The AI Daily Brief: Artificial Intelligence News and Analysis - AI Video Comes to Hollywood and Youtube; Controversy Ensues

Starting point is 00:00:00 Today on the AI Daily Brief, AI Video is coming to Hollywood and to YouTube, and before that in the headlines, an uproar around LinkedIn's new AI data training policies. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI. To join the conversation, follow the Discord link in our show notes. Welcome back to the AI Daily Brief Headlines edition, all the daily AI news you need in around five minutes. We kick off today with a controversy that on the one hand I understand, but on the other hand, I am completely not surprised by, I guess. The Verge writes LinkedIn is training AI models on your data. You'll need to opt out twice to stop LinkedIn

Starting point is 00:00:40 from using your account data for training in the future. But anything already done is done. They continue, if you're on LinkedIn, then you should know that the social network has, without asking, opted accounts into training generative AI models. LinkedIn apparently introduced the new privacy setting and opt-out form before rolling out an updated privacy policy. The policy said, we may use your personal data to improve develop and provide products and services, develop and train artificial intelligence models, develop, provide, and personalize our services, and gain insights with the help of AI automated systems and inferences so that our services can be more relevant and useful to you and others. So if you are someone who wants to revoke permissions for this,

Starting point is 00:01:13 which is completely reasonable, if you go to the data privacy tab in the account settings and click on data for generative AI improvement, there will be a toggle that you can turn to off to opt out. Now, that is specifically for training generative AI models. However, if you also want LinkedIn to turn off training for its personalization and moderation and other things that don't generate content, you have to fill out something called the LinkedIn data processing objection form. And like I said, on the one hand, people reasonably were really upset about this. Rachel Tobac writes, LinkedIn is now using everyone's content to train their AI tool. They just auto-opted everyone in. I recommend opting out now and that orgs put an end to auto-opt-in. It's not

Starting point is 00:01:49 cool. Theo writes, LinkedIn has opted you in by default for generative AI and is given no word or update regarding its data policies beforehand. Turn the damn thing off in the data privacy section ASAP. Microsoft is a crock of poo. Now, like I said, on the one hand, I completely understand where people are coming from being frustrated about being auto-opted into something trained on their data. It's an inherently violating sort of feeling. At the same time, this has long been what it means to be on any sort of social network. It's just now they're using your data for something different. I don't know, maybe I'm cynical as someone who's spent the last half decade or more in the crypto industry, which has tried to sell people on caring about

Starting point is 00:02:23 things like data privacy and data ownership. Completely, I will add, unsuccessfully. But I'm a little surprised, and probably LinkedIn is surprised as well, that people are this up in arms. Now, I would love to see consumer pressure actually push people to not auto-opt-in and to reward companies that don't, but I am currently not holding my breath for that. Next up, follow-up to a story that we discussed earlier this week. While Gavin Newsom might be quote-unquote concerned about the implications of SB 1047, he did sign some AI-related laws specifically around election deepfakes. That led to, Elon Musk, very aggressively posting an AI parody video of the Kamala Harris campaign and basically begging California to sue him. The AP writes this morning, California laws cracking down on election

Starting point is 00:03:03 deepfakes by AI face legal challenges. AP writes, two of the three laws, including one that was designed to curb the practice in the 2024 election, are being challenged in court through a lawsuit filed Tuesday in Sacramento. The lawsuit filed by a person who created parody videos featuring altered audios of vice president and Democratic presidential nominee Kamala Harris, says the laws censor free speech and allow anybody to take legal action over content they dislike. The person who filed the lawsuit was the person who created the video that was shared by Elon Musk. This, like so many legal questions surrounding AI, is 100% I believe destined to end up in front of the Supreme Court. And this one might get there faster, given that it is a free speech and First Amendment question,

Starting point is 00:03:41 which tends to be fairly sacrosanct here in America. On the AI safety front, the US has announced that they will convene a global AI safety summit coming up in November. On November 20th and 21st, Commerce Secretary Gina Raimondo and Secretary of State Anthony Blinken will host the first meeting of the International Network of AI Safety Institutes in San Francisco. The network members include groups from Australia, Canada, the European Union, France, Japan, Kenya, South Korea, Singapore, Britain, and the United States. Ramando had previously announced the launch of the International Network of AI Safety Institutes back in May at the AI Soul Summit.

Starting point is 00:04:12 Ramando said the aim is, quote, close, thoughtful coordination with our allies and like-minded partners. Reuters writes that the San Francisco meeting will include technical experts from each member's A.I. Safety Institute to discuss priority work areas and advanced global collaboration and knowledge sharing on AI safety, which is great because after Governor Newsom vetoes or signs into law, SB 1047, will need another AI safety-related thing to debate. Building off the theme of Newsom's concern of the potential chilling effect that SB 1047 could have, META has warned the European Union in an open letter that the bloc's regulations risk hampering innovation and economic growth. The letter was also signed by Spotify,

Starting point is 00:04:46 Prada, and others. If you are a regular listener to this show, you'll know that meta is not releasing its latest models, at least not the multimodal versions in Europe, for fears of some of these concerns. And Apple is also backing off of putting Apple intelligence on European iPhones. This means that this is no longer theoretical, and the companies are actually making decisions to remove key features from the European market, which, while perhaps protecting Europeans, also denies them access to state-of-the-art technology. As always, big societal questions that surround AI, some of the most interesting for us to discuss here. For now, though, that is going to do it for today's idea. Brief Headlines edition. Next up, the main episode.

Starting point is 00:05:24 Today's episode is brought to you by Fractional. When we wanted to build an AI-powered feature of Superintelligent, our AI tool finder, I went straight to Fractional. The Fractional team is a group of senior engineers in San Francisco working on some of the most exciting projects in applied AI. Working with them is basically like hiring an absolute top-flight AI engineering team, but in a way that you can customize exactly for your particular needs. Like I said, that AI tool finder feature that we built with them is already a key part of the super intelligent platform and we are working on something new as well. Fractional works with everyone from startups to the Fortune 500 and to request a free consultation you can go to fractional.aI. If you want help

Starting point is 00:06:03 identifying and building AI projects for your business, then I highly recommend you hit pause, open a web browser and go to fractional.aI to request a free consultation. Today's episode is brought to you by Venice. The leading AI company store your entire conversation history and attach it to your identity forever. That's every question you ask, every answer you receive, every image you generate, every thought you share with the machine it's all being spied on. If you trust all the company's hackers and NSA board members that will ever have access to your AI conversations, then rejoice, for you are well served. For the rest of us, Venice is an alternative. Venice is a powerful AI app for text, image, and code generation that respects you as a sovereign

Starting point is 00:06:41 individual and believes privacy and free speech are not only human rights but necessary for civilizational advancement. Private, permissionless, and uncensored, you can try it for free without an account. AIA Daily Brief listeners receive a 20% discount on Venice Pro. Visit venice.a. slash NLW.LW. Daily Brief. That's NLW Daily Brief. All one word. Today's episode is brought to you by Superintelligent, which is, of course, our platform that helps you learn how to use AI tools and perhaps even more importantly, gives you ideas on the best use cases that are actually going to help you achieve whatever it is you want to achieve. To recognize the end of summer and back to school slash back to work, we are running our best

Starting point is 00:07:25 promotion ever when you sign up for Super Intelligent between now and the end of August, using code so back, your first month will be 100% free. The platform features over 600 fun, highly practical AI tutorials that get you using AI fast and with an eye to actually transforming how you get things done. We've just launched Super for Teams, so if you have a group of people at your company that want to figure out how to use AI together, I highly suggest you check it out. But for those of you who are using Super Intelligent as an individual, once again, if you sign up for Super Intelligent between now and the end of the month using code so back, you will get your first month 100% free. Go to B-Supert.a.i and check it out today. Welcome back to the AI Daily Brief.

Starting point is 00:08:09 Ever since February, when OpenAI announced SORA, there has been a ton of it. of attention in AI land on the increasing capacity to create AI video. Now, of course, the story ever since then has not been about SORA, but about all the other models that have been released in the subsequent time. We got Google's VEO, Luma Lab's Dream Machine, plus new models from PICA, and the latest models from runway. Point is, there has been a huge amount of advancement in the generative video space, and it seemed inevitable that it's only been a matter of time before it would come and infiltrate the other areas where videos live. Well, that is now happening in a big way. Yesterday, Runway announced what they are calling a first-of-a-kind partnership with Lionsgate.

Starting point is 00:08:49 In short, they're partnering with Lionsgate to bring AI into the filmmaking process. The core of the deal is that Runway will be creating and training a new model that is customized on Lionsgate's proprietary catalog. They write, fundamentally designed to help Liongate Studios, its filmmakers, directors, and other creative talent augment their work. The model generates cinematic video that can be further iterated using Runway's suite of controllable tools. Lionsgate Vice Chair Michael Burns said we view AI as a great tool for augmenting, enhancing, and supplementing our current operations. Runway CEO, Cristobal Valenzuela, said, The History of Art is the History of Technology, and these new models are part of our

Starting point is 00:09:24 continuous efforts to build transformative mediums for artistic and creative expression. Now, some things haven't been answered about this partnership yet. Specifically, will the new model incorporate other training data, or is it going to be exclusively trained on Lionsgate? Nor is there exact information on how Lionsgate intends to deploy this model. Still, it's a clearly notable deal. It was headline news and variety in the Hollywood reporter, and in additional interviews, we got a little bit more about what Lionsgate is looking to get out of it.

Starting point is 00:09:50 Burns again told the Wall Street Journal that Lionsgate expects to save, quote, millions and millions of dollars. He also said, we do a lot of action movies, so we blow a lot of things up, and that is one of the things runway does. Now, of course, AI in the movie and television industry is an extremely controversial area. It was one of the first places last year where we got actual strikes relating to AI, and there remains a ton of hostility, scorn, and major questions around how AI is going to impact the industry. A lot of that was on display in the commentary around this on X.

Starting point is 00:10:20 Neil Tirkowitz called it unsavory news and said this is nothing to celebrate. Zachary Berger, who bills himself as the lead creature designer on Avatar The Way of Water and Avatar sequels, said that International Alliance of Theatrical Stage employees members, quote, like myself, were told by leadership that the reason there were major concessions to the studios on the AI portion of the recent contract deal was because the studios were so scared of the copyright implications of the technology. At the time, I and others called that out for being extremely short-sighted and suggests that there was nothing stopping the studios from creating their own proprietary AI models. And here we are only a few months later.

Starting point is 00:10:51 We know that Runway has trained their AI products on copyrighted content. So even if Runway is providing Lionsgate a proprietary model trained off of Lionsgate properties, it is incredibly likely that it is built off a model with his other stolen content in the mix. Additionally, Berger points out Runway is still being sued for potential copyright infringement. and ultimately he says, this sets a dangerous precedent of allowing a film studio to ask forgiveness, not permission of using stolen copyrighted data in an AI product. We shouldn't let this go on challenged. Union should be pushing back on this in the press now. We don't want to send a signal to other major film studios that this is okay. Now, holding aside, the studio employee impressions of this,

Starting point is 00:11:27 others just didn't think it was going to produce good work. Gail on X writes, Lionsgate partnering with runway to integrate generative AI into film production is incredibly lame. Again, I'm not going to watch your productions if you can't bother producing it yourself. Expect no money from my pockets, thank you. We'll watch movies made by people. For what it's worth, even if this isn't the opinion that I happen to take, I think that voting with your dollars is one of the strongest ways that people can send signals about things that they don't like. Still, I think to many, this feels completely inevitable. It is inevitable, yes, on the one hand, that businesses are, of course, going to look for cost savings, particularly

Starting point is 00:11:59 an incredibly beleaguered industry like Hollywood is right now. At the same time, I personally don't think it's BS that new technologies have always created new mediums for art. I think that if you spend any time on AI Twitter, there is incredible stuff being created with AI video that's going to influence a whole new generation of filmmakers without a doubt. I think the questions that are being brought up by those in the industry are important, and I certainly am going to continue to follow how this all plays out. In the meantime, though, like I said, Hollywood wasn't the only place AI video was showing up. Google DeepMind also tweeted yesterday, our most advanced generative video model Vio is coming to YouTube, shorts to help creators bring their ideas to life. In an expansion post called

Starting point is 00:12:38 empowering YouTube creators with generative AI, Google writes that new video generation technology and YouTube shorts will help millions of people realize their creative vision. So what is actually coming? Well, there are a couple things. With the Dreamscreen feature, YouTube creators will be able to generate backgrounds for their shorts. Starting with an initial text prompt they write, Dreamscreen uses Image in 3 to generate four different images. Creators pick an image in their preferred style, composition, or aesthetic from these options, and with the selected image, Vio generates a high-quality six-second background video. However, the bigger deal is this throwaway line, at least in the blog post, but one that everyone picked up on, early next year we're also making

Starting point is 00:13:12 it possible to generate six-second standalone video clips. The Verge captured, I think, what is the most common sense that I saw. There are some exciting possibilities for what could happen when creators have an easier time making new things, but it's also possible that YouTube is about to be flooded with AI-conceived, AI-written, and even AI-produced videos that I'll look and sound and feel kind of the same. Most of these new features can be useful tools or shortcuts to slop creation, and each creator will have to decide what they want them to be. However, they conclude, from YouTube's perspective, the company has spent the last few years trying to lower the bar to becoming a YouTube creator, particularly through shorts, as it tries to compete with TikTok

Starting point is 00:13:45 and Instagram and countless other places people make things now. It seems confident that AI can make practically every part of a creator's job easier, and maybe even get them to create even more. Daniel Battal summed up the exciting side of this saying, imagine opening up YouTube studio and having AI give you ideas for your next video based on what it knows about your channel. Now imagine it generating thumbnail ideas as well. Mark Egan points out, it looks like we soon won't have to look for AI tools. The tools will come to us. Still, comedian and commentator Hank Green summed up a lot of the other side of this

Starting point is 00:14:15 when he shared a wired article called an avalanche of generative AI videos is coming to YouTube shorts, adding his caption, look, I'm having a bad morning and so nuance is hard for me right now, but this is stupid. And to be fair to the This is Stupid Crowd, there are two different reasons that people are skeptical of this. One is all of their qualms and questions about the ethics of AI creation. But hold aside that. I think there's a general belief that this is likely to just decrease the average overall

Starting point is 00:14:38 quality of videos on that channel. Now, of course, for that, we'll have to wait and see. And there are theoretically other ways to use AI to also figure out what's good and what people are watching. But whether we like it or not, this is happening. Indeed, if you look, video is everywhere. This morning, Amazon released a video generator, but for ads. Alibaba also introduced a new text-to-video model.

Starting point is 00:14:58 One of the more interesting questions, honestly, with all of this, is where the hell is OpenAI? Now, admittedly, they just did release a new model, which has the potential to fundamentally change how we think about creating new models, so it's not like they've been sloughing off. But as Andrew Curran puts it, I thought Sora would get here first.

Starting point is 00:15:15 I think that was the sense that many people had. Indeed, when Sora hadn't been released for so long, it's kind of been assumed that that was because they were focusing on a sort of Hollywood use case rather than a general consumer use case. Are we then going to see the first SORA deal announced in response to this in the next week or so? It wouldn't be out of character. And if we do, of course, you will hear about it here first.

Starting point is 00:15:35 For now, that is going to do it for today's AI Daily Brief. Until next time, peace.

The AI Daily Brief: Artificial Intelligence News and Analysis - AI Video Comes to Hollywood and Youtube; Controversy Ensues

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.