Filmtools
Filmmakers go-to destination for pre-production, production & post production equipment!
Shop Now
We’ve talked about AI as a major theme in the industry for years now, but there was a notable change in 2025. While the focus has shifted toward its impact on budgets and workflows, 2026 promises even deeper integration. But AI isn’t the only story. What else is in store for film and TV production with camera tech, lighting, gaming engines and more?
Below is how various PVC writers explored those answers in a conversation took shape over email. You can keep the discussion going in the comments section.
No 2025 look back or 2026 look forward, can happen without bringing up AI. It has transformed the media creation industry and will continue to transform the media creation industry. Whether this is for better or worse is not something I will debate here, even though it is often the most common debate. I hope that we all tire of the AI slop that the generative AI space is capable of creating. The tools don’t do that creating it’s the humans around it that make the tools generate the slop. But we’re in a place now where the lazy and the dishonest can easily produce content that is viewed by many and often makes them money. If there is money to be made from AI slop, there will be slop made day in, day out. I repeat, I hope that we all tire of AI slop. I wonder what good the Sora app and all of the generated content do for the world. While I do admit swiping through it now and then, it’s ultimately forgettable the instant I leave the app. While I would never want to spend time watching AI generate clips of things that are in the real world, even the most clever prompts that generate some of the more engaging Sora videos are forgotten just a few moments after they leave the vertical screen.
But in that real world of leaner budgets and tighter deadlines, I do enjoy exploring the AI tools that are less about generating slop and more about helping create good work. Many new tools seem to come online daily, and I think part of that is that we’re living in a world of AI-assisted coding, so almost anyone with a good idea can at least attempt to make a great new tool. Tools are those things that make the editors work faster and more efficiently, and the tools are the things I look forward to AI making better so we can create real content, not AI slop.
This past year has been negative for many editors, thanks to artificial intelligence, the consolidation of TV networks and streamers, and the general impact Netflix has had and will have on post in markets like New York, Los Angeles, and London. Many editors working in television and film haven’t had a gig in months.
If you work in commercials, you’ve probably been busier; however, for many clients, the focus has shifted to social media over traditional TV commercials. Those social media pieces are often handled in-house by less experienced editors, which impacts post companies working mainly in that space.
On the technical front, 2025 has been an extension of the shifts made during and after covid. Specifically, most facilities have added remote capabilities and workflows, thanks to tools like Jump Desktop and LucidLink. Some facilities are now completely remote, which is not good news for office building landlords.
2026 will see an acceleration in the use of AI, which is here to stay – like it or not. My own opinion is that generative AI is a net negative for society and humanity. The legal issues of LLM training are slowly moving through the courts. But, recent deals between record labels and Suno and Udio can’t bode well for songwriters, musical artists, recording engineers, and producers.
Generative AI is already used to produce complete television commercials and social media content. This includes scriptwriting, translation and transcription, voice-overs, animation, artificial on-camera talent, and more. The quality is mostly substandard and full of artifacts, including random AI hallucinations. To get something useful – but never perfect – requires multiple attempts through different pieces of software using different AI models.
The next level of AI development will be the ability to modify an existing result, instead of a complete re-render (the current method). I doubt AI generation will ever be completely perfect and free of hallucinations and artifacts. The models are flawed and the AI needs of our industry are a small part of the overall AI pie. There just is no financial incentive for AI companies to massively improve music and video rendering, as compared with other uses of AI. After all, most of the AI companies aren’t turning a profit yet.
Unfortunately for video purists, many clients don’t seem to care. The trade-off in results and costs are justified for many companies and ad agencies. I suspect the biggest job impact in the near term will be on animators, VFX compositors, and graphic artists. Most likely we’ll see new positions emerge – AI specialists. The person who can properly leverage the existing AI tools to generate the results that the client wants.
I’ve been deep diving into the Generative AI pool for the past few years and now I’m using some form of it in every production I’m involved with.
Some overzealous marketers still don’t really understand that what you see in the eye-popping demos that fill our social media streams, are NOT indicative of simple one-two step results. I have had to reset expectations and teach colleagues and clients exactly WHAT can be created, modified and automated with the available tools now. It’s still pretty minimal. Yet people think that AI is doing all the work for them. It reminds me of the early desktop publishing days of the 90s and early Website builders in the early 2000s. Every major industry shift toward technology and self-producing software has only led to mediocrity in the media. Everything ends up noisy and pedantic. Viewers/readers/listeners start to recognize repeating patterns and long for something different. More meaningful. Something they can identify with. So the deluge of AI noise out there is only going to make people want for connection. Pushbutton automation doesn’t deliver that right now.
What IS working in regular corporate client productions now, are ElevenLabs for marketing VO, and some light image generation for backgrounds/stock. Replacement Broll and major characters, etc. are still not ready for primetime. I have had some good Broll animations I’ve generated with some lifestyle and property photos that I’ve animated with various tools. But other than making some animated characters for fun and tools demos, the clients aren’t buying it completely. Not yet. It’s slowly creeping into production workflows in many areas though, so if you haven’t yet explored the possibilities, then you’re going to be in for a surprise.
As they say, if you can’t beat ’em – join ’em! The people who can embrace the changing world and find out their place in it will be able to navigate the “new tech waters”. Everyone else will be standing on the shore wondering what happened. What our viewers/readers/clients still NEED is that human connection to their customers and themselves. Even in the Gen AI world, the works I’ve seen to date that move me are created by real artists and storytellers. Most of us are exactly that – we’re not just technicians pushing buttons. We’re still editing with the same level of passion and communicative storytelling (even in the commercial works) that somebody’s nephew with an iPad and a bunch of AI software can’t begin to understand.
Like everyone else, this year has been one of remarkables – good and not so good – from travelling the world accepting various film festival awards, winning the C21 Content Canada Int’l Co-Production Pitch Competition at TIFF, leaving LA (impossible financially), to settle near my daughter in a small rural town in a deep red state, – (the writer in me is taking copious notes – yes, it’s everything you think, and so much more y’all!), to getting stiffed on two directing contracts (big and small) and going through a lot of torture in this brand new experience. The level of incompetence with a cavalier attitude out there has been a big shock to me.
In 2026, I expect more developments both in audiobook production + distribution along with increasing demand & technological development for archiving and restoring of classic analog videotape formats:
- I expect JVC to release a VHS/S-VHS archival deck (player only) with built-in line and frame TBC, pristine noise reduction and dropout compensation. Given Apple’s abandonment of IEEE-1394 FireWire with Tahoe, I expect JVC to offer it with HDMI for monitoring and USB4 for capture, although backward compatible with USB3 using a USB-C connector. The device will certainly be multi-standard, with capability of playing all variants that ever existed of NTSC, PAL and SECAM, with menu-switchable Yadif de-interlacing with palindromic BoB for 50p or 59.94p realtime output derived from the original interlaced signal. Of course, there should be AI detection (with manual override) to compensate for NTSC-US tapes recorded with 7.5 IRE setup/pedestal. I also expect internal realtime upscaling and options to mask or show headswitching in the final output. In addition to the USB4 over USB-C, there should also be a full-sized built-in SDXC slot to capture directly to Apple ProRes422 or H.265.
- I expect Sony to release a 3/4″ U-Matic SP archival deck (player only) with built-in line and frame TBC, pristine noise reduction and dropout compensation. Given Apple’s abandonment of IEEE-1394 FireWire with Tahoe, I expect Sony to make it with HDMI for monitoring and USB4 for capture, although backward compatible with USB3 using a USB-C connector. The device will certainly be multi-standard, with capability of playing all variants that ever existed of NTSC, PAL and SECAM, with menu-switchable Yadif de-interlacing with palindromic BoB for 50p or 59.94p realtime output derived from the original interlaced signal. Of course, there should be AI detection (with manual override) to compensate for NTSC-US tapes recorded with 7.5 IRE setup/pedestal. I also expect internal realtime upscaling and options to mask or show headswitching in the final output. In addition to the USB4 over USB-C, there should also be a full-sized built-in SDXC slot to capture directly to Apple ProRes422 or H.265.
- I expect the quality of cloned voices to improve exponentially from services like FlexClip, ElevenLabs, Descript and DaVinci Resolve Studio.
- Via an update, I expect an improvement from Blackmagic to allow the Voice Convert in DaVinci Resolve Studio option to work directly from text too, not only from audio. I also expect Blackmagic to add the proper output for exporting audio using the AAC códec with the .m4a system-level file extension. In this article, Blackmagic acknowledged that this is currently missing.
- Given the increasing number of authors who publish a high quantity of books with audiobooks, many with take advantage of TecnoTur’s new Prolific website offerings with built-in e-commerce for direct sales and maximum royalties, with payment within a maximum of three business days, rather than having to wait 60-90 days from most online or physical bookstores. These Prolific authors can have the equivalent of 11 different websites for up to 11 different book titles, each one with its own proper and dedicated domain and proper SEO to be visible both by humans and search engines, an author page with its own visible domain, all for less than they would traditionally pay for two independent e-commerce websites. In addition, with direct sales, the author gets the email address of the purchaser, which never happens with online or physical bookstores.
- Given the increasing competition from both direct audiobook sales and much fairer royalties from Spotify, I expect Audible.com and Amazon to react accordingly, with more competitive royalties than the miserable ones they have been offering to date to authors.
- I expect Amazon.es (Spain) to fix its terrible mistranslation of the word Editor in book credits, which it is currently mistranslated as «Redactor». The proper word for Editor in Castilian is identical: Editor. Amazon.mx (México) got it right. «Redactor» means Writer, which is quite different from the Editor of a book. My Spaniard colleagues agree with me that Amazon.es got it very wrong.
In 2026, we will see those who capture footage and photography that cannot be easily replaced by AI become even busier, or at least continue with their expected workload. Here is my thinking, and I reserve the right to be wrong, as I often am.
There will be a split in our worlds: those who use AI and those who do not. As someone who films and edits, I think the editor in me may have to embrace AI to stay relevant, but I want to do so in ways that are less obvious. I do not love the overtly obvious use of AI visually. It stands out, and not always in a good way. The slop will become a waterfall on social media. Most of the AI slop will be in jest, but it’s easy to manipulate, which makes me value in-person experiences even more than anything online.
As I stated above, AI will become even more present in our creative space, from helping editors create once-highly-expensive transitions from a production standpoint to voice-over narration.
Some of that AI voice-over narration is already decent. AI VO is an obvious cost savings for many businesses. The adoption of AI narration will mean a complete collapse of the middle- and low-end of the voice-over business, and only those at the stratosphere of celebrity will retain any semblance of narration work. If voice-over narration is your side hustle, I’m sorry to say but it might be time to pivot.
In the world of cameras, the market is not running at full speed, and I don’t think it’s really recovered fully since COVID. The big players will always find ways to save on costs and/or buy up smaller competitors. This isn’t all bad news or mundane; great things can happen when people solve problems. Look at Nikon and RED as a great example. Now that the Nikon ZR is out we can see the first real example of Nikon and RED combining tech to make an interesting camera, which I am testing now. We may see more medium format footage in 2026 since Fuji and Blackmagic Design both have reasonably prices options. Though, the Blackmagic URSA Cine 17K will become the one most adopted by filmmakers.
When it comes to production lighting, 2026 will continue to see LEDs dominate. Look at Aputure, NanLux, GoDox, and other companies building high-output lighting systems. The benefits LED fixtures give to a production are real and immediate. From saving on power and generator rental to immediate color-temperature changes to very real cost savings, expect LEDs to replace most HMIs and Tungsten lights, except for the very high-wattage options that LEDs have yet to challenge.
About 40 years ago, the combination of the Apple Mac, Pagemaker software, Postscript & desktop laser printers completely revolutionised the publishing industry. In only 3 years many printing trades which had existed for centuries were rendered completely obsolete. That period is known as the “desktop publishing revolution”.
As the medium of video continues to evolve, both in the way it is produced and the way it is consumed, I am always wondering how close we are to a similar revolution – one which would see the vast majority of established professions rendered obsolete almost overnight. I have written before about how desktop video technology mostly eradicated the “million dollar rooms”, with companies including SGI and Quantel gone and nearly forgotten. But although the massive shift to desktop video production can rightly be considered a “revolution”, during the late 1990s it was only a revolution in the production process. The same products were being made, just using newer & cheaper technology.
But are we in another revolution, one that is more about the types of content being consumed and not how it’s being produced? There are many established workflows for producing video content for different industries, but we can’t only focus on changes to one side of the equation. Adapting to the changing appetite of the industry/ industries, and what types of content are in-demand and who wants them, is just as crucial to economic survival as adapting to newer production technologies.
I don’t know what next year will hold, and I certainly don’t want to try and guess. I’m still a bit stunned at how quickly 9×16 video became a mainstream format, so I’m certainly no prophet. But whatever production tools may dominate headlines, I suggest we all pay just as much attention to market forces and where the demand for content is shifting to.
This past year has brought big changes for me and my family – we moved from Cape Town, where we’d spent the last ten years, back to the UK. And the world of the remote workflow, which so sped up during the pandemic and has become a part of most editors lives, was a central part of my experience – in that I can’t imagine too many jobs where you could move half way round the world in the blink of an eye (extra points if you get the reference there) without any disruption at all. My clients didn’t care where I sat and my edit files were all on the cloud (LucidLink in this case). Of course, I now owe tax in both countries, but that’s another story.
But I wonder if we did a poll of our readers, how you feel about working remotely, if you do. I wonder if it’s taken away some of the creativity that used to come in the edit suite when you were sitting next to a director all day. That said, I feel I can get more done without someone in the room with me. I think in an ideal world I’d have my director or edit producer with me one day of the week (or maybe 2 hours each day) and have the rest of the time to myself.
One of the big markers of 2025 was the closing down of a fair few post houses, which shows that the market has shifted and is unlikely to go back. This was covered on this month’s Alan Smithee roundtable podcast and if you didn’t get a chance to listen, I highly recommend it (it’s how you can sound clever at meetups). I do feel for the next generation – especially with the lack of opportunities to break in at post shops. I had to sit at a careers fair recently for my child’s school to represent post production and I really didn’t know what to say to the kids – part of me wanted to say that this is not a career they should even think of pursuing and the other half wanted to say that it’s the best job in the world.
In terms of how AI affected me this year, it’s mainly continued to be the machine learning side of things. I used Topaz Video AI’s Starlight model for the first time to restore some poor phone footage and it was remarkable. It was a massive step up from the other models and equally takes a lot longer to render. So you either use credits to do it in the cloud, or use the local model (which is somehow better quality) and let it take over your computer for a good while. Topaz have just released their Realism update across their apps which I’m excited to test. I also used UVR (Ultimate Vocal Remover) to separate dialog, music and effects and found specific models (eg Bandit – works in UVR beta only) that have been trained to do just that, rather than the ones that trained on music and vocals. Very useful when an archive has only M&E and not music and effects separately for example. Next year I’m awaiting the release of iZotope RX 12 to see if they’ve kept pace with the industry.
All in all, I am very sad about the industry, the loss of jobs and seeing friends out of work, and the use of AI in the seasonal commercials (yuk). I have a kind of survivor’s guilt that I’ve been working so much. I do hope 2026 brings a change, but I am no longer waiting for it and honestly I think many will have to finally make the decision to seek their livelihoods elsewhere.
TL;DR We’re afraid of change. We always have been.
2025: AI Fatigue is Real
Disclaimer: I’ve consulted on ML for a decade, AI tools since 2019—before the pandemic.
AI promised to accelerate post-production pipelines. Instead? Constant failures, endless double-checking, and massive Dunning-Kruger around output validity. The refrain: “This is the worst it’ll ever be.”
Reality check: Some tools work great—noise reduction, transcription, translation. Most are problematic—unreliable, lacking controls, creating new headaches.
I embrace these tools as an early adopter. But one-touch instant results? Still a fantasy.
You Must Understand the Fundamentals
Best case? AI removes drudgery, adds creativity.
But you can’t skip knowing how things work. Using GenAI to cut corners you don’t understand won’t work. You’re pulling a slot machine lever over and over.
How can you produce a great shot if you can’t describe what makes one great? When AI fails the roto work, you still need to know how to fix it.
WHEN (not if) AI fails, how will you fix it? How can you trust a tool that hallucinates?
Test it yourself: Open any LLM (ChatGPT, Claude, Gemini) and ask about something you’re expert at. See if it gets it right. Note how confident the results sound. It’s just as bad everywhere—you just can’t tell when it’s not your domain.
The Real Problem
Clients, middle management, C-suite want to believe these tools work instantly and make people more productive (cheaper! innovative!). Rose-colored glasses prioritizing speed and cost over reality and quality. They’ll blame the person holding the football not the planning.
Looking Ahead to 2026
TL;DR: Pace accelerates. Sloppy low-hanging fruit multiplies—transcription, animated text, vibe-coded garbage. Real wins? Unsexy infrastructure work.
Networks Matter More Than Tools
Remote workflows won. C-suite can keep paying for empty offices—we’re not coming back. But we lost something: connection.
Invest in your network—virtual and physical. Weekly minimum, daily if possible.
Slow Adoption is Smart
Run a large facility? Your slow adoption is a feature, not a bug. Let early adopters absorb the pain. And there will be pain, especially around AI.
Expect AI Slop Everywhere
Failed products chasing hype, subscriptions and looking for that golden VC money. Cheap clients adopting broken tools. Audiences tuning out inauthenticity and blaming the medium.
We’re already deaf to AI-generated text. Soon we’ll be deaf to AI social content too. I’m already summarizing every YouTube video before watching.
What Actually Excites Me
Inexpensive vocal translation across languages. (Douglas Adams warned us about the Babel fish.) We’ll be able to ship wider than ever. Hopefully we won’t make any brand blunders
Finally, I’m watching/waiting for a real AI MAM to happen. Asset management tools powered by AI—nobody wanted to tag assets. Now nobody has to.
Looking at all the AI-related responses so far, it seems timely to have spent my spare time this year writing a book called AI for Creative Production — which I hope you enjoy. There’s so much going on, so much change, and so much hype that it’s hard to find the truth out there.
First up, it’s important to realise that not all AI is the same. Borrowing from Michael Cioni and extending a little, I like to split AI up into Utility AI, Generative AI and Automation AI.
Utility AI: recognition, classification and understanding
Generative AI: creation of text, images, video, audio and more
Automation AI: performing tasks humans usually would
If you’ve selected a person or object with Magic Mask in Resolve, Magnetic Mask in Final Cut Pro, or Object Selection in Premiere (in beta), that’s Utility AI. Sorting, categorising, intelligent MAMs like Peakto and tools like Jumper that help you find needles in haystacks. Stem splitting in Logic Pro and elsewhere. Beat detection. This is all good.
Generative AI can find some great uses in ideation and mockups and in some tedious animation tasks, but in finished work? Copyright, ethics, and a general icky smoothness still get in the way. Just as you can usually still spot a fully CG character, you can usually spot AI. Gen AI photos are better than videos, but none are perfect. I don’t agree that this will just “keep getting better” because there isn’t enough power — compute or electrical — for it to keep on growing at its current pace.
Today, most Gen AI video is pretty bad, leading consumers to associate “Gen AI” with “cheap”. Big clients like McDonalds are finding that AI ads don’t do what they wanted, and I suspect the reputational risk of using AI will push some clients back towards traditional production.
Worse, even if you manage to create something believable, there’s also the risk that you’ve recreated a real professional’s work and can be sued for it. Even Netflix doesn’t want you to use AI in final deliverables. It’s messy, but as an augmentation of existing, real work? Sure.
Gen AI audio has some clear workflow wins too. I’ve already taught multiple video editing students using Gen AI synthetic voices to narrate internal corporate training videos. If there’s no VO budget, it’s not for public consumption, and they don’t like their own voices… why not?
Automation AI isn’t as good as it probably will be, but right now, there are some surprising winners in this space. In video, Eddie AI is smarter than I expected, while other solutions like Gling are useful for specific video workflows. But AI’s pretty good at writing scripts to automate other apps, so you can gain some workflow boosts today if you want to get your hands dirty.
My book should be available by the time you read this, so if you’re interested in a round-up of many AI tools for creative purposes across text, images, video and audio, please check it out.
However, I don’t think AI is the biggest threat to our careers in video production. To me, it’s attention. The democratisation of production and delivery has meant we all have more choices than ever of what, when and how to watch video content. It’s not surprising that the video equivalent of french fries is the most popular choice.
As consumers get used to watching shorter and shorter content, and get ever more used to swiping away when they’re bored, or annoyed, or think they’re watching an ad, there’s less opportunity for higher quality, longer-form storytelling.
On the flip side, the viewing audience is more diverse than ever, and social platform analytics are not the entire story. We need to keep reminding clients that not every video should be as short as possible, and that a small but interested audience is still worth talking to. But the high end is struggling.
Cinema is dying slowly, both in an obvious way (not as many movies in cinemas) and a more subtle way (viewers spending more time watching shorter stories). Netflix buying Warner Brothers is not the beginning of the end, but more likely a notable point in the middle of a long, gradual decline. Stories aren’t going away — we’ve never had so much great TV — but they’re changing.
The future still has careers for video professionals, but there will be many more careers for semi-professionals. That’s not a bad thing, but the world still needs good videos. I hope we can all keep making videos that shift the needle, that tell people something they didn’t know, that make people feel, because the world needs more than vertical reaction videos to AI-made memes.
In camera and lighting equipment, I’ve recently found myself repeatedly pointing out that we’ve reached the end of two big changes almost simultaneously: the switch to low-energy lighting and digital cinematography. Those processes of change are done. That could have been said to have occurred at any time in the last five years, probably, but it always seemed safest to wait until things were really, seriously established, because otherwise someone will object to the idea, but change in on-set gear has become more incremental than at almost any time since the mid-2000s. That’s had a number of outcomes. First, and most positively, is the unavoidability of the idea that results are far more about technique than technology. That’s something experienced people have always known but that entry-level people often take time to work out, but entry-level camera gear is now so good it’s impossible to blame the hardware. Audio reached this point more than a decade ago and it’s interesting to see it happen in camera.
The other thing is interest in large-sensor cameras for live broadcast, which seems to be for two reasons. First is the sheer creative competitiveness of high-end live work, which is often overlooked; OB providers are under pressure to provide the glossiest-possible show. Given that apex of camera technology, resolution has been pushed as far as it needs to be pushed, and there’s very little technical reason to do this; it’s a creative choice, as well as a big push from camera manufacturers who’d rather sell sixteen units (plus fibre converters, CCUs, etc) to an OB truck than three to a rental house. How practical or necessary any of this is, and whether this sort of stuff will become a long-term growth market, remains to be seen. Certainly it’s impossible to make ultra-long-range zooms for Super-35mm sensors which match the behaviour of those for 2/3″ broadcast cameras. Still, several of the broadcast specialists have built large sensor cameras, and cinema camera manufacturers have offered conversion kits.
None of this addresses the thing which actually occupies the mind of professionals with a background in single-camera drama, which is the sheer paucity of work. Frankly, this was always going to be inevitable; the huge production boom of 2018 and 2019, and the post-pandemic rush, was never sustainable, being based on absurd levels of production spend provoked by streaming companies which did not, at that time, have to show a profit. I think a lot of the industry was wilfully blind to that on the basis it was making people short-term money. What’s less understandable is the long-term investment that was made on the apparent assumption it would go on forever. Obviously it’s miserable that people are having to seek work elsewhere but it was never going to last forever.
Considering all of the strange(r things) shifts in the industry, one may look for a special formula(one) or something fantastic(four) to lead us into a super(man) new future in video production. Even when things feel just okay(pop demon hunters), we can still find ways to create sentimental(value) stories of this life(of a showgirl) and reflect on especially amazing things, or as they say in Boston, wicked(for good). A holiday break might feel welcome(to derry) and for some, time off feels a bit alien(earth). What might our Big Bold Beautiful Journey hold for 2026? A few elements we might want to be aware of:
Vintage and retro still reign (or maybe I just want them too): The revival of VistaVision…the popularity of Point-and-Shoot cameras… even in an age when the crew of F1 built new ways to see inside cars using iPhone sensors, vestiges of the past continue to play roles in our cultural conversations. From a momentary usage of 8mm in Stranger Things to seeing Expired Film Club at matches, these moments facilitate joy and, better yet, can thrive alongside hugely technological and modern approaches in storytelling (Sinners, F1). I personally look forward to incorporating vintage lenses at F1 someday (with my dad’s lenses that he used to take photos of race cars) and at IBC in Amsterdam (Delft lens abroad) in September.
An awareness of the younger generation’s dislike of generative AI: Every so often, the institution at which I work will experiment and post an AI-created image. The response by our college-aged audience is overwhelmingly similar: Why use AI when you have talented graphic designers you can work with? My brief conversations with younger creatives at Adobe MAX revealed some dismay at the AI-centric announcements at both keynotes. Even at the inspirational speeches, a time usually reserved for highly emotive storytelling, there was a display of Firefly’s abilities. Keeping that awareness in mind as generative AI continues to innovate might help its long-term adoption. Creatives hope for the ability to BE highly creative with AI helping to automate our mundane tasks to free up our brainspace.
– Updated media literacy training for all (not just school-aged individuals): The ability to recognize and question the sources of content, particularly in the online sphere, is a skill that not only needs to be taught in schools, but also simultaneously developed (somehow) by everyone who consumes content. As AI increases its ability to mimic and represent reality, so too must our awareness and diligence in questioning what we see on social media.
– Immersive Content Accessibility: When Apple announced they were the exclusive broadcast partner of F1 in 2026, my first thought was: are we going to be able to watch F1 in the Apple Vision Pro? My second thought was: will the Apple Vision Pro ever get cheap enough to be popular and light enough to wear for the entirety of a race? Imagine feeling like you’re in the paddock on race day! But at a price that’s more expensive than a trip to Vegas for F1, customers will put their money elsewhere. It feels like Apple may only see success if they create an enticing deal (make the price much cheaper when you buy a computer), or lower the price significantly. In marketing, we talk about highlighting the problem that a product can solve. But right now, the Vision Pro has no value proposition. Its only purpose is for entertainment. But consider: creating educational experiences for everyone and everyone? And yes, getting “trackside” at your favorite race, too. The Vision Pro can’t just be cool; it also has to be solution-oriented.
It’s tough to talk about this year from a Cinematographers perspective because, truthfully, I didn’t see many people working at all. Obviously the guests I’ve had on Frame & Reference are, but even some of the biggest DPs I know have told me privately how they haven’t worked in 2 years, are working for Scale, or getting denial after denial interviewing for gigs. These are people who have made some of the most popular films in modern memory. From an indie perspective, the only thing keeping me personally alive and housed are a few Documentaries and the odd corporate gig here or there. I won’t lie and pretend it’s been great in LA; Rising costs across the country have squeezed absolutely everyone, so a lot of those commercial and corporate gigs have dried up in favor of cheap/AI/no work at all. I’ve personally been on two gigs from a couple of the largest companies on the planet trying to figure out how to use AI to save money (like they didn’t have enough), with various levels of efficacy, and in one instance all the stuff they had me “fixing” would have just been saved by shooting the photos in question for real (I haven’t personally run in to AI video yet).
Something I am noticing is a general aire of “luddatry” becoming more prevalent in the world. Not ignorance but a yearning for anything that doesn’t whif of “new” “tech” “AI” “cutting edge” “never before seen” type stuff, across all markets. Everyone seems to be asking for more “real” things, which seems to be anything prior to 2012. Digital is down, film and tape are up. Cellphones are down, point and shoots are up. Blockbusters are down, Indies are up. Internet ad revenue is out, Patreon is up. Fitted jeans are down, JNCOs are up. The convergence of the 30 year nostalgia treadmill and modern internet technology “ruining everything” is creating a HUGE backlash against tech progression and invasion into our lives (let alone the type of surveillance and lack of ownership that brings). Also, can we STOP with everything being a subscription!? How is anyone supposed to do anything when everything they use day-to-day takes a percentage of their paycheck in perpetuity? That alone might be tech’s greatest sin brought upon the world.
While right now EVERYTHING is selling their “AI Powered So-and-so” (I just got an ad for a “100% AI powered David Blaine Magic learning App” for gods sake) I can almost guarantee that you’ll soon see the exact opposite, with companies telling you how their product is “made by real people, for real people” and that “no AI at all” was used to make it. I know in my bones that’s probably going to happen within… let’s call it 5 years. Beyond tech bros and people who (rightfully, unfortunately) believe Chat GPT to be a better research tool than Google, AI doesn’t seem to be making anyone’s lives better, and is almost immediately making everything worse. Have you seen the price of RAM now!? That entire swing away from consumer computing is going to be catastrophic for editors and anyone who wants an affordable memory card, let alone a GPU or SSD.
What I do think the push against AI and associated “stuff” means is that, while right now might be really tough, if you’re in the position where you or a friend you know can create what is essentially B2C videos, or are on an “A24” production, or are in the education space as a personality, or can even sell a physical product, you may be in good shape in the near future. I fear that the next 3 years are going to see further corporate consolidation, fewer worker protections, more rising prices, and less opportunity for the middle-class film worker to succeed. I’m not saying it’s impossible, but just that unless you’re established and in a narrow range of lifestyle where taking a “lower paying” gig is kind of in your wheelhouse, things are dire (to quote a major DP I know). I don’t have much to offer other than to say that dips like this in any field are why it’s important to always have diverse sources of income (or a very kind, more stably-careered spouse). I also don’t think “The Industry is Dead” as many people seem to be screaming. What I do think is that we’ve got a perfect storm of factors that make this industry very difficult to work in at the moment, especially if you’re coming from the high-highs of the “streaming bubble” where everyone was working all the time and shows were being shot all over the place. Except, obviously, here in LA. With that being said, Atlanta, New Orleans, Vancouver, all are seeing reduced shooting. Perhaps this will localize productions to their native hubs, but I’m not confident right now. I’m also not an analyst, I’m just a camera guy. So again, since I can’t really say “what” happened this year because I basically wasn’t part of it beyond a couple Docs, what I can say is that my gut tells me that in the future we actually will see more productions, but they’ll all be smaller. Kind of the way I see theaters: many may go away, but the ones that stay will be serving a smaller audience of rabid film lovers, and the films that go in those theaters will be supported by fans of the filmmakers themselves, not necessarily the stars in them as we’re used to. So that could be interesting, although that also means we as film workers have to be more personally exposed, which is obviously not what any of us signed up for. Or you can just be so good at your job that you’re in demand by the folks more willing to be “out there”. That’s an option.
There’s already been such great insight laid out here but I felt the need to share a recent story that comes from the world of video games and some of you may not be aware of: the Clair Obscur: Expedition 33 saga.
A French cinematic videogame created by a small development studio, Sandfall Interactive, Clair Obscur: Expedition 33 sold 500,000 copies within 24 hours after its release on 24 April 2025, and by October 2025, the game had sold over 5 million copies worldwide. In December 2025, Xbox Wire reported the game was the biggest third-party title launch on Xbox Game Pass that year.
The game was nominated for thirteen awards and won nine at the Game Awards 2025, the highest number of nominations and wins ever for a single title in the event’s history to date; the game also won all seven of the awards it was nominated for at the 2025 Golden Joystick Awards, including the 2025 Ultimate Game of the Year award… and has been ranked No. 1 on multiple publications’ year-end lists of 2025. Voted “Best Game of 2025” at the IGW Awards and IGN Community Awards, it also won the “Best Story”, “Best Soundtrack” and “Best Independent Game” categories at the PlayStation Blog Game of the Year awards.
Clair Obscur: Expedition 33 also won the “Indie Game of the Year” and “Indie Debut Game” awards at the 2025 Indie Game Awards… but shortly after the announcement it was disqualified because of the use of generative AI. The decision of the Indie Game Awards jury added fuel to a fire that has been burning for a while, as the discussion around generative AI continues, with people entrenched on opposite sides regarding its use. Comments on social media go from a “What’s so bad about using AI? If the result is a good game, who cares?” to “Stop using AI for placeholder assets. Use sketches. Use lorem ipsum.”
Why was Clair Obscur: Expedition 33 disqualified? Because the Indie Game Awards have a hard stance on the use of gen AI throughout the nomination process and during the ceremony itself. When it was submitted for consideration, a representative of Sandfall Interactive agreed that no gen AI was used in the development of Clair Obscur: Expedition 33, so the game was announced as a nominee.
In light of a published interview with Sandfall Interactive confirming the use of gen AI art in production being brought to the jury of the awards attention on the day of the Indie Game Awards 2025 premiere, the jury decided to disqualify Clair Obscur: Expedition 33 from its nomination. While the assets in question were patched out – adds the IGA jury, “and it is a wonderful game, it does go against the regulations we have in place. As a result, the IGAs nomination committee has agreed to officially retract both the Debut Game and Game of the Year awards.”
The fact that Sandfall Interactive did not mention, initially, the use of generative AI, even for placeholders that were later removed, is the main reason why the game was disqualified. If Sandfall Interactive had confirmed the use of generative AI when asked, the game would automatically be excluded from the awards. It’s unfortunate, but those are the rules and while the discussion rages on social media, the Indie Game Awards states that “As gen AI becomes more prevalent in our industry, we will better navigate it appropriately. The organizational team behind the ceremony is a small crew with big ambitions, and The Indie Game Awards can only grow with your help and support. We already can’t wait for the 2026 ceremony!”
Although this story may be seen as unrelated to the movie industry, video games production is now – and has been for a while – similar to movie productions. It’s not just that video game tech is powering a revolution in Hollywood – one example is that both the video game Clair Obscur: Expedition 33 and the movie Megalopolis, directed by Francis Ford Coppola, use Unreal Engine – video games, just like films, use actors for voiceovers and motion capture. In the future, according to some, AI-generated dialogue, adaptive storylines, and AI-driven characters that react to the choices made by the player will take smart interactive storytelling to heights previously impossible. Apparently, less and less actors – and probably other professionals – will be needed in that future… both for video games and movies.
The decision taken by the Indie Game Awards regarding Clair Obscur: Expedition 33 should make us pause and think. With all the discussion, in 2025, about AI in its different forms, and 2026 round the corner, maybe this is the moment to make choices. According to an article in Psychology Today, “Our choices will decide whether AI turns out to be Pandora’s Box or a GIFT to humanity.”
Links to check
Clair Obscur: Expedition 33 producer confirms generative AI use in Game of the Year winner
I’m Getting Real Tired Of Not Being Able To Trust That A Video Game Doesn’t Have AI Crap In It
https://aftermath.site/the-alters-ai-controversy/
Is AI a Treasure Chest or Pandora’s Box?
