WEBVTT - Week In Tech: Zuck Takes the Stand 0:00:12.640 --> 0:00:16.520 Welcome to Tech Stuff, a production of iHeart Podcasts and Kaleidoscope. 0:00:16.880 --> 0:00:19.120 I'm as Valoshian and today Cara Price and I will 0:00:19.120 --> 0:00:22.960 bring you the headlines this week, including some surprisingly rapid 0:00:23.040 --> 0:00:26.760 changes in the way people are using generative AI. Then 0:00:26.800 --> 0:00:29.760 on tech Support, we'll talk to the Washington Posts Naomi 0:00:29.840 --> 0:00:33.600 Nix about the first week of the Meta antitrust trial. 0:00:33.960 --> 0:00:35.879 This latest trial is a Now they're a piece of 0:00:35.920 --> 0:00:39.199 evidence that the companies haven't yet, to put it crudely, 0:00:39.200 --> 0:00:40.520 gotten what they've paid for. 0:00:40.880 --> 0:00:48.080 All of that. On the Weekend Tech It's Friday, April eighteenth, 0:00:49.120 --> 0:00:50.640 Hello us, Hello Cara. 0:00:51.080 --> 0:00:52.960 So I was thinking about you this week. I was 0:00:52.960 --> 0:00:58.240 seeing something Oliver Instagram, which is my preferred social media platform. 0:00:58.320 --> 0:01:01.280 Your preferred social media platform has nothing to do with socializing. 0:01:01.360 --> 0:01:02.000 It's called Lincoln. 0:01:02.440 --> 0:01:05.319 Well, maybe you're too old for TikTok and I'm too 0:01:05.319 --> 0:01:09.200 old for Instagram. So yes, LinkedIn is my absolutely. 0:01:08.760 --> 0:01:11.280 Right, at least you're not using Facebook. So on this 0:01:11.560 --> 0:01:15.600 I saw action figures. Yeah, that looked a lot like 0:01:15.680 --> 0:01:17.679 my friends, like my friend who works at Tory Birch. 0:01:17.720 --> 0:01:20.560 There was like there was an action figure of like 0:01:20.600 --> 0:01:23.479 a jewel and an iced coffee and a Tory Birch handbag. 0:01:23.920 --> 0:01:26.440 In a weird way, I've looked at it as a 0:01:26.480 --> 0:01:29.760 sort of LinkedIn's own version of the Jibbli portraits that 0:01:29.760 --> 0:01:31.679 we talked about a few weeks ago. You know, to 0:01:31.800 --> 0:01:34.360 make these, you upload a picture of yourself to chat 0:01:34.400 --> 0:01:37.440 GBT and prompt the model to turn you into a 0:01:37.480 --> 0:01:39.560 picture of a toy action figure. 0:01:39.880 --> 0:01:43.920 Imagine going into a toy store in nineteen eighty four 0:01:44.480 --> 0:01:48.040 and you have plastic packed action figurines like could be 0:01:48.040 --> 0:01:51.280 a Bobbie, could be a Gi Joe with some special 0:01:51.440 --> 0:01:54.560 swag that the character has which reflects on who the 0:01:54.640 --> 0:01:55.200 character is. 0:01:55.320 --> 0:01:57.240 Yeah, and I think in a weird way, it sort 0:01:57.280 --> 0:02:00.880 of acts as this like pictorial resume of like this 0:02:00.920 --> 0:02:03.840 is who I like. I am podcast fundraiser, I have 0:02:04.040 --> 0:02:06.080 jim bag, I have a six sneakers. 0:02:06.080 --> 0:02:06.960 What are you talking about? 0:02:07.160 --> 0:02:10.760 I am here asking for some money. I don't know 0:02:10.760 --> 0:02:11.600 what I'm talking about. 0:02:11.960 --> 0:02:13.480 Do you know what's very interesting? I used to work 0:02:13.520 --> 0:02:15.760 for a branding agency. I know that you know that, 0:02:15.880 --> 0:02:20.919 and our business cards were a pictorial representation of us 0:02:21.000 --> 0:02:24.680 and then a few representations of things that were important 0:02:24.680 --> 0:02:27.160 to us. So I had like a British flag and 0:02:27.240 --> 0:02:28.400 like a tennis racking or. 0:02:28.320 --> 0:02:31.200 Whatever one visa on visa. 0:02:31.120 --> 0:02:33.120 Exactly, you know. And there was like kind of a 0:02:33.160 --> 0:02:35.240 talking point at the beginning of the meeting and this 0:02:35.360 --> 0:02:38.760 is kind of the LinkedIn meets chatchybt version of that, 0:02:39.160 --> 0:02:40.360 and it's going completely viral. 0:02:40.520 --> 0:02:43.639 It's going completely viral and people love. I don't know 0:02:43.680 --> 0:02:45.880 if it's uncanny. I think it is uncanny Valley, like 0:02:45.919 --> 0:02:49.440 people love to see themselves represented in an animated format 0:02:49.639 --> 0:02:52.760 because it's sort of like the last frontier of like 0:02:53.440 --> 0:02:56.000 reality turned into surreality one hundred percent. 0:02:56.040 --> 0:02:59.120 I'm sure you've seen The Real Housewives as Pixar characters, 0:03:00.160 --> 0:03:01.120 which is so good. 0:03:00.960 --> 0:03:02.800 Which they're starting to look like anyway with all the 0:03:02.800 --> 0:03:05.160 filters that you can use to post them on Instagram. 0:03:05.280 --> 0:03:07.920 But you know what this is what is driving something 0:03:07.960 --> 0:03:12.680 extraordinary that happened, which is that CHATGPT excluding games, became 0:03:12.720 --> 0:03:15.560 the most downloaded app in the world last month. 0:03:15.720 --> 0:03:18.760 Yeah, and actually at last week's TED conference, Sam Altman 0:03:18.880 --> 0:03:22.000 let slip that the usage of chat gbt had doubled 0:03:22.080 --> 0:03:26.560 in just a few weeks, indicating it now has eight 0:03:26.680 --> 0:03:29.880 hundred million weekly active users. 0:03:30.160 --> 0:03:32.079 Yeah, I mean, it is an extraordinary number, and I 0:03:32.160 --> 0:03:36.119 think you use the word indicating advisedly because he said 0:03:36.160 --> 0:03:39.280 it was about ten percent of the world's population. You know, 0:03:39.320 --> 0:03:41.440 at the beginning of this year, we were talking about 0:03:41.440 --> 0:03:44.840 would the business model for these charative ami companies ever work. 0:03:45.200 --> 0:03:48.320 There's a story in Information this week that the revenues 0:03:48.440 --> 0:03:53.400 at OpenAI are picking up dramatically. They've grown thirty percent 0:03:53.520 --> 0:03:56.120 so far this year, So that's in one quarter to 0:03:56.160 --> 0:03:59.800 around four hundred and fifteen million dollars per month in revenue. 0:03:59.400 --> 0:04:01.600 Which is very interesting for a company that built itself 0:04:01.600 --> 0:04:04.000 as a non brop no. But in some ways, I'm 0:04:04.000 --> 0:04:06.640 not surprised because it feels like there's a new use 0:04:06.680 --> 0:04:09.960 case for chat GBT every week. I'm more invested in 0:04:09.960 --> 0:04:12.360 it every week, and so I am sort of always 0:04:12.360 --> 0:04:17.080 interested when I see the data reflected in my own personal. 0:04:16.800 --> 0:04:18.960 Use I mean, you get hooked on, oh my god, 0:04:19.120 --> 0:04:21.800 what would my family photo look like as a Pixar movie? 0:04:22.120 --> 0:04:24.680 And then it's like, huh, you know, I what if 0:04:24.720 --> 0:04:28.280 I put this legal contract through chat GPT and asked 0:04:28.279 --> 0:04:29.880 it to advise me on what were the key points 0:04:29.880 --> 0:04:31.760 I should be paying attention to, which I did for 0:04:31.800 --> 0:04:34.159 the first time last week after I used the image 0:04:34.160 --> 0:04:40.080 generation feature. So I mean, it's a remarkable moment where 0:04:40.120 --> 0:04:45.120 I think a true consumer adoption is creating the business model, 0:04:45.600 --> 0:04:47.599 or could be beginning to create the business model, which 0:04:47.600 --> 0:04:49.599 is a question some people had, would that ever really 0:04:49.600 --> 0:04:52.920 actually happen? But we are not the only people observing 0:04:53.000 --> 0:04:56.000 what's happening. There was something actually fascinating I read last 0:04:56.040 --> 0:04:58.520 week in the Harvard Business Review that might shock you. 0:04:58.560 --> 0:05:01.360 So tell me about something fat fascinating that you actually 0:05:01.360 --> 0:05:04.000 were able to find in the Harvard Business Review. 0:05:04.120 --> 0:05:08.200 Well, the article was about AI use cases, and according 0:05:08.240 --> 0:05:11.680 to Harvard Business Review, the most common reason for generative 0:05:11.880 --> 0:05:17.840 AI usage over the last twelve months was therapy and companionship. 0:05:18.320 --> 0:05:20.919 It's actually very funny you say this, and this is 0:05:20.960 --> 0:05:24.039 not a setup. I've actually met a woman the other 0:05:24.160 --> 0:05:27.880 day who told me she'd created her own therapy bot, 0:05:28.240 --> 0:05:30.920 calling it a GPT. And I've been seeing a ton 0:05:30.960 --> 0:05:34.080 of stuff about using AI models for therapy everywhere, and 0:05:34.120 --> 0:05:36.479 I was honestly starting to wonder if all these articles 0:05:36.480 --> 0:05:38.960 on AI relationships were a bit blown out of proportion, 0:05:39.279 --> 0:05:41.120 But it seems to be a very real trend. 0:05:41.279 --> 0:05:45.039 Well, you've taken endorsement after slamming me for being fascinated 0:05:45.040 --> 0:05:47.400 by the Harvard Business Review, You've been swayed by their 0:05:47.800 --> 0:05:49.000 validation of your worldview. 0:05:49.120 --> 0:05:52.120 Judge where information comes from unless it's the Harvard Business Review. 0:05:52.360 --> 0:05:56.680 So this is a study and the methodology is quite interesting. 0:05:56.880 --> 0:05:58.760 Some might question it, but I thought it was interesting. 0:05:58.960 --> 0:06:03.880 They're basically a tool to scrape online forums like primarily 0:06:03.920 --> 0:06:06.960 Reddit but also kra and a few others that scrape 0:06:07.000 --> 0:06:09.800 for every single mention of how people are using AI, 0:06:10.240 --> 0:06:14.279 categorize them and then counted the number of posts about 0:06:14.360 --> 0:06:16.440 each use case, you know, filtering out the garbage and 0:06:16.480 --> 0:06:18.960 whatever else, to come up with the stat rank of 0:06:19.000 --> 0:06:21.400 the top one hundred use cases that people are talking 0:06:21.400 --> 0:06:22.520 about how they use AI. 0:06:22.920 --> 0:06:27.680 And so Reddit really is the treasure trove of this discovery. 0:06:28.040 --> 0:06:29.400 I think it is, and I think you know there's 0:06:29.440 --> 0:06:32.200 a reason why is because people post without using their 0:06:32.240 --> 0:06:35.080 real names in a very unfiltered way on Reddit, and 0:06:35.120 --> 0:06:37.400 so I think as a proxy for how people are 0:06:37.480 --> 0:06:40.320 using technology, you could do a lot worse than Reddit. 0:06:40.800 --> 0:06:42.719 And in fact, I also read in the Harvard Business 0:06:42.720 --> 0:06:46.039 Review that today ten percent of reddits revenues actually come 0:06:46.160 --> 0:06:49.799 from selling its user generated content as training data to llms, 0:06:50.400 --> 0:06:51.839 so you know that there is. 0:06:51.920 --> 0:06:53.679 Just like our friends of the Atlanta though there's. 0:06:53.520 --> 0:06:56.240 Gold in them hills. This is actually the second time 0:06:56.320 --> 0:06:58.920 this study has been published in the Harvard Business Review. 0:06:58.920 --> 0:07:01.120 It's published in twenty twenty five as well. What I 0:07:01.160 --> 0:07:05.239 found really really interesting was what change from twenty twenty 0:07:05.279 --> 0:07:08.840 four to twenty twenty five. Thirty eight new use cases 0:07:08.880 --> 0:07:11.600 have been added to the list, and last year's top 0:07:11.720 --> 0:07:15.320 use case was generating ideas, which has now fallen down 0:07:15.360 --> 0:07:18.400 to sixth place. This year, the second and third most 0:07:18.480 --> 0:07:21.760 popular use cases were new entrants to the list. They 0:07:21.800 --> 0:07:24.080 weren't in the top hundred mentions last year and now 0:07:24.080 --> 0:07:28.000 they're in second and third place. They were organizing my life, 0:07:28.400 --> 0:07:30.400 followed by finding purpose. 0:07:31.000 --> 0:07:33.040 You know, I think this speaks to something which is 0:07:33.040 --> 0:07:36.280 that people are lonely, and people don't know how to 0:07:36.320 --> 0:07:39.240 talk to other real people about these things. And I think, 0:07:39.440 --> 0:07:42.000 sort of like tinders the game game right that we 0:07:42.080 --> 0:07:46.880 did last week, it's much easier to kind of test 0:07:47.000 --> 0:07:51.320 these very human interactions on computers. 0:07:50.840 --> 0:07:53.320 Absolutely or in a computer no judgment judgment free zone. 0:07:53.440 --> 0:07:55.239 The other thing that I would mention is that regular 0:07:55.280 --> 0:07:57.720 therapy the kind that you have with the human therapist 0:07:58.160 --> 0:08:03.080 will put your bank account into negative numbers, absolutely, And 0:08:03.160 --> 0:08:06.800 so you may not want to constantly burden your friends 0:08:06.800 --> 0:08:10.240 with your issues and be the carry Bradshaw in your 0:08:10.240 --> 0:08:12.320 friend group. And so I can see why a person 0:08:12.360 --> 0:08:14.920 would turn to AI to help them through a hard time. 0:08:15.320 --> 0:08:16.840 I actually don't judge it at all. 0:08:17.160 --> 0:08:19.960 No, no, me neither. But on the contrary, I mean, 0:08:20.400 --> 0:08:23.400 I think what the HBr sort of pointed out as 0:08:23.400 --> 0:08:25.640 the kind of larger takeaway in the twenty twenty four 0:08:25.720 --> 0:08:29.600 versus twenty twenty five comparison was that in twenty twenty four, 0:08:29.920 --> 0:08:32.680 the most popular use cases were all around quote unquote 0:08:32.720 --> 0:08:36.640 technical assistance and troubleshooting, whereas this year they're in quote 0:08:36.679 --> 0:08:41.000 unquote personal and professional support, which kind of mirrors what 0:08:41.040 --> 0:08:43.000 we were talking about just a few minutes ago, which 0:08:43.040 --> 0:08:46.880 is how AI has become more ubiquitous and therefore like 0:08:47.120 --> 0:08:49.560 normal people are using it for more normal reasons. 0:08:49.840 --> 0:08:52.600 I think, more ubiquitous and also like imbued with our 0:08:52.600 --> 0:08:55.240 own humanity as opposed to like a place where we 0:08:55.320 --> 0:08:57.280 find how do I fix this thing? 0:08:57.480 --> 0:08:57.600 Right? 0:08:58.040 --> 0:08:59.920 I think it's changed from how do I fix this thing? Too? 0:09:00.160 --> 0:09:04.000 How do I fix myself, which to me is both 0:09:04.080 --> 0:09:06.240 a little bit scary and also a little bit exciting. 0:09:06.559 --> 0:09:10.600 Absolutely, there is one really important exception to what we've 0:09:10.600 --> 0:09:14.200 been talking about just now, which is that two use cases, 0:09:14.679 --> 0:09:20.080 firstly generating code and secondly improving code, both had meteoric rises. 0:09:20.440 --> 0:09:24.040 So generating code was down in lowly forty seventh place 0:09:24.200 --> 0:09:27.080 on last year's list, is now in fifth place. And 0:09:27.120 --> 0:09:30.040 that brings me to my next headline, which is by Bloomberg. 0:09:30.160 --> 0:09:35.400 Under the headline AI Coding Assistant, Cursor draws a million 0:09:35.480 --> 0:09:38.760 users without even trying. So there's a hot new AI 0:09:38.800 --> 0:09:41.000 start up in town, one that you've probably never heard of, 0:09:41.120 --> 0:09:43.320 because I hadn't heard of it either until now. Its 0:09:43.320 --> 0:09:47.240 Bloomberg story. But also they don't advertise or market. Bloomberg 0:09:47.320 --> 0:09:51.160 reports that Cursor hasn't spent a single dollar on paid marketing. 0:09:51.720 --> 0:09:54.800 The startup behind it is any Sphere, Inc. And they 0:09:54.840 --> 0:09:57.199 make this AI powered coding editor called Cursor. 0:09:57.559 --> 0:10:01.200 Yeah, it's this popular tool for both formally trained computer 0:10:01.280 --> 0:10:05.560 programs and this thing that I love, Vibe coders. A 0:10:05.600 --> 0:10:07.840 coding editor, from what I understand, is a program that 0:10:07.880 --> 0:10:10.720 does things like check your code for errors. Think of 0:10:10.720 --> 0:10:13.560 it like spell check for coding, and then on top 0:10:13.640 --> 0:10:17.520 of that, more recent AI coders like Cursor can also 0:10:17.640 --> 0:10:20.360 suggest lines of code for you based on what you've 0:10:20.360 --> 0:10:21.240 previously written. 0:10:21.400 --> 0:10:23.520 So Gen one is like spell check and Gen two 0:10:23.679 --> 0:10:26.400 is shifted response for emails. 0:10:25.920 --> 0:10:28.840 Exactly, And that's what vibe coding actually is. You just 0:10:28.880 --> 0:10:32.600 like accept all of the suggestions versus having to correct 0:10:32.600 --> 0:10:35.439 your code, sort of like writing a whole email and autocomplete, 0:10:35.440 --> 0:10:38.240 which I do when I send an email says thank you, Comma, 0:10:38.640 --> 0:10:41.840 Jennifer exclamation point. That is an email that's written entirely 0:10:41.840 --> 0:10:44.600 by autocomplete. And just to be clear, as of today, 0:10:44.720 --> 0:10:47.560 vibe coding does not generate code at the level of 0:10:47.559 --> 0:10:48.120 real coding. 0:10:48.320 --> 0:10:51.640 Nonetheless, Cursor has quietly become one of the fastest growing 0:10:51.640 --> 0:10:55.559 startups of all time. It's even used by programmers at 0:10:55.600 --> 0:10:59.600 companies like OpenAI, Instacra, and Uber. Although most of the 0:10:59.600 --> 0:11:03.280 revenue you comes from individuals people using it and not 0:11:03.360 --> 0:11:05.960 using it through corporate descriptions. They're getting their own subscription 0:11:06.080 --> 0:11:08.440 to help them with their work. As it turns out, 0:11:08.520 --> 0:11:10.680 coders are wanting to pay cash for a good user 0:11:10.679 --> 0:11:13.959 interface and adaptability. The passion is real for the product 0:11:13.960 --> 0:11:17.000 back In January, any Sphere reached one hundred million in 0:11:17.080 --> 0:11:20.760 annual recurring revenue. By March that number of doubled and 0:11:20.840 --> 0:11:23.640 over a million people are now using Cursor every single day. 0:11:24.000 --> 0:11:26.760 I mean, we've talked about this many, many times that 0:11:26.800 --> 0:11:29.880 there's a lot of anxiety about AI replacing people's jobs. 0:11:30.080 --> 0:11:33.080 Cursor is actually interesting because it's being used to help 0:11:33.120 --> 0:11:36.640 people with their jobs. It's a productivity tool, and it's 0:11:36.720 --> 0:11:39.320 what these AI companies have been parroting all along. 0:11:39.440 --> 0:11:41.760 It's an interesting paradigm thing here where it's not like 0:11:42.080 --> 0:11:43.880 here's a new product to make your redundant. It's like, 0:11:44.040 --> 0:11:46.040 here's a new product that makes you better so much 0:11:46.080 --> 0:11:47.920 so that you pay your own money to use it 0:11:47.960 --> 0:11:49.320 to make you better at your job. I mean that 0:11:49.400 --> 0:11:51.720 is I think, sort of a high watermark for what 0:11:51.760 --> 0:11:54.000 we can hope for from AI use cases. 0:11:54.120 --> 0:11:56.360 Yeah, and it's actually a concept that investors are falling 0:11:56.400 --> 0:11:58.560 in love with. Any Sphere, which is, as we said, 0:11:58.559 --> 0:12:01.560 the parent company has raised one hundred and seventy five 0:12:01.679 --> 0:12:04.400 million dollars in the likes of Andrees and Horowitz as 0:12:04.440 --> 0:12:07.520 well as one of open AI's co founders, among many others. 0:12:07.720 --> 0:12:10.880 Any Spheres and talks to raise more money at evaluation 0:12:10.960 --> 0:12:12.640 of nearly ten billion. 0:12:13.080 --> 0:12:16.000 Yeah, I mean it's very competitive space. Open AI and 0:12:16.160 --> 0:12:20.960 Anthropic both seem to be eyeing the arena for their 0:12:21.040 --> 0:12:25.200 own AI for coding tools, but nothing beats user adoption. 0:12:25.320 --> 0:12:28.800 So we'll see how this story unfolds. The CURSA. We're 0:12:28.840 --> 0:12:30.880 going to take a quick break now. When we come back, 0:12:31.480 --> 0:12:44.040 carasches a rather surreal headline. Stay with us. 0:12:47.480 --> 0:12:50.719 So as we're back, and I'm warning you that our 0:12:50.760 --> 0:12:52.760 next story is something that you're really gonna love. 0:12:52.840 --> 0:12:54.160 Okay, tell me. So. 0:12:54.600 --> 0:12:58.320 Google has a generative video model called VO two, and 0:12:58.360 --> 0:13:00.440 the way it works is you type in a prompt 0:13:00.559 --> 0:13:03.959 or feed it some media like an image, and VO 0:13:03.960 --> 0:13:07.200 two will generate a video based on these materials. So 0:13:07.320 --> 0:13:10.160 late last year, Google announced that they were launching a 0:13:10.200 --> 0:13:13.439 new version of the model, which includes a better grasp 0:13:13.480 --> 0:13:18.520 of physics and more quote nuances of human movement and expression. 0:13:18.720 --> 0:13:22.480 When a tech company tries to capture the nuance of humanity, 0:13:22.520 --> 0:13:24.560 it's always you always know there's going to be some 0:13:24.640 --> 0:13:26.240 excitements in store, yet. 0:13:26.080 --> 0:13:28.120 They still try to do it. The news is that 0:13:28.240 --> 0:13:31.480 Google recently showed off the updated VO two at the 0:13:31.559 --> 0:13:34.560 Google Cloud next conference, which we were not invited. 0:13:34.720 --> 0:13:37.600 Two our invitations must have got lost in the mail. 0:13:37.640 --> 0:13:39.640 And maybe they could have made a bigger splash if 0:13:39.640 --> 0:13:42.199 they invited us. But they did want to make a splash, 0:13:42.240 --> 0:13:46.160 so they decided to appeal to both you and the 0:13:46.280 --> 0:13:47.320 art community. 0:13:46.960 --> 0:13:49.680 High risk bet the art community and not notoriously such 0:13:49.720 --> 0:13:51.400 big fans of the big tech world. 0:13:52.080 --> 0:13:53.360 Are you going to ask me what they show? 0:13:53.520 --> 0:13:54.520 Yes? What did they show? 0:13:54.800 --> 0:13:59.880 So they showed a trailer based on Salvador Dolly's unmade screenplay. 0:13:59.360 --> 0:14:02.679 Salvador Dolly's unmade screenplay? What's that about? 0:14:02.840 --> 0:14:05.920 So let me set the scene. I was not there, 0:14:05.920 --> 0:14:09.800 but it's nineteen thirty seven. You're Salvador Dolly, you're in 0:14:09.840 --> 0:14:13.360 your early thirties, you're my age, and you were hanging 0:14:13.360 --> 0:14:15.880 out with Harpo Marx, one of the Marx brothers, who 0:14:15.920 --> 0:14:16.440 I adore. 0:14:16.520 --> 0:14:18.360 This is a great collection of characters, and it's like, 0:14:18.600 --> 0:14:20.200 you can make it up that these people were hanging 0:14:20.200 --> 0:14:22.000 out together and thinking how do we make a movie? 0:14:22.080 --> 0:14:24.520 And my favorite character, who's of course the naysayer, also 0:14:24.560 --> 0:14:28.479 factors into this. So you, as Dolly, write a fantastical 0:14:28.520 --> 0:14:30.960 screenplay for the Marx brothers to star in. But when 0:14:30.960 --> 0:14:34.800 you bring it to MGM, Louis B. Mayor of MGM 0:14:35.200 --> 0:14:37.840 kills the project and the screenplay is never realized as 0:14:37.840 --> 0:14:38.280 a film. 0:14:38.840 --> 0:14:40.880 Was there a reason for spiking it? 0:14:41.240 --> 0:14:44.800 Well, in a world where now tent pole films are 0:14:44.840 --> 0:14:48.080 very important, and then being able to sell a film 0:14:48.200 --> 0:14:52.160 was very important. Pitching a film called Giraffes on Horseback 0:14:52.320 --> 0:14:55.800 Salad is not exactly something that they think jumps off 0:14:55.840 --> 0:14:56.520 the screen. 0:14:56.240 --> 0:14:58.880 The biting Dali and the Marx brothers hands off for 0:14:58.920 --> 0:14:59.240 this one. 0:14:59.320 --> 0:15:01.960 They weren't umping at the salad fit for I'm sorry, 0:15:01.960 --> 0:15:04.800 I couldn't help myself. But the movie itself was a 0:15:05.360 --> 0:15:08.440 kind of broadly about an aristocratic man who falls in 0:15:08.480 --> 0:15:12.040 love with a woman from a world where dreams are reality, 0:15:12.360 --> 0:15:16.240 and it's so surreal it was likely unfilmable at the time. 0:15:17.080 --> 0:15:19.080 The other important piece I think that kind of made 0:15:19.080 --> 0:15:21.000 me laugh is that some of the Marx brothers like 0:15:21.120 --> 0:15:22.280 didn't even think it was funny. 0:15:22.280 --> 0:15:26.440 They were like, eh, so it never got made, no until. 0:15:26.240 --> 0:15:29.400 Now, until now, Google is trying to at least right now, 0:15:29.400 --> 0:15:32.280 it's just a trailer because this is a big feat. Uh, 0:15:32.520 --> 0:15:34.920 it's a trailer that is produced by the Dolli Museum 0:15:35.200 --> 0:15:38.160 and Gooldbee Silverstein and partners. And I'm gonna play you 0:15:38.320 --> 0:15:40.720 just a little sample of the trailer, and then you're 0:15:40.720 --> 0:15:43.400 gonna tell me if you think this is true to 0:15:43.480 --> 0:15:45.120 one of your I don't know if he's one of 0:15:45.120 --> 0:15:48.280 your favorite artists, but you do like him Catalan culture. 0:15:48.440 --> 0:15:49.640 So let's see. 0:15:51.080 --> 0:15:55.280 Let me tell you how the strangest movie never made 0:15:56.440 --> 0:15:59.360 the world wasn't ready for into No. 0:16:00.800 --> 0:16:01.480 I called it. 0:16:03.280 --> 0:16:12.120 Horseback Solid. Prepare yourself because the impossible is coming to life. 0:16:13.560 --> 0:16:14.920 Tell me a little bit about what you saw. 0:16:15.040 --> 0:16:19.360 Well, there's a giraffe on fire. There is obviously Salvador 0:16:19.520 --> 0:16:23.600 Daly's voice recreated. It does have the kind of vibe 0:16:23.840 --> 0:16:28.000 of a movie trailer. However, the esthetic is one that 0:16:28.120 --> 0:16:32.280 I think would make Salvador Dahi personally turn in his grave, 0:16:32.680 --> 0:16:36.200 because it is basically what you might imagine one of 0:16:36.240 --> 0:16:40.680 the camps at Burning Man might reinterpret Salvador Dali through 0:16:40.720 --> 0:16:43.520 the lens of It is so on the nose quote 0:16:43.600 --> 0:16:45.960 unquote surrealist, with people wearing like funny hats. 0:16:46.080 --> 0:16:48.640 It looks like clouds people on fire. 0:16:48.880 --> 0:16:51.560 It's got a very Burning Man vibe. 0:16:51.600 --> 0:16:54.640 It also has a very google commercial vibe, yes, which 0:16:54.680 --> 0:16:57.200 I think isn't the greatest thing ever, But yeah, I 0:16:57.200 --> 0:17:00.520 think fever dream is how I describe it. And actually, 0:17:00.600 --> 0:17:03.640 as you said, the narration is supposed to be Salvador, 0:17:07.119 --> 0:17:08.280 there's enough of his audience. 0:17:08.280 --> 0:17:10.560 Does he actually sound like that? Who knows? Did we 0:17:10.560 --> 0:17:11.640 did in fact shake his voice? 0:17:11.720 --> 0:17:13.480 We did? I did listen to it. Well, I did 0:17:13.520 --> 0:17:17.680 listen to him on radio program a little bit later, 0:17:18.000 --> 0:17:21.080 and he does sound a little bit like you know, 0:17:21.320 --> 0:17:23.000 But I always think about him twirling his. 0:17:23.080 --> 0:17:25.480 Mustache actually character in Rettaitui. 0:17:25.680 --> 0:17:29.240 There's no evidence to me that this is exactly what 0:17:29.280 --> 0:17:32.000 Salvador Dolly sounds like. But I think they definitely tried 0:17:32.000 --> 0:17:33.840 to recreate his voice, which we know by now is 0:17:33.920 --> 0:17:34.919 like what. 0:17:35.200 --> 0:17:36.960 They tried to do it. They probably did do it. Yeah. 0:17:37.320 --> 0:17:40.159 According to Art News, this isn't the first time that 0:17:40.200 --> 0:17:43.280 there's actually been an AI generated Dolly. The same Dolly 0:17:43.320 --> 0:17:46.560 museum who's co producing this movie, actually had an exhibition 0:17:46.680 --> 0:17:50.560 called Ask Dolly that allowed museum goers to talk to him, 0:17:50.920 --> 0:17:54.360 meaning to talk to an AI created based on his voice. 0:17:54.720 --> 0:17:57.560 At the time. The museum COO said that if these 0:17:57.560 --> 0:18:00.720 technologies had been around when Dolly was alive, he would 0:18:00.760 --> 0:18:01.720 have played around with them. 0:18:02.240 --> 0:18:06.880 My take was, it's not horrible, but it's pretty horrible. 0:18:07.160 --> 0:18:10.320 And I don't disagree that Darli would have played around 0:18:10.320 --> 0:18:14.520 with AI tools if he were around today, but I 0:18:14.520 --> 0:18:17.040 don't think he would have used them like this. I mean, 0:18:17.080 --> 0:18:19.480 the whole point of Dahali as a painter was that 0:18:19.560 --> 0:18:22.120 he made the medium new. He basically reinvented the media 0:18:22.119 --> 0:18:26.480 of paintings through his incredibly interesting and unforgettable serialus take. 0:18:26.600 --> 0:18:29.080 So I think if he did use Jennet of AI tools, 0:18:29.280 --> 0:18:33.439 he would come't with the considerably more interesting application than this, 0:18:33.480 --> 0:18:35.959 which ultimately feels a bit druotism. You would imagine if 0:18:36.000 --> 0:18:38.679 an artist like Darli used Jenet AI, he would have 0:18:38.720 --> 0:18:41.160 done so to critique or to push the medium forward, 0:18:41.280 --> 0:18:43.840 or to make us really think, versus to kind of, 0:18:44.600 --> 0:18:45.040 you know. 0:18:45.440 --> 0:18:48.200 Make a gimmick, make a gimmick. Yeah, that's very true. 0:18:48.480 --> 0:18:49.920 Can we run through the rest of the headlines? 0:18:50.000 --> 0:18:54.160 Yes? And can you start yes? So Ours Technica reports 0:18:54.160 --> 0:18:57.399 that Nvidia is producing AI chips for the first time 0:18:57.680 --> 0:19:01.080 outside of Taiwan. Blackwell chip and are now being made 0:19:01.119 --> 0:19:06.560 at TSMC. That's the Taiwan semiconductor manufacturing company at their 0:19:06.560 --> 0:19:10.560 plant in Phoenix, Arizona, and other companies in the state 0:19:10.640 --> 0:19:13.920 will test and package these chips. Nvidia also announced that 0:19:13.960 --> 0:19:18.560 they plan to build complete supercomputers on us soil, and 0:19:18.600 --> 0:19:20.720 their reports of the promise of up to five hundred 0:19:20.800 --> 0:19:24.760 billion dollars of investment in us AI infrastructure having been 0:19:24.760 --> 0:19:28.439 agreed to at a dinner between Jensen Huang and you 0:19:28.560 --> 0:19:33.680 know who recently at Mari Lago. In return, Kwang reportedly 0:19:33.720 --> 0:19:37.760 hoped to avoid yet more stringent export controls on chips 0:19:37.800 --> 0:19:41.360 to China, but seemly that hope has not materialized. 0:19:41.440 --> 0:19:43.040 I do wonder what's going to happen there, so please 0:19:43.119 --> 0:19:45.119 keep me posted. I also do wonder if the food 0:19:45.200 --> 0:19:46.320 is good at mar Laga. 0:19:46.440 --> 0:19:48.119 Yeah, I bet it's old school. 0:19:48.560 --> 0:19:53.120 Well, it's food that has to benefit a discussion about semiconductors, 0:19:53.400 --> 0:19:56.359 so it can't stand out too much. So if you 0:19:56.400 --> 0:20:00.000 are a big enough nerd to be covetous of Sam's 0:20:00.000 --> 0:20:04.960 Sung technology, but you're loyal to the Apple ecosystems, it's 0:20:04.960 --> 0:20:08.880 a very it's slender. There may be a phone for you. 0:20:09.240 --> 0:20:11.640 Nine to five Mac reports that there are rumors of 0:20:12.359 --> 0:20:17.280 and this is very mind blowing a foldable iPhone, which 0:20:17.280 --> 0:20:22.560 has been dubbed very surprisingly the iPhone fold be like 0:20:22.600 --> 0:20:27.120 I need the fold pro. I have to say this, 0:20:27.560 --> 0:20:30.000 and I hope it's not too embarrassing. If I do 0:20:30.080 --> 0:20:32.800 get this, I will be walking around saying my iPhone, 0:20:32.800 --> 0:20:34.760 don't jiggle, jiggle it folds. 0:20:35.520 --> 0:20:35.720 But no. 0:20:35.840 --> 0:20:38.800 According to the one report, the rumored phone may look 0:20:38.840 --> 0:20:41.720