WEBVTT - Tech News: A Big Week for Big Tech 0:00:04.400 --> 0:00:07.800 Welcome to Tech Stuff, a production from I Heart Radio. 0:00:11.880 --> 0:00:14.600 Hey there, and welcome to tech Stuff. I'm your host, 0:00:14.800 --> 0:00:18.480 Jonathan Strickland. I'm an executive producer with I Heart Radio. 0:00:18.640 --> 0:00:22.239 And how the tech are you? Y'all? It's time for 0:00:22.320 --> 0:00:26.040 the tech news for January thirty one, twenty twenty three. 0:00:26.200 --> 0:00:31.840 Saying goodbye to January already, And first off, this week 0:00:31.920 --> 0:00:34.760 is actually a really big one for several tech companies 0:00:34.800 --> 0:00:39.720 because it's when they report their earnings for the previous quarter. 0:00:40.280 --> 0:00:42.920 Uh Snap has their call today. By the time you 0:00:43.080 --> 0:00:45.440 hear this episode, that news may already be out there, 0:00:45.920 --> 0:00:48.519 but at the time I'm recording this, it hasn't yet happened. 0:00:49.120 --> 0:00:53.400 Meta has their earnings called tomorrow. Everyone's gonna really pay 0:00:53.440 --> 0:00:55.680 a lot of attention to that. And then Thursday is 0:00:55.720 --> 0:00:58.360 what I like to call a triple a day, and 0:00:58.400 --> 0:01:02.600 by that I mean Amazon, Apple, and Alphabet all have 0:01:02.720 --> 0:01:08.120 their earnings call this Thursday. There's always increased scrutiny into 0:01:08.200 --> 0:01:11.360 these companies around the earnings calls, and I admit that 0:01:11.400 --> 0:01:14.360 can get a little frustrating if you're really just interested 0:01:14.360 --> 0:01:17.600 in the tech, But honestly, it's a really good indicator 0:01:17.640 --> 0:01:20.319 of how these companies are doing, like how the overall 0:01:20.440 --> 0:01:24.039 tech economy is doing, and what we can expect going 0:01:24.120 --> 0:01:28.000 forward because of that, so if investors are unhappy with 0:01:28.040 --> 0:01:30.600 the results, he can see stock prices take a hit 0:01:31.120 --> 0:01:34.040 that in turn can affect a company's plans moving forward. 0:01:34.440 --> 0:01:37.960 For example, Meta has been under some serious fire from 0:01:38.000 --> 0:01:40.959 investors for a while now because of some obstacles that 0:01:41.000 --> 0:01:44.200 the company has faced as far as getting you know, 0:01:44.280 --> 0:01:49.120 more people on the platforms and increasing engagement. Also, Zuckerberg's 0:01:49.160 --> 0:01:52.320 continued commitment to building out the metaverse hasn't won him 0:01:52.320 --> 0:01:55.760 a lot of fans in the investor community. Will probably 0:01:55.800 --> 0:01:59.080 also get a feel for how digital ad spending is 0:01:59.120 --> 0:02:02.280 going across the tech industry because a lot of these 0:02:02.320 --> 0:02:06.280 big tech companies that's where they really depend upon, you know, 0:02:06.360 --> 0:02:09.040 for the revenue, Like Alphabet, you know, the parent company 0:02:09.080 --> 0:02:13.640 for Google, and Meta both really depend on advertising for 0:02:13.760 --> 0:02:17.600 the vast majority of the revenue they generate. Of course, 0:02:18.320 --> 0:02:21.960 these are also companies that have recently engaged in massive layoffs, 0:02:22.000 --> 0:02:24.520 which we'll talk about a bit more in this episode. 0:02:24.760 --> 0:02:28.000 And layoffs, obviously they have a terrible impact on the 0:02:28.000 --> 0:02:31.760 people who are directly affected. Uh, but the investment world 0:02:31.760 --> 0:02:35.160 often reacts positively because I mean it means the company 0:02:35.200 --> 0:02:38.680 has reduced its costs by eliminating workforce, which is a 0:02:38.720 --> 0:02:41.960 pretty clinical way to say. A lot of people found 0:02:42.000 --> 0:02:46.160 themselves without jobs. So it doesn't mean that a slim 0:02:46.200 --> 0:02:49.000 down company is going to perform better, but at least 0:02:49.040 --> 0:02:51.239 in the short term, you end up seeing a bigger 0:02:51.240 --> 0:02:56.200 return on investment because costs drop. And once again I 0:02:56.240 --> 0:02:59.720 get all sour on the concept of capitalism. Now, don't 0:02:59.720 --> 0:03:01.800 get me wrong, I don't have a better alternative up 0:03:01.800 --> 0:03:04.600 my sleeve, so I'm just complaining about something I can't change. 0:03:05.160 --> 0:03:07.560 It just it really feels gross to me to have 0:03:07.600 --> 0:03:11.440 a system where investors are encouraged when people are losing 0:03:11.480 --> 0:03:15.640 their jobs. But that's enough commentary from me. I'm sure 0:03:15.800 --> 0:03:18.400 on Thursday, I'll talk a little bit more about what happened, 0:03:18.400 --> 0:03:21.679 at least during Meta's call, but it probably won't be 0:03:21.760 --> 0:03:24.760 until next week before I can talk about Amazon, Apple, 0:03:24.880 --> 0:03:28.720 and Alphabet. Now, one thing I imagine Apple execs will 0:03:28.760 --> 0:03:32.040 not be addressing on Thursday is how the United States 0:03:32.280 --> 0:03:36.200 National Labor Relations Board or in l r B, has 0:03:36.240 --> 0:03:39.880 accused the company of engaging in practices meant to discourage 0:03:39.880 --> 0:03:43.840 employees from organizing and unionizing. In fact, the n l 0:03:43.960 --> 0:03:48.840 r B says that various official Apple rules quote tend 0:03:49.080 --> 0:03:55.640 to interfere with restraint or coerce employees end quote from organizing. 0:03:56.160 --> 0:03:59.520 This was reported in Bloomberg and y'all. This is not great, 0:04:00.120 --> 0:04:02.920 but it's also not a surprise because tech companies in general, 0:04:03.400 --> 0:04:07.840 and big tech companies in particular, have a pretty well 0:04:07.960 --> 0:04:13.000 documented history of forming policies intended to protect the company 0:04:13.080 --> 0:04:16.640 above the rights of the employees. You hear it all 0:04:16.680 --> 0:04:19.720 the time. That comes in the form of arbitration agreements. 0:04:20.279 --> 0:04:25.080 This is where employees have to agree to settle issues 0:04:25.240 --> 0:04:28.120 within the company by going through like HR as opposed 0:04:28.160 --> 0:04:32.760 to going outside the company to resolve sometimes really serious 0:04:32.800 --> 0:04:36.400 problems like things like sexual harassment or cases of discrimination. 0:04:36.839 --> 0:04:40.560 So the company is compelling employees to to sign an 0:04:40.600 --> 0:04:43.400 agreement that says you won't go to, say, a law firm, 0:04:43.760 --> 0:04:46.359 in order to pursue your case. You have to go 0:04:46.400 --> 0:04:51.000 through the company. There's also the case of being compelled 0:04:51.040 --> 0:04:55.279 to sign non disclosure agreements, which frequently do way more 0:04:55.400 --> 0:04:59.360 to restrict an employee's freedom than anything else. Anyway, Apple 0:04:59.480 --> 0:05:02.000 is not unique to the situation. It's not like they're 0:05:02.040 --> 0:05:04.719 the bad guy and the only one doing this, but 0:05:05.000 --> 0:05:08.120 it is currently facing these accusations from the n l 0:05:08.240 --> 0:05:10.479 r B, and there are a couple of options the 0:05:10.520 --> 0:05:14.359 company could take. One is that Apple could settle with 0:05:14.440 --> 0:05:17.960 the n l r B, presumably that settlement would require 0:05:17.960 --> 0:05:21.960 a commitment to change these various rules and policies, or 0:05:22.400 --> 0:05:25.680 Apple could end up facing a formal complaint from the 0:05:25.800 --> 0:05:28.880 n l RB. I'd say this is another recent example 0:05:28.920 --> 0:05:31.120 of how we're seeing a shift in philosophy when it 0:05:31.160 --> 0:05:34.280 comes to big companies and employee rights here in the 0:05:34.360 --> 0:05:38.479 United States, although whether this ultimately results in massive changes 0:05:38.520 --> 0:05:43.680 throughout business remains to be seen. I'm somewhat pessimistic simply 0:05:43.760 --> 0:05:49.159 because for a very long time, companies essentially had the 0:05:49.520 --> 0:05:52.160 advantage in this area, and it's very hard to kind 0:05:52.160 --> 0:05:55.560 of fight that sort of inertia. While we wait for 0:05:55.600 --> 0:05:59.200 Alphabet's earnings call, we actually have a couple of Google 0:05:59.240 --> 0:06:02.080 stories to talk about out so again a reminder, Alphabet 0:06:02.240 --> 0:06:05.080 is the big umbrella company that is parent to Google 0:06:05.160 --> 0:06:08.520 and YouTube and all that. Anyway, Google recently sent out 0:06:08.560 --> 0:06:11.640 an email to Google THI customers alerting them of a 0:06:11.720 --> 0:06:14.480 data breach now, Google Fi, in case you're not aware, 0:06:15.080 --> 0:06:19.440 is Google's cellular service company, so it's like a telecommunication service, 0:06:19.680 --> 0:06:23.240 and Google piggybacks on top of other companies infrastructure to 0:06:23.320 --> 0:06:27.080 provide these services, so it's not like Google has its 0:06:27.120 --> 0:06:31.000 own cell towers out there. Instead, they essentially are leasing 0:06:31.040 --> 0:06:35.120 that from another company, companies like T Mobile, which had 0:06:35.160 --> 0:06:38.520 its own data breach recently, and it is possible that 0:06:38.560 --> 0:06:42.200 this data breach for Google Fi is a direct result 0:06:42.440 --> 0:06:45.360 of the T Mobile breach, but we don't have confirmation 0:06:45.400 --> 0:06:48.839 on that. According to Google, hackers access to database that 0:06:48.880 --> 0:06:52.400 included information on customers, but the info was limited to 0:06:52.440 --> 0:06:55.920 stuff like phone numbers, account status so whether they're active 0:06:56.000 --> 0:07:00.600 or inactive, simcards, serial numbers, and details about the service 0:07:00.600 --> 0:07:03.719 plan that each customer subscribed to. Now, what was not 0:07:03.839 --> 0:07:09.200 included were names and other personal information. Also, Google says 0:07:09.279 --> 0:07:13.080 no passwords or payment data or text message content was 0:07:13.120 --> 0:07:16.360 included in the data, so that's good. At least. Google 0:07:16.400 --> 0:07:19.160 did warn customers to be on the lookout for phishing scams, 0:07:19.200 --> 0:07:21.880 which the company stressed is a good idea anyway, but 0:07:21.920 --> 0:07:25.280 it's particularly good when you have a data breach like this, 0:07:25.760 --> 0:07:28.640 And according to tech Crunch, at least one Google THI 0:07:28.760 --> 0:07:32.360 customer has reported being the victim of SIM swapping. That's 0:07:32.400 --> 0:07:34.840 when a hacker hijacks a SIM card to take control 0:07:34.920 --> 0:07:38.080 of a phone number, and that allows them to send 0:07:38.120 --> 0:07:41.280 and receive phone calls as if they were that phone, 0:07:41.920 --> 0:07:45.400 and as well as as be able to access voicemail, 0:07:45.400 --> 0:07:49.120 which is a big old yikes. Speaking of Google, as 0:07:49.120 --> 0:07:53.800 I mentioned earlier, the company recently laid off some employees 0:07:54.000 --> 0:07:57.040 like it has plans to lay off twelve thousand total, 0:07:57.760 --> 0:08:00.840 and one of those employees was christ In max o 0:08:01.600 --> 0:08:05.440 Um or max co. I apologize christ and I do 0:08:05.480 --> 0:08:07.280 not know how to pronounce your last name, and that's 0:08:07.320 --> 0:08:10.760 my fault. But Kristen joined Google back in two thousand 0:08:10.800 --> 0:08:13.240 and eight and had served as the company's head of 0:08:13.320 --> 0:08:16.960 mental health and well being, and apparently her team was 0:08:17.080 --> 0:08:21.000 hit significantly by these layoffs. And I think that sends 0:08:21.040 --> 0:08:25.720 a pretty bleak message to Google employees who remain at Google, 0:08:26.120 --> 0:08:28.640 because I imagine hearing that the mental health and well 0:08:28.680 --> 0:08:33.040 being department has been largely cleared out. It's not great 0:08:33.080 --> 0:08:36.560 news when you start thinking about what that implies, because 0:08:36.559 --> 0:08:39.320 it suggests that Google does not view mental health and 0:08:39.360 --> 0:08:43.920 well being of its employees as a critical component for 0:08:44.160 --> 0:08:48.360 its business. And that's just not a good sign. I've 0:08:48.440 --> 0:08:52.800 worked for companies where I ultimately felt like the employees 0:08:52.840 --> 0:08:58.240 were unappreciated or taken advantage of, and it is the worst, right, 0:08:58.320 --> 0:09:01.200 even if you love what you do, if you feel 0:09:01.240 --> 0:09:04.240 like you're working for an organization that doesn't have an 0:09:04.280 --> 0:09:08.680 appreciation for its own employees, that's demoralizing and that could 0:09:08.679 --> 0:09:11.719 be enough to convince someone to leave, even if they 0:09:11.720 --> 0:09:15.440 love their work. Trust me, I know, because I did 0:09:15.520 --> 0:09:21.680 that anyway. According to Insider, uh, some employees hit by 0:09:21.679 --> 0:09:25.360 the layoffs received kind of a terse impersonal email that 0:09:25.480 --> 0:09:27.760 essentially just told them that they no longer have a 0:09:27.880 --> 0:09:30.800 role at the company and that and at least some 0:09:30.920 --> 0:09:34.280 of the cases, a few employees said their managers didn't 0:09:34.280 --> 0:09:36.719 even know that this had happened, which suggests that the 0:09:36.760 --> 0:09:41.800 managers weren't consulted before these layoffs started to roll across 0:09:41.840 --> 0:09:45.040 the company. So the impression I get from the article 0:09:45.160 --> 0:09:48.719 in Insider says that you know that these layoffs may 0:09:48.720 --> 0:09:53.120 have lacked forethought and consideration. Now, maybe I'm reading too 0:09:53.200 --> 0:09:54.960 much into it, but that's what it says to me. 0:09:55.800 --> 0:09:58.160 And again that could really hurt the company in the 0:09:58.240 --> 0:10:00.400 long run if it turns out that some of these 0:10:00.480 --> 0:10:03.800 layoffs were for people that really they needed in order 0:10:03.840 --> 0:10:07.520 to keep certain projects going and to ensure success. But again, 0:10:08.320 --> 0:10:10.839 I don't know how I would go about telling twelve 0:10:10.880 --> 0:10:13.560 thousand people they don't work for me anymore, because I 0:10:13.600 --> 0:10:17.679 think singing telegrams would get too expensive. All right, We've 0:10:17.720 --> 0:10:20.000 got a lot more news to cover before we get 0:10:20.000 --> 0:10:32.280 to that. Let's take a quick break. Okay, we're back, 0:10:32.320 --> 0:10:34.640 and I got a little bit more Google news. One 0:10:34.679 --> 0:10:38.200 person who had some words of warning for Google is 0:10:38.280 --> 0:10:41.839 Paul Butchit or book Hite. I don't know how to 0:10:41.880 --> 0:10:44.200 say his last name either. I'm I'm batting a thousand 0:10:44.200 --> 0:10:48.160 when it comes to names today, y'all. But anyway, Paul 0:10:48.280 --> 0:10:51.400 as I will call him lead development on Gmail, so's 0:10:51.440 --> 0:10:56.080 he's credited as the person who created Gmail. And he 0:10:56.200 --> 0:11:00.680 predicts that chat gpt will completely we're a place Google 0:11:00.760 --> 0:11:02.839 search within a year or two, as people start to 0:11:02.920 --> 0:11:06.520 lean on chat gpt to get their questions answered. Instead 0:11:06.559 --> 0:11:08.640 of going to search for the right length and then 0:11:08.679 --> 0:11:11.400 searching for their answer there, they'll just ask the question 0:11:11.440 --> 0:11:13.760 to chat gpt. They'll get the answer and then they'll 0:11:13.760 --> 0:11:17.080 just skip Google entirely, and this in turn will destroy 0:11:17.160 --> 0:11:21.160 Google because Google depends so heavily on ads that are 0:11:21.160 --> 0:11:25.080 served within search results. Paul points out that even if 0:11:25.080 --> 0:11:29.280 Google deploys its own AI tool, which we know that 0:11:29.320 --> 0:11:32.320 Google has been developing, that's not a secret. The company 0:11:32.360 --> 0:11:36.080 has been very uh forthright on that, saying yes, they're 0:11:36.080 --> 0:11:38.480 working on AI, but it's not something that they're ready 0:11:38.520 --> 0:11:42.080 to release yet because it could be potentially harmful. Well, 0:11:42.160 --> 0:11:44.120 Paul says, the effect would still be the same if 0:11:44.160 --> 0:11:48.000 Google launches its own version of this, because again, it 0:11:48.120 --> 0:11:52.120 bypasses search, which means it bypasses where you would normally 0:11:52.160 --> 0:11:55.640 serve up ads. But y'all, I want to just point 0:11:55.640 --> 0:11:59.120 out that it would do way more than just that. 0:11:59.320 --> 0:12:02.960 It would hurt way more than just Google, because think 0:12:02.960 --> 0:12:05.320 about what shows up in search results in the first place. 0:12:05.480 --> 0:12:09.880 It's it's a list of links to external websites. That's 0:12:09.920 --> 0:12:12.400 where all this data lives. Right when you're asking a 0:12:12.480 --> 0:12:16.000 question from Google, the data lives on these websites, and 0:12:16.600 --> 0:12:18.800 you know, the goal is to have you click on 0:12:18.840 --> 0:12:20.960 the links so that you visit the website, and then 0:12:21.200 --> 0:12:24.480 chances are that website is generating revenue from ad results too. 0:12:25.280 --> 0:12:29.720 But if chat GPT disrupts web search to that extent, 0:12:30.600 --> 0:12:32.880 it's not just gonna hit Google. It's going to hit 0:12:33.000 --> 0:12:36.000 all the sites that Google links out to, and that 0:12:36.559 --> 0:12:39.880 you know chat GPT is drawing information from these sites 0:12:40.480 --> 0:12:43.520 because it's not like chat gpt just magically knows the 0:12:43.559 --> 0:12:48.640 answers to questions. It's pulling data from sources in a 0:12:48.679 --> 0:12:52.720 way that's not entirely or even particularly transparent. So we 0:12:52.720 --> 0:12:57.719 don't actually know where chat gpt is pulling data from 0:12:57.720 --> 0:13:00.400 in any given moment. But think about it's for just 0:13:00.480 --> 0:13:06.000 a moment. If search essentially dies, then how do websites 0:13:06.080 --> 0:13:09.480 attract users to their sites? Now? Subsites are kind of 0:13:09.520 --> 0:13:13.280 like landing pages. Folks might bookmark them, they might go 0:13:13.360 --> 0:13:16.920 to them regularly and not rely on search. Right, you 0:13:17.000 --> 0:13:19.920 might have certain news sites that you go to on 0:13:19.960 --> 0:13:23.319 a regular basis. Those cases, you know, they get a 0:13:23.400 --> 0:13:26.600 lot of traffic that's coming in just from folks jumping 0:13:26.640 --> 0:13:30.079 straight to the website. But a lot of sites depend 0:13:30.160 --> 0:13:34.360 heavily on traffic from search engines. People are searching for 0:13:34.400 --> 0:13:37.600 specific terms, they come up with responses, they go to 0:13:37.679 --> 0:13:41.400 those websites, and that's where most of the traffic coming 0:13:41.400 --> 0:13:44.400 to those websites comes from, the search engines. So if 0:13:44.400 --> 0:13:49.800 the search engine goes down, that traffic disappears. Then eventually 0:13:50.120 --> 0:13:52.360 these sites start to go out of business because no 0:13:52.400 --> 0:13:54.840 one's visiting them, so they're not getting any ad revenue, 0:13:55.440 --> 0:13:58.120 so they you know, it costs money to keep those 0:13:58.160 --> 0:14:01.600 sites up. It's eventually just gonna die. Now, ultimately that 0:14:01.640 --> 0:14:05.440 means that the information sources that chat GPT depends upon 0:14:06.040 --> 0:14:08.680 will start dying off, and that would mean that the 0:14:08.720 --> 0:14:12.280 tool would become less reliable or maybe more biased, depending 0:14:12.360 --> 0:14:14.920 upon what sites are sticking around. So, in other words, 0:14:15.880 --> 0:14:20.160 I don't think a future where chat GPT takes over 0:14:20.240 --> 0:14:22.640 for search is bad just for Google. I think it's 0:14:22.640 --> 0:14:27.720 bad for everyone, including chat GPT, in the long run. 0:14:28.640 --> 0:14:30.480 So that's something to think about. I don't know that 0:14:30.480 --> 0:14:34.000 there's any way to avoid it, but I definitely see 0:14:34.000 --> 0:14:37.640 it as being a bad thing like that would require 0:14:38.160 --> 0:14:42.680 some careful work in order to either avoid or fix 0:14:42.760 --> 0:14:47.320 that problem. Okay, now let me paint a scenario for you. 0:14:47.600 --> 0:14:50.040 This is a new story. Let's say that you're you 0:14:50.080 --> 0:14:55.000 are an AI company that specializes in creating synthetic human voices, 0:14:55.320 --> 0:14:58.760 including voices that could sound really close or maybe even 0:14:58.800 --> 0:15:02.760 identical to the voices of specific people, Like you could 0:15:02.840 --> 0:15:07.040 replicate a voice like famous people, and you've got this 0:15:07.120 --> 0:15:10.800 tool that can learn how to reproduce the sound and 0:15:10.920 --> 0:15:15.800 vocal quality of a real, specific human beings voice. Now 0:15:16.360 --> 0:15:19.680 would it shock you to learn that some people on 0:15:19.720 --> 0:15:22.400 the Internet would use this kind of tool to do 0:15:22.520 --> 0:15:27.760 awful things? Oh? You wouldn't. Oh I So you've been 0:15:27.760 --> 0:15:30.680 on the internet before, you already you realize this would 0:15:30.680 --> 0:15:32.760 be a problem, right the gate right, Like, yeah, of 0:15:32.800 --> 0:15:37.520 course it's obvious. Well, that means you're ahead of eleven Labs. 0:15:37.520 --> 0:15:40.920 That's a company that makes a voice synthesis tool and 0:15:41.280 --> 0:15:45.680 launched a beta for its AI generated voice tool, and 0:15:45.720 --> 0:15:48.680 then even more recently than that has announced that it 0:15:48.760 --> 0:15:54.640 had detected a quote increased number of voice clothing misuse 0:15:54.760 --> 0:15:58.240 cases end quote. So this is all reported in Vice, 0:15:58.640 --> 0:16:02.080 which has a very long art call detailing how members 0:16:02.160 --> 0:16:06.640 of four chan had enrolled in this beta and then 0:16:06.720 --> 0:16:09.880 use the tools to create synthesized AI replications of voices 0:16:09.920 --> 0:16:14.680 belonging to folks like the actress Emma Watson or podcaster 0:16:14.920 --> 0:16:18.120 Joe Rogan and others. And then, and I'm sure this 0:16:18.200 --> 0:16:21.560 is gonna shock you, they use these synthetic voices to 0:16:21.680 --> 0:16:24.560 make them say all sorts of truly awful stuff. I'm 0:16:24.600 --> 0:16:30.560 talking like homophobic and transphobic comments, racial slurs, violent threats, 0:16:30.600 --> 0:16:34.040 all sorts of stuff like that. Again, this is only 0:16:34.120 --> 0:16:36.920 shocking for someone who maybe was born yesterday, but for 0:16:36.960 --> 0:16:39.800 anyone who actually is aware that four chan is a thing, 0:16:40.360 --> 0:16:44.000 this is not surprising even a little bit. And considering 0:16:44.000 --> 0:16:47.600 we've already seen plenty of people use deep faked technology 0:16:47.640 --> 0:16:51.200 to create content without the consent of the folks who 0:16:51.240 --> 0:16:55.000 were replicated in those videos, you can only imagine the 0:16:55.040 --> 0:16:57.360 sort of stuff we could see in the future. And 0:16:57.400 --> 0:16:59.320 I think I'm gonna need to do a full episode 0:16:59.400 --> 0:17:03.000 about this problem. I probably want to talk with someone 0:17:03.040 --> 0:17:06.680 who has either experienced this kind a violation of their 0:17:06.720 --> 0:17:10.440 identity themselves, or someone who has worked closely with people 0:17:10.480 --> 0:17:13.920 who have experienced it, because I honestly think it's really 0:17:13.960 --> 0:17:18.800 important to realize the the extent of the harm this 0:17:19.000 --> 0:17:23.199 abuse causes. Uh, And I wanted to come from a 0:17:23.240 --> 0:17:25.800 genuine place, so I'll probably be on the lookout see 0:17:25.800 --> 0:17:27.399 if I can find someone who would be willing to 0:17:27.440 --> 0:17:32.400 talk about that, because it's a really despicable practice too, 0:17:33.080 --> 0:17:38.280 to replicate someone without their consent. It is a true violation. Anyway, 0:17:38.400 --> 0:17:41.159 eleven Labs has chosen to lock down the beta and 0:17:41.200 --> 0:17:44.520 revisit their approach to safeguards, which he has a good idea. 0:17:44.560 --> 0:17:48.359 It sounds like they're going to require identification verification before 0:17:48.359 --> 0:17:52.439 you're allowed to replicate a specific voice, and to to 0:17:52.760 --> 0:17:55.199 show that you have the consent to do that. I 0:17:55.200 --> 0:17:59.399 think that's the bare minimum you you can do. And 0:17:59.400 --> 0:18:01.720 obviously you have to keep the tool underlock and key. 0:18:01.760 --> 0:18:04.560 You can't just release it because otherwise people would just 0:18:04.640 --> 0:18:10.000 exploit it. So interesting story, Uh, disturbing look at the future. 0:18:10.280 --> 0:18:12.240 But we knew this was coming, I mean because deep 0:18:12.280 --> 0:18:14.720 fakes have been a thing for a while now. The 0:18:14.720 --> 0:18:19.120 Wall Street Journal reports that despite US export restrictions that 0:18:19.200 --> 0:18:23.160 should prevent certain Chinese companies and organizations from being able 0:18:23.200 --> 0:18:26.719 to purchase stuff like cutting edge computer chips, the China 0:18:26.800 --> 0:18:29.959 Academy of Engineering Physics seems to be able to do 0:18:30.000 --> 0:18:32.679 that without any real problem. Uh. This is one of 0:18:32.680 --> 0:18:35.680 the organizations that the United States has already put on 0:18:35.760 --> 0:18:40.760 a blacklist for semi conductor companies and chip manufacturers. In fact, 0:18:40.840 --> 0:18:44.879 the US has told companies like Nvidia and Intel expressly 0:18:45.000 --> 0:18:48.880 do not sell hardware to the China Academy of Engineering 0:18:48.880 --> 0:18:54.200 Physics because the organization's work includes work in nuclear weapons research. 0:18:55.040 --> 0:18:58.639 So if that ban is in place, how is this 0:18:58.760 --> 0:19:04.000 organization acquiring the chips? But largely through resellers in China. 0:19:04.480 --> 0:19:07.960 You have other entities in China buying up these these 0:19:08.000 --> 0:19:12.440 components and then reselling those two organizations like the China 0:19:12.480 --> 0:19:16.200 Academy of Engineering Physics. It is very hard. It's almost 0:19:16.280 --> 0:19:20.439 impossible to prevent all sophisticated chips from being shipped to 0:19:20.520 --> 0:19:23.880 China because some of these components were clearly originally part 0:19:23.920 --> 0:19:26.800 of like a larger system, like like a PC or 0:19:26.840 --> 0:19:29.800 a gaming rig. The Wall Street General reports that the 0:19:29.840 --> 0:19:33.080 chips in question are typically a couple of years old, 0:19:33.240 --> 0:19:36.200 so they aren't the most powerful, most recent ones on 0:19:36.240 --> 0:19:39.680 the market. It's hard to imagine how companies can make 0:19:39.720 --> 0:19:42.720 certain that all retailers that they work with comply with 0:19:42.760 --> 0:19:46.200 these rules, none of them are shipping to China. Meanwhile, 0:19:46.680 --> 0:19:50.199 China has to purchase these chips because it lacks the 0:19:50.240 --> 0:19:55.080 manufacturing facilities to make the most sophisticated processors, particularly at scale. 0:19:55.480 --> 0:19:59.399 China lags behind everybody else by a few generations, and 0:19:59.440 --> 0:20:02.720 so these black lists do slow down China's military use 0:20:02.720 --> 0:20:05.920 of such technology. But the article really does point out 0:20:06.080 --> 0:20:10.199 how hard it is to enforce a total ban. Okay, 0:20:10.560 --> 0:20:12.720 I've got a few more tech stories I want to 0:20:12.760 --> 0:20:15.280 talk about. Before I get to that, Let's take another 0:20:15.359 --> 0:20:29.560 quick break. We're backed. Tech Dirt has an informative piece 0:20:29.600 --> 0:20:33.880 by Mike Masnick about how a paralegal named Katherine Twoson 0:20:34.640 --> 0:20:39.240 is pulling at the threads holding together Joshua Browner's story 0:20:39.400 --> 0:20:42.560 and image. So Browner, in case you don't know, is 0:20:42.600 --> 0:20:46.080 the CEO of a company called Do Not Pay. I 0:20:46.119 --> 0:20:49.280 have mentioned Do Not Pay a few times already this year. 0:20:49.800 --> 0:20:53.160 Uh It's best known for being an AI focused company 0:20:53.320 --> 0:20:56.000 that aims to use artificial intelligence to help people do 0:20:56.160 --> 0:21:00.400 stuff ranging from canceling subscription services that they're not using 0:21:00.440 --> 0:21:06.720 anymore to fighting parking tickets and other legal related work. Well, 0:21:06.880 --> 0:21:10.600 Twoson decided to put some of Do Not Pays services 0:21:10.640 --> 0:21:13.919 to the test. She reported that her attempts to have 0:21:14.040 --> 0:21:17.880 the AI generate certain legal documents were largely a failure. 0:21:18.280 --> 0:21:21.199 That in one case, it was pretty much like the 0:21:21.240 --> 0:21:23.480 service was just kind of filling in the blanks on 0:21:23.520 --> 0:21:26.920 a form letter kind of approach, and that in another 0:21:27.440 --> 0:21:30.760 she put uh in for a more complicated legal document 0:21:30.800 --> 0:21:32.800 and received a response that it would take a few 0:21:32.800 --> 0:21:38.080 hours to process. She's suggests that maybe that indicates that 0:21:38.119 --> 0:21:41.680 there is little to know AI involvement in that process 0:21:41.720 --> 0:21:44.560 at all, That if it were a computer generated you 0:21:44.560 --> 0:21:48.240 wouldn't expect it to be something that takes a long time, 0:21:49.080 --> 0:21:52.120 and that perhaps it means that there are actual humans 0:21:52.160 --> 0:21:54.920 putting together those kind of requests, or at least having 0:21:55.640 --> 0:22:00.399 a heavy involvement in editing an AI generated piece. In 0:22:00.400 --> 0:22:03.520 other words, that it seems to indicate the AI is 0:22:03.560 --> 0:22:09.160 not quite as sophisticated and effective as the company might 0:22:09.480 --> 0:22:13.720 otherwise indicate. She has also investigated some of Browder's claims, 0:22:13.720 --> 0:22:17.359 including one where he tweeted he would buy up ten 0:22:17.440 --> 0:22:21.320 dollars of medical debt for every retweet and follow that 0:22:21.480 --> 0:22:26.160 This particular tweet message got Tuson suspected the Browner had 0:22:26.200 --> 0:22:30.920 not followed through on that promise at that point. Browder 0:22:30.960 --> 0:22:34.280 then produced a receipt showing a five hundred dollar donation, 0:22:34.440 --> 0:22:37.800 which bought up around fifty grand of medical debt because 0:22:37.880 --> 0:22:41.200 debt is sold for pennies on the dollar. But Twoson 0:22:41.560 --> 0:22:46.200 found this receipt somewhat suspicious and hypothesized that Browder had 0:22:46.320 --> 0:22:49.040 not made the donation when he claimed he did. He 0:22:49.080 --> 0:22:52.240 claimed he had made it back in early December last year, 0:22:52.920 --> 0:22:57.159 but Twoson looked closer at the receipt and saw that 0:22:57.240 --> 0:23:01.520 the date on the receipt and that was next to 0:23:01.520 --> 0:23:04.600 the line item failed to be on the exact same 0:23:06.560 --> 0:23:09.960 line horizontally across the receipt. She said, oh, it's it's misaligned. 0:23:10.280 --> 0:23:13.840 She even did her own donation to buy up medical 0:23:13.920 --> 0:23:18.639 debt and showed that her receipt everything was in perfect alignment, 0:23:18.960 --> 0:23:22.679 but his receipt showed a little misalignment, which suggested the 0:23:22.720 --> 0:23:26.239 possibility of a photoshop job to change the date. And 0:23:26.280 --> 0:23:28.960 then Tuson decided to look into it further by actually 0:23:29.000 --> 0:23:32.520 contacting the organization that sold the medical debt in the 0:23:32.560 --> 0:23:37.160 first place, and found that Browner's purchase happened four minutes 0:23:37.640 --> 0:23:43.160 after she brought into question his follow through. So she says, hey, 0:23:43.560 --> 0:23:49.440 whatever happened to this guarantee? Four minutes later? He apparently 0:23:49.480 --> 0:23:53.400 makes this purchase and then subsequently says, yeah, I did 0:23:53.400 --> 0:23:56.840 it way back in December. So this piece really paints 0:23:56.880 --> 0:23:59.520 Browder in a critical light, suggesting that he is leaning 0:23:59.520 --> 0:24:02.399 perhaps a little too hard on publicity stunts and hype. 0:24:03.080 --> 0:24:07.080 We've been hearing these kinds of of thoughts about Browner 0:24:07.160 --> 0:24:09.879 for a while now, particularly in the wake of his 0:24:09.960 --> 0:24:13.080 offer to pay a million dollars to a lawyer who 0:24:13.080 --> 0:24:16.359 would use do not pays AI to argue a case 0:24:16.400 --> 0:24:18.639 before the U. S. Supreme Court. So I guess the 0:24:18.760 --> 0:24:22.080 lesson here is to rely on critical thinking and investigation 0:24:22.160 --> 0:24:25.040 like Tucson did, and do your best to try and 0:24:25.560 --> 0:24:30.360 hear through the hype. Speaking of hype, Mercedes Benz announced 0:24:30.440 --> 0:24:33.320 last week that the company will introduce a level three 0:24:33.440 --> 0:24:37.440 autonomous driving feature in vehicles here in the United States 0:24:37.840 --> 0:24:41.359 later this year. So the various levels of autonomous driving 0:24:41.560 --> 0:24:44.520 come out of the Society of Automotive Engineers that kind 0:24:44.520 --> 0:24:49.000 of defined the levels. They identified six levels from zero 0:24:49.119 --> 0:24:52.000 to five. So a level zero vehicle essentially has no 0:24:52.080 --> 0:24:55.879 autonomous features and is strictly manual. A Level five vehicle 0:24:55.920 --> 0:24:59.800