When Hollywood Meets the Pentagon: AI's New Battlegrounds

Sam Hinton: The Academy just handed AI companies the biggest gift they could possibly ask for by banning AI actors and scripts from Oscar consideration.

Alex Shannon: Wait, what? That sounds completely backwards. You have thirty seconds to justify that take.

Sam Hinton: Think about it - Hollywood just told every AI company ‘your content isn’t good enough for our prestigious awards.’ That’s basically a challenge. Now every AI lab is going to pour resources into proving them wrong, and when they do…

Alex Shannon: Oh, I see where you’re going with this. They’re going to create a whole parallel entertainment industry just to spite the Academy.

Sam Hinton: Exactly. Netflix, Amazon, Apple - they’re all going to start their own ‘AI Excellence Awards’ and suddenly the Oscars are going to look like the old guard clinging to relevance.

Alex Shannon: That’s… actually kind of brilliant. And terrifying.

Alex Shannon: You’re listening to Build By AI, I’m Alex Shannon, and that cold open should tell you everything you need to know about today’s episode.

Sam Hinton: And I’m Sam Hinton. Today we’re talking about AI getting banned from Hollywood while simultaneously getting embraced by the Pentagon, a Chinese model that’s making OpenAI sweat, and medical AI that can spot cancer before it even shows up.

Alex Shannon: It’s May 3rd, 2026, and AI is basically reshaping every industry at once. Some are rolling out the red carpet, others are building walls.

Sam Hinton: Right, and the pattern of who’s doing what is fascinating. Let’s dive in.

AI-generated actors and scripts are now ineligible for Oscars

Alex Shannon: Alright, so let’s start with this Hollywood story that got us fired up. The Academy of Motion Picture Arts and Sciences has officially announced new rules making AI-generated actors and AI-written scripts completely ineligible for Oscar consideration.

Alex Shannon: This isn’t just a footnote in the rulebook either - this is the Academy drawing a hard line in the sand about what they consider ‘real’ filmmaking versus AI-generated content.

Sam Hinton: Yeah, and look, I get why they did it from a traditional filmmaking perspective. The Oscars are supposed to celebrate human artistry, craft, performance. But strategically? This feels like the music industry trying to ban digital recording in the 1980s.

Alex Shannon: That’s a strong comparison. But let me play devil’s advocate here - isn’t there something to be said for preserving human creativity? I mean, we’re talking about acting and screenwriting, which are fundamentally human expressions.

Sam Hinton: OK but here’s what people are missing - this rule doesn’t stop AI from being used in filmmaking. It just stops it from getting awards recognition. So now you have this weird situation where a movie can use AI for visual effects, for editing, for sound design, but the moment you use it for dialogue or digital actors, suddenly it’s ‘not Oscar-worthy.’

Alex Shannon: Right, so the line is completely arbitrary. And what happens when AI gets so good that you can’t tell the difference? Are we going to have AI detection tests for Oscar submissions?

Sam Hinton: Exactly! And here’s the kicker - this is going to accelerate AI adoption in entertainment, not slow it down. Because now every streaming platform is going to see this as an opportunity. ‘Come to our platform where we celebrate ALL kinds of storytelling, not just the Academy-approved kind.’

Alex Shannon: That’s actually a really interesting point. Netflix and Amazon already have complicated relationships with traditional Hollywood. This gives them a reason to lean even harder into experimental content.

Sam Hinton: And think about the economics here. If you can create compelling content with AI actors who never demand raises, never have scheduling conflicts, never have scandals - why wouldn’t you? The Academy just told the industry ‘we don’t care how cost-effective or innovative you are.’

Alex Shannon: But wait, let me push back on that cost-effectiveness argument. Yes, AI actors might be cheaper in the long run, but what about the human element that audiences connect with? Don’t you think viewers will be able to tell the difference?

Sam Hinton: I think that’s the million-dollar question. Right now, sure, most people can probably spot AI-generated content. But we’re talking about technology that’s improving exponentially. How long before AI actors are indistinguishable from human performers?

Alex Shannon: And here’s what’s really wild - this Oscar rule might actually incentivize that development. If AI companies want their content taken seriously, they now have a clear target: create performances so convincing that the Academy can’t justify excluding them.

Sam Hinton: Exactly! It’s like the Academy just created the ultimate benchmark challenge. ‘Make AI so good that we have to change our rules.’ That’s not discouraging innovation - that’s fuel for it.

Alex Shannon: So your prediction is that this creates a two-tier entertainment industry? Traditional ‘Oscar-eligible’ content and everything else?

Sam Hinton: I think we’re going to see new award shows, new recognition systems, and honestly, in ten years the Academy might be begging AI-generated content to come back because that’s where all the innovation and audience attention went.

Alex Shannon: You know what’s really interesting though? This could actually be great for independent filmmakers. If big studios start using AI to cut costs, but AI content can’t win Oscars, suddenly there’s this protected space for human-made independent films.

Sam Hinton: That’s a really good point. The Academy might have accidentally created affirmative action for human filmmakers. But I still think that’s a temporary protection at best.

Alex Shannon: Keep an eye on this because it’s not just about movies - this same debate is coming for music, for literature, for every creative industry. Hollywood just fired the first shot in what’s going to be a much bigger war.

US Navy signs deal with AI firm for training underwater drones to detect mines in Strait of Hormuz — $100 million would allow drone minesweepers to update their detection algorithms in days instead of months

Alex Shannon: Now let’s flip completely to the other end of the spectrum. Early reports suggest the US Navy just signed a $100 million deal with an AI firm to train underwater drones for mine detection in the Strait of Hormuz.

Alex Shannon: The key detail here is speed - if confirmed, this system would let drone minesweepers update their detection algorithms in days instead of months. That’s a massive operational advantage.

Sam Hinton: Dude, this is huge. The Strait of Hormuz handles like 20% of global oil shipments. If you can clear mines faster, you can keep that shipping lane open during conflicts. This isn’t just military tech - this is economic security.

Alex Shannon: Right, but let’s talk about what ‘updating algorithms in days instead of months’ actually means. Traditional military systems are notorious for slow update cycles because of security and testing requirements.

Sam Hinton: Yeah, that’s a big deal because warfare changes fast, but military tech usually doesn’t. If enemy forces deploy a new type of mine or change their placement patterns, historically it would take months to retrain detection systems. Now we’re talking about adaptive AI that learns and deploys new tactics in real-time.

Alex Shannon: But hold on, doesn’t that rapid updating create new vulnerabilities? I mean, if you can push updates that quickly, what’s stopping bad actors from compromising those updates or the AI making dangerous mistakes?

Sam Hinton: That’s actually the million-dollar question - or in this case, the hundred-million-dollar question. You’re right to be concerned. Fast iteration is great until your AI decides that school of fish looks suspiciously like naval mines and starts raising alarms.

Alex Shannon: Or worse, misses actual threats because the algorithm updated too aggressively. There’s got to be some kind of human oversight built into this system.

Sam Hinton: But here’s the thing about human oversight in underwater mine detection - humans are terrible at it. The whole point of using drones and AI is that the environment is too dangerous and the patterns too complex for human operators to handle effectively.

Alex Shannon: That’s fair, but there’s still the question of accountability. If an AI system fails to detect mines and ships get damaged or lives are lost, who’s responsible? The Navy? The AI company? The algorithm itself?

Sam Hinton: That’s probably being hammered out in contract negotiations right now. But honestly, given the alternative - which is potentially having one of the world’s most critical shipping lanes shut down by mines - I think they’re willing to accept some AI risk to avoid much bigger economic risk.

Alex Shannon: What I find interesting is the location choice though. The Strait of Hormuz isn’t just strategically important - it’s also where Iran has historically threatened to disrupt shipping. This feels like the US saying ‘we’re not going to let critical infrastructure be held hostage by mine warfare.’

Sam Hinton: That’s a geopolitical signal as much as a technological deployment. And $100 million suggests this isn’t just a pilot program - they’re going all-in on AI-powered naval defense.

Alex Shannon: And think about the implications if this works. Every critical shipping chokepoint in the world - the Suez Canal, the Panama Canal, the South China Sea - they’re all going to want similar AI mine detection systems.

Sam Hinton: Right, and if this works, every navy in the world is going to want similar systems. We might be looking at the beginning of AI-powered underwater warfare becoming the new standard.

Alex Shannon: But here’s what I’m really curious about - the timeline. Days instead of months for algorithm updates sounds great, but how do you validate those updates? How do you test whether your new mine detection algorithm actually works without, you know, setting off real mines?

Sam Hinton: That’s a really good question. They’re probably using simulation environments and historical data, but at some point you need real-world testing. And in a high-stakes environment like the Strait of Hormuz, the margin for error is basically zero.

Alex Shannon: Keep watching this space because maritime AI is about to become a major factor in global trade security. When your Amazon package depends on AI underwater drones keeping shipping lanes clear, we’ve entered a new era.

Kimi K2.6 just beat Claude, GPT-5.5, and Gemini in a coding challenge

Alex Shannon: Alright, let’s talk about something that should have everyone in Silicon Valley paying attention. Early reports from Hacker News suggest that Kimi K2.6 just outperformed Claude, GPT-5.5, and Gemini in a coding challenge competition.

Alex Shannon: Now, I have to emphasize this is from a single source, so we’re using some caution here. But if confirmed, this is a Chinese AI model beating the best that Anthropic, OpenAI, and Google have to offer.

Sam Hinton: Yeah, that’s a big deal because coding benchmarks are usually where Western models have dominated. Claude and GPT have been the gold standard for developers. If Kimi is beating them, that’s not just a technical achievement - that’s a competitive shift.

Alex Shannon: What makes this particularly interesting is the timing. We’re seeing increased tech tensions between the US and China, export restrictions on AI chips, and then boom - a Chinese model that outperforms everything else in coding.

Sam Hinton: OK but here’s what I’m curious about - coding challenges are very specific benchmarks. They test algorithmic thinking, syntax knowledge, problem-solving in constrained environments. But that doesn’t necessarily mean Kimi is better at real-world software development.

Alex Shannon: That’s a fair point. There’s a difference between solving coding puzzles and building production software. But still, if you’re a developer choosing which AI to help with your work, benchmark performance matters.

Sam Hinton: Absolutely. And think about the psychological impact here. For two years, developers have been saying ‘Claude for coding, GPT for everything else.’ If that changes, it opens up market share in a huge way.

Alex Shannon: Right, and this isn’t happening in a vacuum. We’ve got export restrictions on AI hardware going to China, but here’s a Chinese company potentially leapfrogging American AI with what they have. That suggests either the restrictions aren’t working or they’re finding other ways to compete.

Sam Hinton: That’s interesting but I’m a bit skeptical because we don’t know the full details of this coding challenge. What languages? What types of problems? Was it just algorithmic puzzles or actual software engineering tasks? The devil’s in the details.

Alex Shannon: Good point. But even if it’s just algorithmic challenges, that’s often where AI coding assistants shine. And for a lot of developers, that day-to-day problem-solving capability is exactly what they need.

Sam Hinton: True. And if Kimi can do this well in coding, what else might it be good at? Coding ability often translates to logical reasoning, math, even some types of creative problem-solving.

Alex Shannon: But here’s what I keep coming back to - if a Chinese AI model is beating GPT-5.5, that suggests they’re working with some pretty sophisticated technology. Either they’ve made breakthrough improvements in efficiency, or they have access to more computing power than we thought.

Sam Hinton: Or they’ve found completely different approaches to training models. Maybe they’re not just throwing more compute at the problem - maybe they’re being smarter about how they use what they have.

Alex Shannon: That’s actually kind of scary from a competition perspective. If you can achieve better results with less resources, that’s a huge strategic advantage. It means export restrictions might not be as effective as policymakers hoped.

Sam Hinton: The bigger question is access. Even if Kimi is better, can Western developers actually use it? Or is this going to create fragmented AI ecosystems based on geography?

Alex Shannon: Yeah, we might be heading toward a world where the best AI tools aren’t available everywhere. That’s going to create some weird competitive dynamics in software development.

Sam Hinton: And imagine being a startup trying to compete globally but being locked out of the best AI tools because of where you’re located. That could really skew innovation patterns.

Alex Shannon: Keep an eye on this because if it’s confirmed and if Kimi becomes widely available, the whole AI coding assistant market just got a lot more competitive. And that’s usually good news for developers.

US military reaches deals with 7 tech companies to use their AI on classified systems

Alex Shannon: Let’s stay in the military AI space for a moment. According to multiple sources, the US military has reached deals with seven technology companies to integrate their AI systems into classified military platforms.

Alex Shannon: This represents a pretty significant shift in how the military approaches AI adoption. Instead of building everything in-house, they’re partnering with commercial tech companies for defense applications.

Sam Hinton: This is massive because historically, the military has been super cautious about letting outside companies anywhere near classified systems. The fact that they’re doing this with seven different companies suggests they’re moving fast and casting a wide net.

Alex Shannon: Right, and it raises immediate questions about security clearances, data handling, and how these companies are going to manage the compartmentalization that classified work requires.

Sam Hinton: Yeah, that’s a big deal because most of these tech companies have global operations, international employees, cloud infrastructure that spans multiple countries. How do you reconcile that with classified military requirements?

Alex Shannon: That’s exactly what I’m wondering. Are we talking about completely separate, air-gapped systems? Special cleared personnel? Or have they figured out new ways to segment AI capabilities for security?

Sam Hinton: What’s interesting is the timing though. This comes right after that Navy underwater drone deal we just talked about. It feels like the military is going all-in on AI integration across multiple domains - land, sea, air, cyber, space.

Alex Shannon: And seven companies suggests this isn’t just the usual suspects like Lockheed or Raytheon. This is probably including pure tech companies - maybe Microsoft, Google, Amazon - companies that don’t traditionally do classified defense work.

Sam Hinton: Right, which means they’re prioritizing AI capability over traditional defense industry relationships. That’s a huge shift in procurement strategy.

Alex Shannon: But think about what this means for those tech companies. Getting cleared to work on classified military systems is a huge business opportunity, but it also comes with massive compliance burdens and restrictions on their other operations.

Sam Hinton: Yeah, and it potentially limits their ability to work with international partners or certain types of research collaborations. There are trade-offs here that go way beyond just the contract value.

Alex Shannon: But let me ask you this - if these AI systems are being used on classified military platforms, what happens when there’s a bug or the AI makes a wrong decision? How do you debug or audit AI behavior in a classified environment?

Sam Hinton: Oh man, that’s a great question. Traditional software debugging is hard enough in classified environments. But AI systems are basically black boxes. How do you explain to a general why your AI recommended a particular action when even the engineers don’t fully understand the decision-making process?

Alex Shannon: Exactly. And what about liability? If an AI system on a classified platform makes a mistake with serious consequences, who’s responsible? The tech company? The military personnel who deployed it? The AI itself?

Sam Hinton: Those are the questions that lawyers and policy makers are probably frantically trying to figure out right now. Because the technology is moving faster than the legal and regulatory frameworks.

Alex Shannon: And here’s another angle - if US military is integrating AI from seven different companies into classified systems, that suggests they’re not worried about vendor lock-in. They want diverse AI capabilities rather than standardizing on one approach.

Sam Hinton: That’s actually really smart from a strategic perspective. Don’t put all your eggs in one AI basket. If one system fails or gets compromised, you have six others to fall back on.

Alex Shannon: This feels like one of those stories where the immediate news is ‘military adopts AI’ but the long-term implications are about fundamentally changing how defense technology development works.

Sam Hinton: Absolutely. We’re probably looking at a new model where commercial AI capabilities get rapidly adapted for military use, rather than the old model of spending decades developing military-specific systems from scratch.

Anthropic in Talks to Buy AI Chips From U.K. Startup

Alex Shannon: Quick rapid-fire round here. Early reports suggest Anthropic is in talks to buy AI chips from a UK startup. This feels significant given all the chip supply chain drama we’ve been seeing.

Sam Hinton: Yeah, this is smart diversification by Anthropic. Everyone’s been so focused on Nvidia that having alternative chip suppliers, especially from allied countries, makes a lot of strategic sense.

Alex Shannon: And it probably helps that UK chip technology doesn’t face the same export restrictions as some other regions. Could be a way to secure supply without getting caught up in trade disputes.

Sam Hinton: Right, plus if this UK startup has something special - maybe more efficient chips or better price points - this could give Anthropic a competitive advantage in training and running Claude.

Alex Shannon: What I’m curious about is whether this is about raw compute power or something more specialized. Are we talking about chips optimized specifically for Anthropic’s model architecture?

Sam Hinton: That would be fascinating - custom chips designed around Claude’s specific computational needs. That kind of hardware-software optimization could be a huge efficiency gain.

Alex Shannon: And if it works out, this could be the beginning of AI companies moving away from general-purpose chips toward specialized hardware partnerships. That’s a big shift in the industry.

Sam Hinton: Absolutely. We might be looking at the birth of a whole new ecosystem of AI-specific chip companies, not just the Nvidia monopoly we have today.

US backs Anthropic AI model amid China tech tensions

Alex Shannon: Speaking of Anthropic, according to early reports, the US government is backing Anthropic’s AI model development amid escalating tech tensions with China.

Sam Hinton: This connects directly to that Kimi story we talked about earlier. If Chinese AI models are getting competitive, the US government probably wants to make sure American companies stay ahead.

Alex Shannon: It’s interesting though - government backing can be a double-edged sword. It provides resources but might also limit international market access if other countries see Anthropic as too closely tied to US government interests.

Sam Hinton: True, but at this point, AI leadership is being treated as a national security issue. Better to have government support and restricted access than fall behind technologically.

Alex Shannon: What’s really interesting is the timing - government backing for Anthropic, deals with seven tech companies for classified systems, the Navy’s underwater drone program. This feels coordinated.

Sam Hinton: Yeah, it’s like the US is making a massive strategic bet that AI dominance is critical for national security and economic competitiveness. They’re not messing around.

Alex Shannon: But I wonder how this affects Anthropic’s independence. When the government is backing your research, how much influence do they have over your development priorities?

Sam Hinton: That’s a great question. We might end up with AI companies that are technically private but functionally operating as extensions of national policy. That’s a pretty significant change from the traditional tech industry model.

AI finds signs of pancreatic cancer before tumors develop

Alex Shannon: Let’s end on a positive note. Early reports suggest researchers have developed an AI system that can detect signs of pancreatic cancer before tumors even visibly develop.

Sam Hinton: This is huge because pancreatic cancer has one of the worst survival rates partly because it’s usually detected too late. If AI can spot it earlier, that could save thousands of lives.

Alex Shannon: What’s fascinating is that this suggests AI is picking up on patterns in medical data that human doctors might miss - subtle biomarkers or imaging features that precede tumor formation.

Sam Hinton: Exactly. This is AI doing what it does best - finding complex patterns in massive datasets. And unlike some AI applications we’ve talked about today, this one has zero downside. Nobody’s going to complain about AI that saves lives.

Alex Shannon: Although it does raise questions about healthcare access. If this AI diagnostic tool is expensive or only available at certain medical centers, it could create disparities in who gets early detection.

Sam Hinton: That’s a fair point. We need to make sure these breakthrough AI medical technologies don’t just benefit people who can afford premium healthcare.

Alex Shannon: But the potential is incredible. If AI can detect pancreatic cancer before tumors develop, what about other hard-to-detect cancers? What about Alzheimer’s or heart disease?

Sam Hinton: Right, this could be the beginning of AI-powered preventive medicine becoming the standard of care. Instead of treating diseases after they develop, we catch them before they even start.

Oscars set new rules for artificial intelligence

Alex Shannon: And just to close the loop on our opening story, this Oscar AI ban isn’t just about actors and scripts. The Academy has set broader rules governing AI use in film production.

Sam Hinton: Yeah, they’re trying to create these detailed guidelines about what’s acceptable AI use versus what disqualifies you from awards. But I still think this is going to be impossible to enforce as AI gets more sophisticated.

Alex Shannon: Right, and it creates this weird situation where filmmakers might have to document every AI tool they used during production. That’s a lot of bureaucracy for creative work.

Sam Hinton: Which again brings me back to my original point - this is going to push innovation toward platforms and award systems that embrace AI rather than restrict it.

Alex Shannon: What’s interesting is they’re trying to draw these precise lines around AI use, but the technology is evolving so fast that their rules might be obsolete before they’re even implemented.

Sam Hinton: Exactly. It’s like trying to regulate a technology that’s changing monthly with rules that get updated annually. The math doesn’t work.

Alex Shannon: I keep wondering if the Academy realizes they might have just accelerated the exact technological development they’re trying to control.

Sam Hinton: They definitely didn’t think through the second-order effects of this decision. Sometimes trying to stop technology just gives it more motivation to prove you wrong.

BIGGER PICTURE

Alex Shannon: Alright, if you zoom out and look at everything we covered today, there’s a really clear pattern emerging. We’ve got Hollywood rejecting AI, the military embracing it, Chinese companies potentially leapfrogging American ones, and medical AI making genuine breakthroughs.

Sam Hinton: Yeah, what strikes me is that the institutions and industries that are winning with AI are the ones that are moving fast and accepting some risk. The military is throwing $100 million at underwater drones, tech companies are rapidly integrating AI into classified systems, medical researchers are deploying AI for early cancer detection.

Alex Shannon: And the ones that are trying to control or limit AI - like the Academy with their Oscar rules - might find themselves left behind. It’s like they’re trying to regulate a technology that’s evolving faster than regulation can keep up.

Sam Hinton: Exactly. And here’s what I think people should be watching: we’re entering a phase where AI adoption is becoming a competitive advantage not just for companies, but for entire countries and industries. The gap between AI adopters and AI avoiders is going to widen really quickly.

Alex Shannon: That’s a really good point. Six months from now, are we going to see a clear divide between AI-forward industries and traditional industries? And what happens to the ones that chose to restrict rather than adapt?

Sam Hinton: I think what’s happening is we’re seeing the emergence of parallel ecosystems. You’ll have the ‘AI-enabled’ version of every industry and the ‘traditional’ version, and they’re going to compete directly against each other.

Alex Shannon: Right, and looking at today’s stories, it seems like the AI-enabled versions are getting massive resource advantages. Whether it’s government backing for Anthropic, $100 million Navy contracts, or breakthrough medical applications - the money is following AI adoption.

Sam Hinton: And the geopolitical angle is getting more intense. China potentially beating US models in coding, the US government backing Anthropic, Anthropic buying chips from UK companies - AI is becoming a proxy for broader international competition.

Alex Shannon: Which raises this question - are we heading toward fragmented global AI ecosystems? Where the best AI tools available to you depend on your geography and political alignments?

Sam Hinton: That seems increasingly likely. And that’s going to create weird dynamics where innovation happens faster in some regions than others, not because of technical capability but because of access to tools.

Alex Shannon: But here’s what gives me hope - stories like that pancreatic cancer detection show that when AI is applied to genuinely important problems, it can create value that transcends political boundaries. Nobody’s going to restrict cancer-detecting AI because of trade disputes.

Sam Hinton: I think we’re about to find out. The next year is going to be fascinating to watch - and probably pretty consequential for anyone trying to figure out where they want their career or their business to be positioned.

Alex Shannon: And the speed of change is just accelerating. Look at the range of stories we covered today - from entertainment to defense to healthcare to international competition. AI is touching everything simultaneously.

Sam Hinton: Right, and that’s why these individual stories matter. They’re not isolated events - they’re all part of this broader transformation of how society and technology intersect. The patterns we’re seeing now are going to define the next decade.

OUTRO

Alex Shannon: That’s our show for today. From Oscar bans to underwater drones to pancreatic cancer detection, AI is reshaping everything at once.

Sam Hinton: Yeah, and if you want to stay on top of these changes as they happen, make sure you’re subscribed because this stuff moves fast. We’ll be back tomorrow with more stories about AI changing the world.

Alex Shannon: Thanks for listening to Build By AI. I’m Alex Shannon.

Sam Hinton: And I’m Sam Hinton. See you tomorrow.