:03.480 this research project, in addition to deactivating those personal pages 0:10:03.559 --> 0:10:06.959 of the actual team involved. Now, on the face of it, 0:10:07.760 --> 0:10:10.120 my danger starts to get up on behalf of the 0:10:10.200 --> 0:10:13.520 research team. However, I have to admit this story is 0:10:13.520 --> 0:10:17.679 actually more complicated than that. See back in two thousand nineteen, 0:10:18.080 --> 0:10:22.120 the Federal Trade Commission or FTC here in the United 0:10:22.160 --> 0:10:26.840 States hit Facebook with a fine of five billion dollars 0:10:26.840 --> 0:10:30.240 as billion with a B, which is a princely sum indeed, 0:10:30.440 --> 0:10:32.880 And the reason for that fine was that the FTC 0:10:33.080 --> 0:10:36.240 found that Facebook had not done enough to protect users 0:10:36.480 --> 0:10:40.679 from developers who were relying on Facebook to collect personal 0:10:40.760 --> 0:10:46.000 information on a large scale, think of things like Cambridge Analytica. 0:10:46.080 --> 0:10:49.840 So Facebook has kind of an obligation to push back 0:10:49.880 --> 0:10:53.080 against groups that are using automated means to collect data 0:10:53.200 --> 0:10:56.400 off the platform, even if that group has a seemingly 0:10:56.480 --> 0:10:59.920 good reason for doing so. Facebook requires groups to see 0:11:00.040 --> 0:11:05.120 out permission from the company before using those kinds of tools, 0:11:05.200 --> 0:11:08.880 and this group, you could argue, failed to secure permission, 0:11:09.360 --> 0:11:12.400 and in fact, Facebook had reached out and given the 0:11:12.440 --> 0:11:15.560 group a couple of warnings in the past. Now that 0:11:15.640 --> 0:11:18.960 being said, Facebook is also under a great deal of 0:11:19.000 --> 0:11:22.360 scrutiny for how it serves up political ads and how 0:11:22.360 --> 0:11:25.720 it targets people and how it chooses who gets to 0:11:25.760 --> 0:11:28.920 see what. So, as I said, this is a really 0:11:28.960 --> 0:11:33.000 complicated issue. The study could call more attention to problems 0:11:33.040 --> 0:11:36.680 with Facebook's methods of accepting and serving ads to users, 0:11:37.080 --> 0:11:41.200 which the company would likely want to suppress. It doesn't 0:11:41.240 --> 0:11:45.000 want those problems to be brought into, you know, harsher light. 0:11:45.520 --> 0:11:49.000 But then Facebook also has this obligation under the FTC, 0:11:49.600 --> 0:11:52.320 and that at least gives the company a fairly compelling 0:11:52.360 --> 0:11:56.120 reason to push back. Though one could imagine that this 0:11:57.480 --> 0:12:00.480 issue could be solved if Facebook and the research group 0:12:01.040 --> 0:12:05.280 could just get that permission to the researchers, but I 0:12:05.440 --> 0:12:08.360 really doubt that Facebook is eager to do that. In 0:12:08.400 --> 0:12:13.200 this particular case, we have more stories to cover. But 0:12:13.280 --> 0:12:23.520 before we do that, let's take a quick break. Reuters 0:12:23.559 --> 0:12:26.880 reports that are ransomware attacked targeting more than one thousand 0:12:26.960 --> 0:12:30.360 organizations last month might be just the beginning of a 0:12:30.440 --> 0:12:35.559 trend of attacks on service providers. The attack leveraged vulnerabilities 0:12:35.600 --> 0:12:38.719 in a management software product from a company called CASA, 0:12:39.320 --> 0:12:44.160 which affected more than fifty managed service provider organizations or 0:12:44.440 --> 0:12:48.520 m sps. Now, an MSP is a company that other 0:12:48.600 --> 0:12:53.280 companies rely upon to outsource services of some sort, like 0:12:53.400 --> 0:12:57.480 they could be human resources services or I T services, 0:12:57.920 --> 0:13:01.480 And that means that the MSP represents a really juicy 0:13:01.600 --> 0:13:06.880 target for hackers because you could attack a specific company. 0:13:06.960 --> 0:13:09.920 Let's say that you wanted to target I don't know, 0:13:10.040 --> 0:13:14.520 like like, uh, Intel, you want to target Intel, and 0:13:14.520 --> 0:13:17.640 you want to compromise one of Intel's systems. Well, that's 0:13:17.640 --> 0:13:20.640 a single target. But let's say you go after an 0:13:20.800 --> 0:13:25.200 MSP which could service dozens or hundreds of clients, and 0:13:25.280 --> 0:13:28.199 Intel happens to be one of them, but it also 0:13:28.240 --> 0:13:31.440 services all these others. Well, you can disrupt all of 0:13:31.480 --> 0:13:36.320 those companies by targeting the MSP specifically, right if they 0:13:36.320 --> 0:13:38.600 are providing a service to all these companies and you 0:13:38.679 --> 0:13:42.920 interrupt it, you've affected not just one organization, but potentially 0:13:43.120 --> 0:13:47.720 hundreds of them. This puts immense pressure on the MSP 0:13:47.920 --> 0:13:51.720 companies to solve the problem quickly, because if all of 0:13:51.760 --> 0:13:54.880 its customers are depending upon those services, and you've got 0:13:54.960 --> 0:13:58.440 like a hundred important clients, you might be answering calls 0:13:58.640 --> 0:14:01.320 all day long because you can't get their widgets to 0:14:01.320 --> 0:14:04.040 work because of a problem with your product. You have 0:14:04.360 --> 0:14:07.959 a high motivation to fix the issue that might include 0:14:08.280 --> 0:14:13.000 paying off an exorbitant ransom or worse. The hackers might 0:14:13.080 --> 0:14:16.000 use the MSPs as a launch point and find a 0:14:16.040 --> 0:14:19.240 way to use an MSP product to have a supply 0:14:19.320 --> 0:14:23.920 chain attack and thus infect the MSPs customers as well. 0:14:24.000 --> 0:14:27.280 So the MSP becomes the entry point and then all 0:14:27.320 --> 0:14:30.440 of its customers become targets. That multiplies the number of 0:14:30.440 --> 0:14:34.120 targets that the hackers are able to to actually compromise, 0:14:34.560 --> 0:14:38.040 and potentially multiplies the number of ransoms that they could 0:14:38.040 --> 0:14:42.040 receive as a result. Reuters states that hacker groups are 0:14:42.200 --> 0:14:46.480 feverishly researching ways to target MSPs for just that reason, 0:14:46.800 --> 0:14:49.440 which suggests we should be on the lookout for similar 0:14:49.480 --> 0:14:51.760 attacks in the future, and it should serve as a 0:14:51.800 --> 0:14:56.960 warning to all organizations MSP and otherwise to be extra 0:14:57.080 --> 0:15:02.160 cautious with security. Moving on on, the state of Massachusetts 0:15:02.320 --> 0:15:05.800 is seeing a group called the Massachusetts Coalition for Independent 0:15:05.840 --> 0:15:09.720 Work Lobby to have a special ballot introduced that would 0:15:09.800 --> 0:15:13.720 create some special exemptions for companies like Uber and Lift 0:15:14.120 --> 0:15:18.040 to be largely excluded from the restrictions and regulations of 0:15:18.080 --> 0:15:22.200 the state's labor laws. And you will be shocked, shocked, 0:15:22.600 --> 0:15:25.560 I tell you to learn that Uber and Lift run 0:15:26.040 --> 0:15:29.560 this special interest group whose name makes it sound like 0:15:29.640 --> 0:15:34.080 it's a worker advocacy organization. The companies have actually done 0:15:34.120 --> 0:15:37.920 this before. In fact, they did so successfully in California 0:15:38.040 --> 0:15:42.160 by introducing a measure called Proposition twenty two, which effectively 0:15:42.360 --> 0:15:46.680 excluded them and several other gig economy companies from having 0:15:46.720 --> 0:15:49.920 to abide by the laws that govern other businesses in 0:15:49.960 --> 0:15:53.000 the state of California. And it looks like the companies 0:15:53.040 --> 0:15:56.480 are going to spend a lot of money in Massachusetts 0:15:56.520 --> 0:15:59.720 to fight the same sort of battle there. Whether that 0:15:59.800 --> 0:16:03.080 were us are not remains to be seen, because generally speaking, 0:16:03.120 --> 0:16:06.520 there's a growing awareness regarding the company's efforts to avoid 0:16:06.560 --> 0:16:10.720 things like acknowledging the drivers are in fact employees as 0:16:10.720 --> 0:16:14.440 opposed to you know, contracted freelancers and so on. But 0:16:14.520 --> 0:16:16.600 we'll have to watch to see how this plays out, 0:16:16.640 --> 0:16:18.760 because we have to keep in mind these companies already 0:16:18.760 --> 0:16:22.280 have a track record for selling a political message effectively 0:16:22.480 --> 0:16:26.480 to the public, and they might do it again. And 0:16:26.560 --> 0:16:29.960 now for the latest of if misogyny in the workplace 0:16:30.040 --> 0:16:34.880 is such a problem, why don't more women come forward? Department, Well, 0:16:34.920 --> 0:16:38.200 to you, I submit the story of Ashley Giovic, a 0:16:38.320 --> 0:16:42.560 program manager at Apple's engineering department. According to the verge, 0:16:42.680 --> 0:16:46.880 Yovic found herself put on administrative leave indefinitely. In fact, 0:16:46.960 --> 0:16:50.240 she actually requested that, but we'll get to that. After 0:16:50.320 --> 0:16:52.720 she tweeted about how she found Apple to be a 0:16:52.760 --> 0:16:57.080 hostile work environment and one that tolerates sexism within the workplace. 0:16:57.320 --> 0:17:01.680 She had already gone through internal system ms numerous times 0:17:01.760 --> 0:17:05.159 in order to address this, so tweeting this out was 0:17:05.240 --> 0:17:08.480 kind of like, you know, the fact that she wasn't 0:17:08.520 --> 0:17:13.080 seeing progress being made within the company. And when you 0:17:13.160 --> 0:17:18.720 see a company essentially sideline someone after they come forward with, 0:17:19.960 --> 0:17:24.040 you know, these these issues, that's really one of the 0:17:24.080 --> 0:17:28.280 reasons why more women don't come forward because they've they've 0:17:28.280 --> 0:17:31.000 been intimidated. They've seen time and time again that the 0:17:31.040 --> 0:17:35.560 people who speak out end up getting sidelined. Gil Vic 0:17:35.600 --> 0:17:40.760 tweeted quote so following raising concerns to hashtag Apple about 0:17:40.880 --> 0:17:47.480 hashtag sexism, hashtag hostile work environment, and hashtag unsafe work conditions. 0:17:47.920 --> 0:17:52.800 I'm now on indefinite paid administrative leave per hashtag Apple 0:17:52.880 --> 0:17:57.359 employee relations while they investigate my concerns. This seems to 0:17:57.400 --> 0:18:02.400 include me not using Apple's internal back end quote. Giovic 0:18:02.480 --> 0:18:05.320 also said that the company offered to provide her medical 0:18:05.440 --> 0:18:09.640 leave rather than address the underlying issues, which is kind 0:18:09.720 --> 0:18:13.879 of like saying, hey, I know Pete in accounting keeps 0:18:13.960 --> 0:18:16.600 stabbing you, but we've got these band aids so you 0:18:16.600 --> 0:18:19.200 can patch yourself up every time it happens, rather than, 0:18:19.280 --> 0:18:23.960 you know, going to arrest Pete from accounting. Apple is, 0:18:24.040 --> 0:18:28.160 of course, just one of dozens of companies currently being 0:18:28.200 --> 0:18:32.199 called to reckon with its culture, and once again, it 0:18:32.680 --> 0:18:35.359 it's stories like these that feed into that feeling that 0:18:35.440 --> 0:18:39.080 human resources isn't there to protect the humans who work 0:18:39.119 --> 0:18:43.359 for the company, but rather to protect the company. Let 0:18:43.480 --> 0:18:47.000 us now segue over to a moment i'd like to 0:18:47.040 --> 0:18:50.040 call we live in the future, which can sometimes be 0:18:50.119 --> 0:18:55.200 on inspiring and sometimes terrifying. We'll start with something terrifying. 0:18:55.640 --> 0:18:59.040 Apparently the Pentagon is making use of an AI application 0:18:59.080 --> 0:19:02.640 to analyze real time information in an effort to predict 0:19:02.920 --> 0:19:06.760 what the enemy will do next, with the Pentagon claiming 0:19:06.920 --> 0:19:10.040 that the AI can let them see quote days in 0:19:10.119 --> 0:19:13.399 advance end quote. As a result, makes me think of 0:19:13.400 --> 0:19:17.680 the precogs in Minority Report, except instead of being put 0:19:17.720 --> 0:19:22.080 to use in law enforcement to predict crime, they are 0:19:22.320 --> 0:19:25.440 being put in military use to predict what the enemy 0:19:25.560 --> 0:19:29.840 is going to do in any given situation. And it 0:19:29.880 --> 0:19:32.439 also makes me think of Grand Admiral Thrown in the 0:19:32.480 --> 0:19:35.720 Star Wars Extended Universe. He could look at a culture's 0:19:35.800 --> 0:19:38.720 work of art and figure out what that culture how 0:19:38.880 --> 0:19:41.600 they would react, you know, to any given situation, which 0:19:41.640 --> 0:19:45.040 is pretty far fetched. But using AI to analyze what's 0:19:45.080 --> 0:19:47.240 going on within a specific region and then make a 0:19:47.280 --> 0:19:51.760 strategic decision seems like it's a little bit less fantastical, right. 0:19:51.800 --> 0:19:54.320 I Mean, you could argue that this is just an 0:19:54.359 --> 0:19:57.960 extremely much more complicated version of a war game, like 0:19:58.000 --> 0:20:01.440 a simulated war game. The AI is part of the 0:20:01.520 --> 0:20:08.200 Global Information Dominance Experiment or g I d E. And boy, 0:20:08.280 --> 0:20:11.919 that name is kind of scary. Obviously. The goal is 0:20:11.960 --> 0:20:15.000 to predict what an enemy might do in order so 0:20:15.040 --> 0:20:19.680 that we could prepare against that very move, potentially protecting 0:20:19.720 --> 0:20:22.879 the lives of soldiers and civilians in the process and 0:20:22.960 --> 0:20:26.520 maximizing the effectiveness of an armed presence in any given region. 0:20:26.720 --> 0:20:30.720 We might dissuade someone from doing something. For example, if 0:20:30.760 --> 0:20:37.080 the analysis says, hey, based upon historical data, such and 0:20:37.160 --> 0:20:40.679 such country is likely to launch a submarine from this 0:20:40.840 --> 0:20:44.320 port in the next two days, if you build up 0:20:44.480 --> 0:20:49.400 a show of force around that area, then you might 0:20:49.520 --> 0:20:53.720 dissuade them from doing that. That kind of thing. Well, 0:20:54.359 --> 0:20:57.439 the Pentagon points out that this is work that people 0:20:57.480 --> 0:21:01.679 have actually been doing for years by analyzing data in 0:21:01.720 --> 0:21:05.680 search of patterns and trying to make predictions. But obviously 0:21:06.040 --> 0:21:08.639 it took a long time in the past, so we 0:21:08.720 --> 0:21:10.800 are now in an era where we can collect and 0:21:11.000 --> 0:21:15.160 more importantly, analyze mountains of information rapidly, so we can 0:21:15.160 --> 0:21:18.159 find those patterns much more quickly and then act on 0:21:18.160 --> 0:21:22.520 that information. Of course, this requires a very careful design 0:21:22.600 --> 0:21:26.399 of the algorithm, and it's always possible for an algorithm 0:21:26.440 --> 0:21:30.800 to mistake noise for a meaningful pattern. But that's why 0:21:30.840 --> 0:21:34.400 this initiative is in a testing phase, and it concerns 0:21:34.560 --> 0:21:37.919 some critics of artificial intelligence in use with, you know, 0:21:37.960 --> 0:21:41.359 military applications, because they see it as a potential starting 0:21:41.400 --> 0:21:44.399 point for a slippery slope in which AI ends up 0:21:44.520 --> 0:21:48.200 not just pointing out possible enemy decisions as a result 0:21:48.200 --> 0:21:50.720 of what's going on in the world, but then going 0:21:50.800 --> 0:21:54.560 on to potentially making life or death decisions of its own, 0:21:55.080 --> 0:21:57.520 and then potentially even getting to the point where it 0:21:57.760 --> 0:22:02.399 enacts those decisions without human oversight, because obviously human oversight 0:22:02.480 --> 0:22:06.359 slows things down. So if you have really powerful AI 0:22:06.520 --> 0:22:10.080 making all the decisions, and you have at least assumed 0:22:10.080 --> 0:22:13.440 that the AI is infallible, you rely on the AI. 0:22:14.119 --> 0:22:17.800 We're always off from that sort of dystopian future, but 0:22:17.920 --> 0:22:21.520 being wary of that eventuality is probably not a bad idea. 0:22:21.720 --> 0:22:26.800 So while you could argue that it's premature to worry 0:22:26.800 --> 0:22:29.680 about that kind of stuff, the counter to that is, well, 0:22:29.760 --> 0:22:32.159 if we don't bring it, bring it up now, we 0:22:32.200 --> 0:22:36.439 could see this evolved slowly enough where it happens before 0:22:36.480 --> 0:22:38.720 we were even aware of it. I've got a couple 0:22:38.800 --> 0:22:40.560 more stories to tell, but before we get to that, 0:22:40.680 --> 0:22:52.639 let's take another break. Okay, it is time, my friends, 0:22:52.760 --> 0:22:55.840 and by time, I mean it's time to talk about 0:22:56.119 --> 0:23:01.320 time crystals, which full disclosure, I did not even know 0:23:01.520 --> 0:23:06.320 we're a thing before this morning. Apparently, time crystals are 0:23:06.400 --> 0:23:09.320 a phase of matter, kind of like you know how 0:23:09.440 --> 0:23:13.800 gas is a phase of matter, solids, liquids, plasmas, and 0:23:13.840 --> 0:23:17.240 that the most likely context we would observe such a 0:23:17.280 --> 0:23:20.600 thing as a time crystal is on the quantum level, 0:23:21.160 --> 0:23:25.040 so you know, they're they're they're very small, like smaller 0:23:25.040 --> 0:23:29.479 than Natman small. Some Google scientists have written a paper 0:23:29.720 --> 0:23:32.240 that still has to go through the pure review process, 0:23:32.359 --> 0:23:36.040 so we can't just assume that they got everything right. 0:23:36.480 --> 0:23:39.400 But that paper claims that they have found a way 0:23:39.440 --> 0:23:43.919 to use quantum computers to study time crystals, which is 0:23:44.720 --> 0:23:47.760 pretty much the most science fiction sentence I think I 0:23:47.760 --> 0:23:50.840 could say today. But let's break this down a bit, 0:23:50.880 --> 0:23:54.119 because that's just words salad, right unless we understand what 0:23:54.160 --> 0:23:59.399 we're talking about. Quantum computers are very complicated machines that 0:23:59.640 --> 0:24:05.240 only somewhat resemble classic computers, and they are very delicate 0:24:05.359 --> 0:24:09.560 as well. They have this tendency toward decoherence, which means 0:24:09.640 --> 0:24:13.480 that your quantum computer kind of breaks down and becomes 0:24:13.520 --> 0:24:17.320 a very low powered classical computer as a result. So 0:24:17.359 --> 0:24:21.560 with a traditional computer, we talk about information in terms 0:24:21.720 --> 0:24:25.479 of the basic unit of a bit or a binary digit, 0:24:25.920 --> 0:24:30.000 which can either be a zero or a one. Quantum 0:24:30.000 --> 0:24:34.080 computers are different. They have a cubit as the basic 0:24:34.200 --> 0:24:37.960 unit of information, and a cubit or a quantum bit 0:24:38.600 --> 0:24:41.959 can kind of sort of be both a zero and 0:24:42.400 --> 0:24:47.080 a one at the same time, and every value in between. Technically, 0:24:47.119 --> 0:24:52.840 this is a thing called superposition. Uh. There's also the 0:24:52.880 --> 0:24:55.760 issue of entanglement, but we're not going to get into 0:24:55.800 --> 0:24:58.320 all that. That would require a full episode by itself. 0:24:59.400 --> 0:25:01.840 But the important thing with quantum computers is that with 0:25:02.080 --> 0:25:06.840 the proper design of algorithms, that is, processes that are 0:25:06.960 --> 0:25:12.600 specifically engineered to take advantage of these quantum properties, computer 0:25:12.680 --> 0:25:20.000 scientists could use these quantum computers to solve particularly difficult problems. Now, 0:25:20.000 --> 0:25:22.680 a quantum computer would not be much good for your 0:25:22.760 --> 0:25:26.880 classic computer processes, so in other words, you wouldn't want 0:25:26.880 --> 0:25:29.359 these as a gaming rig so you can play Overwatch 0:25:29.400 --> 0:25:33.359 on it or something, but quantum computers would be really 0:25:33.400 --> 0:25:38.280 good at tackling certain subsets of computational problems that would 0:25:38.280 --> 0:25:41.960 tie up a classical computer for thousands of years or more. 0:25:43.240 --> 0:25:47.359 So that's quantum computers from a very very high level. Alright, 0:25:47.400 --> 0:25:50.600 what about time crystals. Well, here we got to talk 0:25:50.720 --> 0:25:54.200 about entropy. So you can kind of think of entropy 0:25:54.320 --> 0:25:57.119 as the natural tendency for stuff in the universe to 0:25:57.240 --> 0:26:01.440 kind of fall apart or more accurately, to go from 0:26:01.560 --> 0:26:06.959 more ordered two more disordered. So let's say you've got 0:26:07.119 --> 0:26:10.800 a big room and you've subdivided this big room with 0:26:10.840 --> 0:26:15.240 a removable panel, like it completely bisects the room in half, 0:26:15.920 --> 0:26:20.240 and the panel has like heavy thermal insulation, and you 0:26:20.320 --> 0:26:23.320 make one side of that room really hot, and you 0:26:23.320 --> 0:26:25.040 make the other side of the room that's on the 0:26:25.080 --> 0:26:28.080