In the latest episode of Denoised, hosts Addy Ghani and Joey Daoud take us inside AI on the Lot, a significant industry event that has quickly become the premier gathering for professionals exploring artificial intelligence in media and entertainment. The episode offers a firsthand look at how this event has evolved, the key discussions happening among industry leaders, and what these developments mean for creative professionals navigating the rapidly changing landscape.

The third annual AI on the Lot event marked a substantial step up from previous years, reflecting the increasing importance of AI in entertainment workflows. Held at the Culver Theater in Los Angeles, the event expanded from a one-day program to a comprehensive two-day, multi-track conference that attracted 1,200 attendees—completely selling out available tickets.

What stood out about this year's event was the caliber of attendees. As Addy noted, "We are seeing very, very top level VFX executives walking around amongst us. I saw execs from Netflix...Amazon...Asteria has heavy presence here." The location in Los Angeles proved strategic, making it accessible to industry decision-makers who might not travel to events in other cities.

The hosts highlighted several key observations about the event's growth:

As Joey remarked, AI on the Lot has positioned itself as "the most legitimate show for talking about AI seriously in a professional M&E pipeline and how it can be used at the highest level of media entertainment."

One of the most insightful discussions at the event centered on how AI is developing along two distinct tracks in filmmaking and content creation. As described by Addy after attending a panel featuring leaders from Eyeline, Wonder Dynamics, and Flawless AI, these approaches represent fundamentally different philosophies.

Path 1: Enhancing Existing Workflows

This approach focuses on integrating AI into current production pipelines to make them more efficient and cost-effective. Most established companies prefer this direction, as it presents fewer ethical concerns and more immediate financial opportunities. It's about using AI as a tool to improve what humans already do rather than replacing human creativity.

Path 2: Fully Synthetic Generation

This more experimental approach involves creating completely synthetic content from prompts—generating humans, environments, and entire scenes that don't exist in reality. While this path raises more complex ethical questions, it opens up new creative possibilities, especially for independent creators with limited resources.

As Addy summarized the industry sentiment: "The main sort of thing here is, is AI gonna help me be better and do more, or is it just gonna take my job away? And we of course, want the first one."

Joey added an important point from his own presentation at the event, where he discussed AI-assisted editing tools: "I think that might just take a mindset shift or how you harness it...depending on what your job is or what you're trying to do." He noted that while AI will enable people to do more if they take initiative, those in rigidly defined roles might face replacement.

A growing concern voiced throughout the event was the increasing difficulty of differentiating between real and AI-generated content. Joey shared a telling example of an AI-generated video featuring an "emotional support kangaroo" attempting to board an airplane:

"It was posted on an AI account and clarified that it was AI when it first went around. But it's like a cell phone looking video...it looks like at first glance it looks like someone shot on a cell phone... The original post was clarified, like 'this is AI.' But then people took the video, and now other accounts have reposted it...and it has like millions of views with no clarity that it is AI."

This raises significant questions about media literacy in an age when synthetic content can appear indistinguishable from reality. As Addy pointed out, "A lot of AI companies are building for one thing and one thing only: revenue. And most of the revenue does not come from film and TV. We're a tiny fraction of the market... Most of the revenue's from social media."

The hosts discussed how this might lead to a proliferation of low-quality, engagement-driven content, similar to what happened with the digital revolution but at a much larger scale. Joey noted the unsettling reality that he's begun questioning whether tutorial videos he watches feature real people or AI avatars.

Despite concerns, several promising applications of AI in professional production contexts were highlighted during the event:

SAG-Approved AI Film Production

Joey shared details about a conversation with Kavan the Kid from Phantom X AI, who premiered a fully generative film that was SAG-approved:

"They casted SAG actors... It was gonna be fully generated, but they wanted the characters to have the look of real people and be based on actors that are recognizable... They train a model on each actor and then use that when they were generating."

This represents an important milestone for AI in film production, showing how traditional entertainment institutions are adapting to new technologies. As Addy noted, "This won't be the last SAG actors in an AI film. If anything, this is the very first one of many."

Documentary and Historical Recreation

Another practical application discussed was using AI for historical recreations in documentary production, particularly for markets with limited budgets:

"Gen AI has fit the gap for a lot of these smaller productions where they just don't have options or budgets to do something that looks cool, historically accurate, and just elevates the videos," Joey explained, giving the example of recreating ancient Egypt without the expense of location shooting or elaborate set construction.

This demonstrates how AI can enable productions that previously would have been financially impossible, particularly in educational and documentary contexts.

Character LORAs for Consistent Synthetic Actors

Addy attended a session on "Character LORAs" (Lower Rank Adaptation models), which are fine-tuned models that enable consistent character representation throughout a film:

"The training data for LORAs comes from volumetric data. So in order to train a LORA, they are actually feeding it real human data, consistent data, and then once the LORA is trained, then they're using it as a starting point and then putting VFX on top of that."

This approach allows filmmakers to create synthetic characters with consistent appearances and behaviors across an entire production.

Despite all the technical discussions, a consistent theme throughout the event was that the core purpose of filmmaking—telling compelling stories—remains unchanged. As Addy recounted from a panel featuring cinematographer Michael Goi: "The main thing is that we still consider the main thing to be the main thing... If AI is a distraction, you're doing it wrong. The main thing is still filmmaking and telling a story."

Joey echoed this sentiment from his conversation with Davide Bianca from GRAiL Studio: "The common theme that we kept coming back to is just like at the core of everything they're trying to do is just tell a good story... using these different tools to tell a good story."

This perspective helps ground the technical innovations within the fundamental purpose of the entertainment industry. As AI capabilities advance, the distinction between high and low-quality content may increasingly rest on storytelling fundamentals rather than production values, since AI is democratizing access to professional-looking visuals.

The discussions at AI on the Lot highlight an industry in transition, actively exploring how artificial intelligence will reshape creative processes and production workflows. While some applications are already proving valuable—particularly in enhancing efficiency and enabling previously impossible productions—significant questions remain about how the technology will affect jobs, media literacy, and creative expression.

As both hosts acknowledged, we're likely to see parallel developments: AI will simultaneously enable more sophisticated storytelling for those who embrace it thoughtfully, while also facilitating a flood of lower-quality, algorithm-optimized content. The key for professionals will be understanding how to harness these tools to enhance their creative vision rather than being displaced by them.