The AI Edge Podcast

The New York Times says AI is making digital advertising boring. Meta just posted its biggest earnings quarter ever. Google's CapEx story is looking smarter by the day. And OpenAI is quietly building toward a $100 billion ad business while the three-headed AI monster — Gemini, Claude, and ChatGPT — keeps getting more tangled by the week.
This week, Shiv and Miles sit down with David Danziger, SVP of Partnerships at Dstillery, one of the leading independent targeting companies in advertising. David has spent his career at the intersection of data, partnerships, and the open internet — with stints at LiveRamp, The Trade Desk, and Yahoo — and he brings a practitioner's eye to a week full of big numbers and bigger questions.
David breaks down why the NYT's "boring" framing is ironic at best and wrong at worst, why Meta's CapEx story looks very different from Google's when you factor in the infrastructure play underneath it, and why the open internet is going to see real AI benefits too — not just the walled gardens. He also explains why a wrapper on an LLM isn't enough anymore, why ChatGPT ads are essentially search ads with more granular intent signals, and why the companies with real data assets underneath them are the ones that will win.
They also get into how MCP is replacing what APIs used to do — faster, easier, and already happening in production at Dstillery — why multimodal AI means any input can produce any output and what that actually unlocks for targeting, and how a workflow that used to take two days of back-and-forth emails now takes fifteen minutes inside Dstillery's agentic tools.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Starts May 12th | Use code EARLYBIRDSLEARNAI for 20% off https://uof.digital/ai/accelerator/ Sponsored by Hightouch
New episodes every Monday. Get your AI Edge here.

OpenAI launched GPT-5.5 and called it a step toward a super app. Anthropic dropped Claude Design and quietly kept building what might be today's version of G Suite. Google launched a native agent builder in Gemini Enterprise. And Meta got caught secretly tracking employee keystrokes to train AI — then told the employees who objected: no.
This week, Shiv and Myles sit down with Mark Josephson — three-time CEO, former head of Bitly, and now an executive coach working with founders and senior leaders on how to actually transform their businesses with AI. Mark has a front-row seat to what's working, what isn't, and what separates companies that are going to make it from the ones that are quietly running out of time.
Mark breaks down why Anthropic has a clarity of purpose that OpenAI still hasn't found, why the real Google competitive battle isn't against Claude or ChatGPT but against AWS and Azure at the cloud layer, and why the future is simply binary — companies will either be AI native or they'll be dead, the same way print was always going to die once digital arrived. He also shares the story of a founder who cancelled every meeting on a Monday, handed his entire company Claude Code, and committed to two hours of daily AI training until they got there.
They also get into why Meta secretly tracking keystrokes is a warning sign that should make everyone think harder about who they're trusting with this technology, why smart regulation matters more than the performative policy theater coming out of OpenAI, and what it actually means to be an AI native leader — not the person doing all the building, but the person who understands what's possible.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Starts May 12th | Use code EARLYBIRDSLEARNAI for 20% off https://uof.digital/ai/accelerator/ Sponsored by Hightouch
New episodes every Monday. Get your AI Edge here.

OpenAI wants to be a $100 billion ad business by 2030. Meta just overtook Google as the largest ad company in the world. Anthropic launched a new model, built a design studio, and is now worth more on the private market than OpenAI. And Allbirds — yes, the sneaker company — rebranded as an AI infrastructure company and watched their penny stock spike 600%.
This week, Shiv sits down with David Berkowitz, longtime marketing leader, community builder, and founder of the AI Marketers Guild at Marketecture, to make sense of a week where everything moved at once.
David breaks down why OpenAI's $100 billion ad ambition is a hell of a magic trick that requires nearly tripling their user base, why Meta's flood-the-feed strategy might not be as tenable as their revenue numbers suggest, and why Google — with nine billion-user-plus products and a 15-year head start building AI — is playing a much longer game than anyone is giving them credit for. He also explains why the seed and early-stage AI companies that felt safe six months ago should now be getting nervous, and why the "anti-AI" brand positioning trend is mostly companies talking out of both sides of their mouth.
They also get into what marketers should actually be doing with ChatGPT ads right now, why the organic vs. paid AI debate is just the Google search debate starting all over again, and how the industry has moved from Walmart calling a basic chatbot a "super agent" to an era where your audience is building their own agents and calling your bluff.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Starts May 11th | Use code EARLYBIRDSLEARNAI for 20% off https://uof.digital/ai/accelerator/ Sponsored by Hightouch
New episodes every Monday. Get your AI Edge here.

Sam Altman is either a visionary genius, a pathological liar, or both. OpenAI wants to be a $100 billion ad business by 2030. Anthropic dropped a mysterious model called Mythos that's apparently too powerful for the public to use. And the AEO/GEO industry has officially jumped the shark.
This week, Shiv and Myles sit down with Debra Aho Williamson, founder and chief analyst of Sonata Insights and writer of the fast-growing Substack The AI Ad Economy. Debbie spent nearly two decades at eMarketer building and leading their social media research practice — she was one of the first analysts to identify social media as a force that advertisers needed to take seriously. Now she's doing the same thing with AI.
Debbie breaks down why the organic reach window in AI platforms is closing faster than most marketers realize — and why Facebook's playbook of giving brands free reach and then taking it away is about to repeat itself in AI. She also shares why she bought a product she'd never heard of after ChatGPT recommended it, what that means for brand discovery, and why the explosion of AEO and GEO vendors is a sign the space has already jumped the shark.
They also get into whether OpenAI's industrial policy paper is genuine concern or brilliant PR, why Trade Desk's new Performance Mode is starting to look a lot like the black box ad networks that came before it, and why Time Magazine pitching branded content as an AI visibility play is either genius or a harbinger of chaos — depending on who you ask.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Starts May 11th | Use code EARLYBIRDSLEARNAI for 20% off https://uof.digital/ai/accelerator/
New episodes every Monday. Get your AI Edge here.

OpenAI just raised $122 billion at an $852 billion valuation — and apparently that only gives them 18 to 24 months of runway. They bought a live streaming network for over $100 million. They're building conversational ads with Smartly. Amazon is quietly opting every search advertiser into sponsored prompts inside Rufus. And Walmart is going the opposite direction entirely, building a native app inside ChatGPT.
This week, Shiv and Miles sit down with Joe Root, co-founder and CEO of Permutive, a data collaboration platform that sits at the intersection of media companies, agencies, and advertisers. Joe has a front-row seat to how AI is actually reshaping the ad stack — not in theory, but in the workflows his customers run every day.
Joe breaks down why AI is a tailwind for Permutive by dismantling the old agency margin model, why over half of his customers are now letting agents run tasks on their behalf — up from just 5% eighteen months ago — and why the real shift isn't replacing software, it's changing how humans interact with it.
They also get into why live content is the one thing AI can't replicate and what that means for publishers, why the native ad format for conversational AI is still completely unsolved, and why Amazon's sponsored prompts in Rufus might just be a repackaged version of a search arbitrage play that never really worked the first time around.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Now enrolling for May | https://uof.digital/ai/accelerator/
New episodes every Monday. Get your AI Edge here.

OpenAI killed Sora. Disney pulled its billion-dollar commitment. The headlines called it proof that AI video is dead. It isn't. But the breathless takes — in both directions — are a pretty good sign that marketers are still figuring out how to think about all of this clearly.
This week, Shiv and Tameka sit down with Melissa Grady Diaz, former CMO of Cadillac and now a senior leader at health-tech startup Measured Wellness, to talk about what's actually changing for brand marketers right now — and what most people aren't paying enough attention to yet.
Melissa breaks down why performance marketing is about to have a real problem as agents take over the discovery layer, why brand building outside of paid media is becoming more important than ever, and why you should never fully hand your marketing over to a platform that's optimizing for response instead of brand health. She also gets into what it was like to experiment with AI at the CMO level at Cadillac, why incremental AI adoption can actually make things worse, and the right way to think about building from scratch versus improving what already exists.
They also cover Apple's surprising decision to open Siri up to any AI model, what it means that ChatGPT has become the Kleenex of AI — and why that brand equity might be slipping away — and why agentic commerce is happening whether the skeptics like it or not.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Now enrolling for May | https://uof.digital/ai/accelerator/
Learn more about the AI Literacy Alliance at uof.digital/alliance
New episodes every Monday. Get your AI Edge here.

Mark Zuckerberg says just tell AI what you want to sell and it handles the rest. Brian Tomasette — 25-year ad tech veteran, former Amazon DSP product leader, and current PhD researcher — says that's high school physics without friction. The business rules, privacy constraints, and data ownership that make advertising actually work don't disappear because the model is powerful enough. They just need better tools to operate them.
This week, Miles sits down with Brian to get into the real plumbing behind agentic advertising — what's actually being built today, how far off the utopian vision really is, and why the tools aren't going away anytime soon. Brian is currently researching neuro-symbolic AI at Drexel University and contributing to the IAB Tech Lab's AAMP framework, which is building the buyer and seller agent infrastructure that could eventually reshape how media gets bought and sold.
They break down why AI agents will use DSPs more optimally rather than replace them, what the IAB Tech Lab's buyer and seller agent SDKs actually do, and why Andrej Karpathy is right that it's not the year of agents — it's the decade. Brian also introduces the "AI vampire" phenomenon — the way vibe coding becomes as addictive as TikTok and drains you faster than actual genius work ever could — and why the right move is sometimes just closing the laptop and going for a hike.
Links from this episode:
AAMP — Agentic Advertising Management Protocol https://iabtechlab.com/standards/aamp-agentic-advertising-management-protocols/
IAB Tech Lab Buyer Agent Documentation https://iabtechlab.github.io/buyer-agent/
IAB Tech Lab Seller Agent Documentation https://iabtechlab.github.io/seller-agent/
Andrej Karpathy: "It's Not the Year of Agents, It's the Decade of Agents" https://www.youtube.com/watch?v=BlVnGXEzFow
The AI Vampire — Steve Yegge https://steve-yegge.medium.com/the-ai-vampire-eda6e4f07163
Beads — Steve Yegge's Claude Code library https://github.com/steveyegge/beads
Neuro-Symbolic AI Research Paper https://arxiv.org/abs/2510.05486
HAZ Lab at Drexel — Brian's Research Lab https://zharry29.github.io/hazlab/
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Now enrolling for May | https://uof.digital/ai/accelerator/
New episodes every Monday. Get your AI Edge here.

ChatGPT just paused its buy button. The agentic commerce skeptics are doing a victory lap. Shiv and Myles aren't buying it.
This week it's just the two of them — no guest, no guardrails — working through the biggest story of the week and a mailbag full of questions from the smartest people in the industry. They break down why ChatGPT backing off instant checkout is a strategic pause, not a retreat, why Amazon's real moat has nothing to do with software, and why the iOS App Store is the most useful analogy for understanding where AI commerce is actually headed.
They also get into Amazon's new sponsored prompts inside Rufus, a wave of AI product launches with increasingly questionable names, and why the utopian promise of the open internet is a pretty good warning sign for what's about to happen to AI.
Plus the mailbag: who's best positioned to build the standard for agentic buying, whether walled gardens will eventually swallow AI the way they swallowed the web, and the hair care products powering the AI Edge podcast.
Are you an AI company, agency, or consultancy building for the marketing and advertising industry? Join the AI Literacy Alliance — U of Digital's partner program to help the industry get smarter about AI together. Learn more at https://uof.digital/alliance
And if you're ready to level up your own skills, the AI Accelerator is enrolling now for May. https://uof.digital/ai/accelerator/
New episodes every Monday. Get your AI Edge here.

Meta wants to turn its AI chatbot into a shopping engine — but what happens when a chunk of the catalog is fake? OpenAI just partnered with Criteo to inject ads into ChatGPT, and agencies are vibe-coding their own GEO platforms while unicorns raise $100 million to do the same thing. The commerce media land grab is on, and nobody's quite sure who wins.
This week, hosts Shiv Gupta and Myles Younger bring the conversation inside the house. Their guest is Tameka Kee, U of Digital's new Head of AI Enablement and Strategy — a 20-year ad tech veteran with stints at Rubicon Project, Magnite, and CIMM, who spent her first two weeks on the job going through the AI Accelerator herself.
Tameka breaks down why Meta's AI shopping tool could supercharge the drop-shipper problem and usher in a new era of AI product slop, why OpenAI partnering with Criteo is smarter than it looks (hint: it's as much a data play as a monetization play), and why Google's ecosystem — Maps, Wallet, Chrome, Translate, YouTube, Gmail — is still the greatest monetization machine ever built and ChatGPT has a long way to go. They also get into whether agency AI platforms will survive to 2029, what Stagwell's GEO build is really signaling, and why throwing money at AI adoption without understanding your team almost never works.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Starts March 9, 2026 | Use code PROCRASTINATOR for 25% off https://uof.digital/ai/accelerator/ Sponsored by Yahoo DSP
New episodes every Monday. Get your AI Edge here.

Everyone thought AI would kill the consultancies. Instead, OpenAI just partnered with Accenture, McKinsey, and BCG to form the Frontier Alliance — bringing enterprise AI to the brands that need help figuring out where to start. Turns out, the awkward phase of a new technology wave is exactly when consultancies earn their keep.
This week, hosts Shiv Gupta and Myles Younger sit down with Linda Bethea, Chief Marketing Officer at Danone North America, to talk about what brands are actually doing with AI, why efficiency isn't the same as effectiveness, and what happens when you send your entire C-suite to Seattle to learn about AI from Microsoft.
Linda breaks down how Danone built an AI task force before most companies understood what was coming, why the physical shelf gave way to the digital shelf — and why marketers now need to think about the AI shelf, and the three things every CMO should tackle first when getting started with AI (spoiler: always start with people). They also get into Profound's $96 million raise and unicorn status, a $20 million bet on injecting ads into LLMs, and why Anthropic is pushing back hard on government demands for access to its models and data.
Ready to level up your AI skills? Join the U of Digital AI Accelerator. Starts March 9, 2026 | Use code PROCRASTINATOR for 25% off https://uof.digital/ai/accelerator/
New episodes every Monday. Get your AI Edge here.