The Best Take on AI and Hollywood is from Ben Affleck

0:00Today on the AI Daily Brief, Ben Affleck's impressive and viral comments on AI in Hollywood. Before that, in the headlines, 11 labs is now allowing users to build conversational AI agents. 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.

0:27Welcome 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 new conversational agent launch from Eleven Labs. The voice cloning and text-to-speech platform now offers the ability to build conversational AI bots. Users can customize their own bots, tweaking settings like tone of voice and response length. The company's head of growth, Sam Sklar, told reporters that customers were already trying to use the platform to build agents like this, but had been running into problems around things like knowledge-based integration and handling of interruptions, leading Eleven Labs to decide to build a full pipeline to support this use case.

1:02Now, devs can plug in their choice of LLM to drive the conversation and set limits around creativity and token use. They can also tune voice, latency, stability, authentication criteria, and maximum length of conversation. The agents also have multi-language support to help engage a global customer base. Basically, the big thing here is not any of the specifics. There's nothing super novel. It's the fact that this is just becoming increasingly normalized. And it really is one more signal that the AI agent era is now truly upon us. Next up, Mistral have updated their product suite in an effort to keep up with the larger frontier labs.

1:33Mistral's chatbot, LeChat, now has inline citations as well as a canvas tool to allow users to modify AI outputs. Mistral writes in an announcement blog post, You can use the canvas features to create documents, presentation, code, mockups, the list goes on. You're able to modify its contents in place without regenerating responses, version your drafts, and preview your designs. In addition, LeChat can now ingest large PDF documents and images as inputs. The platform leverages Black Forest Labs Flux Pro model for image generation. LeChat can also now host shareable automated workflows for repetitive tasks, like scanning expense reports and invoice processing, and indeed, Mistral is referring to these workflows as agents.

2:10All in all, these new features bring LeChat basically in line with OpenAI's ChatGPT and Anthropics Cloud. The new features come along with an update to the model range. Pixtro Large is the second model in Mistral's image-capable LLM line. The model clocks in at 124 billion parameters. which is a little smaller than Anthropics Cloud 3 Sonnet and nowhere near the size of Meta's largest Lama 3-400B model, or of course the rumored trillions of parameters for GPT-4-0. Mistral said in their announcement blog post, Mistral Large is able to understand documents, charts, and natural images. The model demonstrates frontier-level image understanding.

2:43Mistral also unveiled a new version of their text-only model, Mistral Large. They said it brings notable improvements in long-context understanding, making it well-suited for use cases like document analysis and task automation. Now, Mistral is one of a very small handful of companies that are credible frontier model developers. They are, along with Meta, holding the banner for open-source approaches, but they are, of course, in an extremely competitive space where resources really do make a difference. The company recently raised$640 million in venture funding, which is no small amount, but is still a small fraction of the billions of dollars raised by companies like OpenAI.

3:17And given that this isn't enough to train a competitive next-generation model, Mistral may end up answering the question of whether AI labs can do more with less by leveraging novel techniques. For their part, Mistral is explicitly taking a different path, writing, At Mistral, our approach to AI is different. We're not chasing AGI at all costs. Our mission is instead to place frontier AI in your hands so you get to decide what to do with advanced AI capabilities. This approach has allowed us to be quite frugal with our capital while consistently delivering frontier capabilities at affordable price points.

3:45So really interesting that you're starting to see this clear differentiation in terms of how they view their competitive set. Speaking of OpenAI, some interesting news on one of their licensing deals. Publisher Dot Dash Meredith, who is the publisher of People, Better Homes and Gardens, and InStyle, disclosed that they were being paid$16 million per year to license content. The multi-year partnership with OpenAI was announced back in May and is one of many that OpenAI has signed in the past year. Speaking of licensing, book publisher HarperCollins has asked its authors to opt into a licensing deal for AI training.

4:15One of the big five publishers, HarperCollins confirmed that they had struck a deal named AI Company. They said in a statement, HarperCollins has reached an agreement with an AI technology company to allow limited use of select nonfiction backlist titles for training AI models to improve model quality and performance. While we believe this deal is attractive, we respect the various views of our authors, and they have the choice to opt into the agreement or to pass on the opportunity. A screenshot of the communication was circulated this week by author Daniel Kibblesmith. It disclosed that the deal was$2 ,500 per book over a licensing period of three years.

4:46Kibblesmith shared his feelings when he captioned the post, a dominable. Interestingly, Kibblesmith took the chance to soapbox a little bit to AV Club and wrote, it seems like they think they're cooked and they're chasing short money while they can. I disagree. The fear of robots replacing authors is a false binary. I see it as the beginning of two diverging markets, readers who want to connect with other humans across time and space, or readers who are satisfied with a customized on-demand content pellet fed to them by the big computer so they never have to be challenged again. Now, while it's clear which side he comes down on, that sort of tale of two markets take is reminiscent of what we're going to get into in our main episode, which is, of course, now coming up as we are done here with the headlines.

5:23Stay tuned for that. Today's episode is brought to you by Plum. Want to use AI to automate your work but don't know where to start? Plum lets you create AI workflows by simply describing what you want. No coding or API keys required. Imagine typing out, AI, analyze my Zoom meetings and send me your insights in Notion, and watching it come to life before your eyes. Whether you're an operations leader, marketer, or even a non-technical founder, Plum gives you the power of AI without the technical hassle. Get instant access to top models like GPT-40, Claude Sonnet 3.5, Assembly AI, and many more.

5:54Don't let technology hold you back. Check out Use Plum, that's Plum with a B, for early access to the future of workflow automation. Today's episode is brought to you by Vanta. Whether you're starting or scaling your company's security program, demonstrating top-notch security practices and establishing trust is more important than ever. Vanta automates compliance for ISO 27001, SOC 2, GDPR, and leading AI frameworks like ISO 42001 and NIST AI risk management framework, saving you time and money while helping you build customer trust. Plus, you can streamline security reviews by automating questionnaires and demonstrating your security posture with a customer-facing trust center all powered by Vanta AI.

6:32Over 8 ,000 global companies like LangChain, Lila AI, and Factory AI use Vanta to demonstrate AI trust and prove security in real time. Learn more at vanta.com slash nlw. That's vanta.com slash nlw. Today's episode is brought to you by Superintelligent. Every single business workflow and function is being remade and reimagined with artificial intelligence. There is a huge challenge, however, of going from the potential of AI to actually capturing that value. And that gap is what Superintelligent is dedicated to filling. Superintelligent accelerates AI adoption and engagement to help teams actually use AI to increase productivity and drive business value.

7:12An interactive AI use case registry gives your company full visibility into how people are using artificial intelligence right now. Pair that with capabilities building content in the form of tutorials, learning paths, and a use case library, and Superintelligent helps people inside your company show how they're getting value out of AI while providing resources for people to put that inspiration into action. The next three teams that sign up with 100 or more seats are going to get free embedded consulting. That's a process by which our super intelligent team sits with your organization, figures out the specific use cases that matter most to you, and helps actually ensure support for adoption of those use cases to drive real value.

7:51Go to bsuper.ai to learn more about this AI enablement network. And now, back to the show. Welcome back to the AI Daily Brief. Today, we are talking about AI and Hollywood, although really it's AI and creativity more broadly. The conversation is specifically sparked by these comments from Ben Affleck on a CNBC show that have just lit the internet on fire for, surprisingly, given that it's the internet, their clarity and nuance. As you will see when we share the clip, Affleck is neither denying the impact of AI, nor is he fearfully warning about its implications. We're actually going to watch and listen to nearly the three and a half minute clip because it's so coherent.

8:31But first, let's talk a little bit of background. Hollywood has actually been one of the front lines in the struggle around AI. We got our first strikes against AI last year, where both writers and actors were concerned around how AI might be used to replace them. Last September, the AP wrote, after a 148-day strike, Hollywood screenwriters secured significant guardrails against the use of AI in one of the first major labor battles over generative AI in the workplace. During the nearly five-month walkout, no issue resonated more than the use of AI in script writing. What was once a seemingly lesser demand of the Writers Guild of America became an existential rallying cry.

9:07And although this strike ended as it did with the writers winning some protections, it felt much more like the beginning of something than it did like the end. There have, of course, also been dust-ups around training material. The Atlantic just this week published a piece there's no longer any doubt that Hollywood writing is powering AI. The analysis is part of the Atlantic's investigation into the open subtitles dataset, and they write, I can now say with absolute confidence that many AI systems have been trained on TV and film writers' work. Not just on The Godfather and ALF, but on more than 53 ,000 other movies and 85 ,000 other TV episodes.

9:40Now, of course, these copyright questions are not just about whether a dataset was used in training the model, but whether that model represents an economic threat to the source of that data. In other words, the big question ultimately is, does AI threaten those writers? And yet at the same time, there has clearly been a countervailing trend as well. perhaps unsurprisingly coming from the studios. Lionsgate, for example, recently announced a deal with Runway to train a new model on Lionsgate's film and TV library. Lionsgate's CEO told financial analysts, the entertainment business is a creative enterprise, but its future growth will require a combination of art and science.

10:15We believe that AI harnessed within the appropriate guardrails can be a valuable tool to serve our talent. And we believe over the longer term, it will have a positive transformational impact on our business. And speaking of more ethical approaches to degenerative video, A company called Moon Valley just announced a$70 million raise to build what they call more ethical video models. The founder of the company told TechCrunch, We shared a belief that video generation was going to transform media and entertainment, but the startups we saw operating in the space didn't have the necessary attributes to be successful.

10:45Existing companies were deeply antagonistic towards artist creators in the broader industries. Basically suggesting that the existing crop of companies feels like it can do whatever it wants vis-a-vis the industry, as opposed to thinking about how it can be augmentative. Still, the point here is that$70 million of funding just went into a company that's trying to specifically connect those dots. And honestly, on any given day, you can find opinions across the gamut. Nicolas Cage recently went on a media tour, warning Hollywood actors that AI, quote, wants to take your instrument. Tim Burton recently called it very soul-sucking.

11:15But then you're also seeing op-eds like this one in the New York Times magazine, What If AI Is Actually Good For Hollywood? Now, this article is specifically about how AI changes fundamentally what creators can even do. But maybe the most interesting part of it is the contextualization. The history of Hollywood, the author writes, can be told as a series of technological leaps, beginning with the invention of the camera itself. And each time something new comes along, jobs are lost, jobs are created, and the industry reorganizes itself. Everyone in town of a certain age has seen this movie before.

11:45Past leaps, though, have tended to have narrower impacts. Home video changed movie distribution. Digital cameras changed movie production. CGI changed visual effects. The difference here is that AI has the potential to disrupt many, many places in our pipeline, says Laurie McCreary, the chief executive of Revelations Entertainment. This one feels like it could be an entire industry disruptor. So with all that context, let's go back to these comments from Ben Affleck and see what he had to say. I want to, as we kind of wrap up here, I do want to come back to AI. Jerry, you mentioned it, but, you know, Ben, how did, you know, earlier you guys weren't here.

12:18we did a demonstration, my colleague Andrew Sorkin and I, recreated ourselves and our voices. How do you see it? I mean, is it a benefit or is it a real threat? Is it possible that a Netflix could say, you know, we're going to do our own, excuse me, James Bond thing out there with a bunch of actors that are completely recreated for this market or that market? A, that's not possible now. B, will it be possible in the future? Highly unlikely. C, movies will be one of the last things, if everything gets replaced, to be replaced by AI. AI can write you excellent imitative verse that sounds a little bit, and it cannot write you Shakespeare.

12:56The function of having two actors, or three or four actors in a room, and the taste to discern and construct that is something that currently entirely eludes AI's capability, and I think will for a meaningful period of time. What AI is going to do is going to disintermediate the more laborious, less creative, and more costly aspects of filmmaking that will allow costs to be brought down, that will lower the barrier to entry, that will allow more voices to be heard, that will make it easier for the people who want to make Good Will Hunting to go out and make it. Look, AI is a craftsman at best.

13:30Craftsmen can learn to make stickly furniture by sitting down next to somebody and seeing what their technique is and imitating. That's how large video models, large language models, basically work. of a library of vectors of meaning and transformers that interpret it in context, right? But they're just cross-pollinating things that exist. Nothing new is created. Not yet. Not yet. Yeah, not yet. And really, in order to do that, look, craftsmen is knowing how to work. Art is knowing when to stop. And I think knowing when to stop is going to be a very difficult thing for AI to learn because it's taste.

14:07and also lack of consistency, lack of controls, lack of quality. AI for this world of generative video is going to do key things more. I wouldn't like to be in the visual effects business. They're in trouble because what costs a lot of money is now going to cost a lot less. And it's going to hammer that space, and it already is. And maybe it shouldn't take 1 ,000 people to render something. But it's not going to replace human beings making films. It may make your background more convincing. It can change the color of your shirt. It can fix mistakes that you've made. It can make it, you know, you might be able to get two seasons of House of the Dragon in a year instead of one.

14:46And if that happens, according to macroeconomics, in, you know, cultures where there are basically oligopolies competing, what should happen is with the same demand and the same spend is they should just make more shows, which should, you know, you should have the same spend and now you can just watch more episodes. And eventually, AI will allow you to ask for your own episode of Succession, where you could say, I'll pay$30, and can you make me a 45-minute episode where, like, Kendall gets the company and runs off and has an affair with Stewie? And it'll do it. And it'll be a little janky and a little bit weird, but it'll know they're sad, it'll know those actors, and it will, you know, remix it, in effect, and it will do that.

15:25That's the value, in my view, long-term of AI for consumers, which is eventually my hope for AI is that it's an additional revenue stream that can replace DVD, which took 15 % to 20 % out of the economy of filmmaking, and there should be negotiated rights. Right. That's a key part of it. Because what do people want to make? Five-minute, 30-second TikTok videos where they look like the Avengers. Well, great. Just like you used to be able to buy your Iron Man costume at the store, you're going to buy your Iron Man pack and you and your buddies are going to look like Iron Man and Hawkeye, like, you know, on Twitch.

16:00That's what's going to really happen. All right. So really interesting comments, right? A lot of the response was just people being impressed and not thinking that Ben Affleck was going to get that deep. Another big theme was the appreciation of the nuance. Pinar Demerdag writes, the reason why I liked his word so much is that he is not deifying nor demonizing AI. AI is evolving to be a partner to assist us in our creative and production journeys where we can streamline all the parts that are devoid of intuition. So when it comes to the parts that involve vision, intuition, and celestial download, we can be as open as possible.

16:30And guess what? It will open up additional revenue streams for Hollywood that are gone with CDs, so it might as well save Hollywood. Huh. We'll come back to that comment around the business model in just a minute. Bilual Sidhu, who's the host of the TED AI show, writes, finally a grounded take on AI and filmmaking from a Hollywood A-lister. Yes, AI will shatter barriers to entry, budgets will drop, and the number of movies will rise, and more voices will have access to filmmaking. Paradoxically, this will raise the bar for creative quality. Yes, AI will hammer the VFX industry. It's already in a downward spiral and AI will accelerate that decline.

17:01Yes, customized episodes on demand are inevitable. We'll see new episodes of Game of Thrones long before AI can create an entirely new, coherent, and compelling series. Incremental episodes will beat full-season automation. I've frequently in the past used entertainment as my example of why my base case is the future of the world just being about more. I think it's very likely that we do get a world where people are able to just generate their own versions of their favorite IPs. I, for example, want West Wing Season 8 through 28. It's also likely that there will be new networks built on top of that as well, where people can share their best-generated episodes.

17:36But there will still need to be new sources of IP. Now on that front, Robbie Starbuck writes, Sorry, but Ben is wrong on multiple fronts. In the future, one small team or maybe even one person alone will be able to make a film that looks and sounds like a Hollywood blockbuster. AI can already write entire films, edit, render, simulate lighting in different environments, etc. Now interestingly, I actually think that he's sort of making just an extreme version of the same point. One small team or one small person alone making something great is just those barriers to entry being lowered. And where Robbie agrees is that human taste is going to be incredibly important.

18:08The rise of AI in filmmaking, he continues, will force real filmmakers to set themselves apart with truly original content. that emotionally separates the film from the AI films. Gilly Moreno writes, Affleck hits the nail on the head with the concept of taste. Those who develop and nurture good taste will be in the golden era of AI. Lastly, I did want to make mention of that revenue stream piece, which was almost a throwaway line of part of it. Here's Ben Affleck's frequent contributor, Matt Damon, talking about how streaming changed DVD economics and what the implications for the industry have been.

18:36A way lots of viewers can relate to is sitting on the couch on a Friday night, going through the streaming services, cycling through the movies and thinking to themselves, they're not making movies for me anymore. As somebody who's been intimately involved in movie making for 30 years, what are the macro Hollywood conditions behind that sentiment? Well, so what happened was the DVD was a huge part of our business, of our revenue stream. And technology has just made that obsolete. And so the movies that we used to make, you could afford to not make all of your money when it played in the theater because you knew you had the dvd coming behind the release and six months later you'd get all you know a whole another chunk it would be like reopening the movie almost and when that went away that changed the type of movies that we could make i did this movie behind the candelabra and i talked to a studio executive who explained it was a 25 million dollar movie i would have to put that much into print and advertising right to market it um what we call pna so i'd have to put that in pna so now i'm in 50 million dollars i have to split everything i get with the exhibitor right the people who own the movie theaters so i would have to make a hundred million dollars before i got into profit and and the idea of making a hundred million dollars on a story about like this love affair between these two people yeah i love everyone in the movie but that's a that's

20:12Suddenly, I think that even Hold aside, the specifics of the entertainment industry, and Tokenbender here really gets what's valuable about this. Ben Affleck gets it, they write, and most AI influencers are busy Doomer posting or AGI hyping. In other words, AI will change the world. It will not end the world. It will reshape industries. It will not end industries. and along the way there will be immense opportunity, opportunity honestly like we've almost never seen to make and shape the future. That's gonna do it for today's AI Daily Brief. Appreciate you listening as always. Until next time, peace.