COEY Cast

Audio is getting weird in a good way. This episode breaks down Alibaba Tongyi Lab’s open source Fun-CosyVoice 3.5 and Fun-AudioGen-VD, why natural language control for voice matters, and how to turn soundscapes into a real sonic identity instead of generic futuristic whooshes. Then the focus shifts to OpenAI’s Responses API and Agent SDK, what actually makes an agent different from workflow automa...

Google drops Gemini 3, OpenAI teases GPT 5.3 Codex and GPT 5 Mini, and ByteDance levels up Seedream 5.0 and Kling 3.0. This episode breaks down what actually changes for creative workflows. Learn how multimodal models can review your ads, why portability of your brand brain matters, and how to treat small models as bouncers not headliners. Get practical patterns for using faster code models to aut...

GLM5 from Zhipu AI drops open weights under an MIT license and everyone screams freedom without reading the chores list. This episode breaks down what “ownable” really means for marketers from infra and governance to long context workflows for brand consistency. Then it dives into Google Ads AI Max Text Guidelines and how banned terms, tone rules, and human choke points keep platform AI from wreck...

Voice agents just leveled up with OpenAI gpt realtime 1.5 and the real win is not small talk. It is inbound and post click voice concierges that can query your product data, booking tools, and CRM without turning into a spam cannon. The conversation breaks down consent, compliance, and “no manipulation” guardrails. Then it shifts to Meta Llama 3.3 70B and open coding models as the hidden engine fo...

Meta open-sources SAM Audio and suddenly sound becomes selectable instead of mysterious. Adobe and WPP push agentic marketing with Firefly and Quick Cut turning coordination and first-draft edits into automated workflows. Google Veo 3 with synced sound and Flow puts AI generated video closer to real filmmaking, from stitched shots to global-ready audio. This episode breaks down how open tooling ve...

Open source models are leveling up and crashing straight into real workflows. This episode breaks down Alibaba’s new Qwen 3.5 open weights for agentic workflows, why “smaller model bigger results” only matters if your loops get cheaper and more reliable, and where governance still fails when agents touch real systems. Then it dives into Llama 4 rumors around huge context windows and local first se...

Mercury 2 is shipping a reasoning diffusion LLM with lower latency and cheaper inference, but does that actually fix your workflow bottlenecks or just move them downstream. Hunter and Riley break down what diffusion style language models mean for multi step agents, real world marketing automation, and repurposing content without turning it into generic sludge. They also cover OpenAI realtime audio...

Google’s Lyria 3 is crashing the ad music party with near instant jingles tailored to campaigns and creators, putting serious pressure on stock audio but boosting anyone who knows taste and risk. Runway is turning into a multi model video mall where teams compare engines inside one workflow instead of tab surfing. Vercel’s AI Gateway is making video generation feel like an API utility for internal...

ChatGPT is quietly testing ads inside conversations, and the stakes are bigger than search. This episode breaks down how assistant ads change trust, intent, and what “sponsored” should look like when users feel like they are talking to a partner, not a feed. Hear how Google is pushing shopping ads into Search AI Mode, what that means for creators and affiliate content, and why attention tools like...

Gemini 3.1 Pro, Claude Code Security, and Meridian Scenario Planner hit the marketing stack all at once. Hear what “reasoning and agents” actually mean for real workflows, why pass at one is just first draft hit rate, and how to judge AI benchmarks without getting hypnotized by hype. Learn how Claude Code Security found hundreds of zero days, what that means for shipping marketing code safely, and...

AI powered digital humans just leveled up with Phoenix 4 from Tavus, Cara 3 from Anam.ai and new multilingual voice cloning from Gnani.ai. This episode digs into what actually breaks when you try to ship real time video agents from latency spikes and audio desync to brittle tool calls and flaky approvals. Learn where avatars truly outperform landing pages and chat, how to test conversion impact, a...

SkyReels V3 R2V 14B is getting world model hype for image to video, but how ready is it for real marketing workflows. This episode breaks down practical tests for believable motion, where brand fidelity breaks, and how to keep image to video ads on brief and on model. Then the focus shifts to OpenAI’s Codex macOS app, parallel agents, worktrees, and how to spend higher limits on QA instead of chao...

Grok 4.20 ships a native multi-agent setup and suddenly “captain plus specialists” moves from meme to workflow. Learn how real teams can use agent teams to kill re-prompts, clean up UTM chaos, and automate painful QA while still keeping a human in the loop. Then zoom out to Ant Group’s open source InclusionAI Ling 2.5 1T with million token context and what long context plus tool use actually chang...

Google is dropping Lyria 3 into Gemini so creators can spin up polished thirty second music beds from text, photos, or video. That shifts brand audio from rare asset to promptable commodity and puts taste and sonic identity back in the spotlight. Meta’s SAM Audio turns messy recordings into usable clips with promptable sound separation, supercharging UGC, podcasts, and ad cleanup while raising new...

Qwen 3.5, Aurora Alpha, and Kani TTS 2 are reshaping what creators and marketers can automate. Hear how open weights change lock-in, why benchmarks often lie about real-world performance, and which agent workflows are actually safe to trust with campaigns. Get a practical due diligence checklist for mystery models, ideas for low-cost internal tools, and clear guardrails for ethical voice cloning....

Alibaba’s Qwen3.5-397B-A17B is out with open weights, Apache 2.0, multilingual chops, and lots of agent talk. This episode breaks down what “open” really means for infra, compliance, and tool use inside real marketing workflows. Then it zooms into JD.com’s JoyAI-LLM-Flash for cheap, fast content ops, and where small models quietly rack up quality debt. Finally, it looks at Rescripted’s Clara as a...

Open AI models keep getting better, but the real plot twist is infrastructure. This episode breaks down Qwen Image 2.0 and why pairing it with Zvec matters for search, reuse, and brand safety. Seedream 5.0 Lite brings fast, visual reasoning inside ByteDance tools and what that means for creators living in CapCut. FoundationVision Alive adds local video plus audio and raises new questions about pri...

ByteDance drops Seed 2.0 and Seedance 2.0 and everyone starts rethinking multimodal workflows and legal risk. Hunter and Riley dig into what “quad-modal” really means for creators running text image audio and video through one model and how ultra cheap tokens change the economics of content meat grinders in a good way. They unpack the Disney cease and desist energy around AI video likeness and IP...

Google DeepMind’s Gemini 3 Deep Think, OpenAI’s GPT 5.3 Codex Spark, and UniAudio 2.0 all dropped with big promises about reasoning, speed, and unified audio. This episode breaks down what those claims actually mean for real teams shipping campaigns and products. Learn how to test “real” reasoning, when to pay for closed frontier models versus lean on open weights, and how to build a role based co...

GLM 5, Emu 3.5 Image, and Google’s AI Mode for shopping collide in one very stacked news cycle. Hunter and Riley break down how to judge new open weight models beyond hype screenshots, what “agent native” really looks like in a production workflow, and where the liability traps hide. They dig into Emu 3.5 Image for in-house creative factories, including typography pitfalls and smart templated work...

Google Gemma 3 lands in everyone’s stack and suddenly “open weights” is the main character. Hunter and Riley unpack what open actually buys creators and marketers, how to use long context as a bigger whiteboard instead of a hallucination machine, and when small models are perfect for low risk, high volume content. They break down Gemma 3 on Together AI, the Llama 5 Avocado leak chatter around ultr...

Aurora Alpha, Runway Gen 4.5, and AudioControlNet show how the AI stack is splitting into three lanes at once. Get practical about using mystery free models for coding without leaking your strategy, and how to treat new tools like a staging environment instead of production. Learn where Runway’s storyboard panels actually fit in a real creative pipeline, plus the review gates needed before AI vide...

Kling 3.0 is pushing AI video from one-off clips to true multi-shot sequences with subject consistency that actually survives the edit. This episode breaks down how marketers can use it for shippable previs, faster storyboards, and compliance friendly variations without reshooting everything. Then it shifts to Google Gemini 3 as the new default assistant and how to stop it from turning your brand...

ChatGPT is officially an ad surface now, and that changes how marketers think about intent, recommendation, and trust. This episode breaks down what good ChatGPT ads should look like, how to measure performance inside a conversation, and why “assistant friendly truth” is the new SEO. Then it zooms out to DeepSeek V3.2 and agent workflows you can actually trust, plus Meta’s Andromeda and Lattice sh...

ByteDance Seedance 2.0, Adobe Firefly unlimited generations, and OpenAI Frontier are all attacking the same thing: faster content and tighter control. Hunter and Riley break down how Seedance inside CapCut could turn short form creators into motion studios, where open source video really stands today, and why “generate a stack of moments” beats one click ads. They unpack Firefly’s unlimited window...