AI Marketing

Sign up free for NitroSend via this link Marketing dashboards were built for humans to click through. George Hartley thinks AI agents will make most of that clicking unnecessary.

In this episode of AI Marketing, Mark Fidelman talks with George Hartley, co-founder of Nitro Send, about the shift from traditional marketing software to agent-first systems that work directly through Claude, Codex and other AI assistants.

George explains why his team designed Nitro Send with the agent as the primary user and the human interface as the backup. The result: 98% of the activity on the platform last month was completed through AI agents rather than its dashboard.

The conversation goes beyond email. Mark and George discuss what happens when CRMs become little more than databases, agents monitor buying signals automatically, follow-ups write themselves and marketers stop spending half their day bouncing between browser tabs.

What you'll learn

• Why traditional marketing dashboards may become largely irrelevant

• How AI agents can build complete email lifecycle campaigns in minutes instead of weeks

• Why most CRM software could eventually become a background data layer

• How agents can enrich leads, identify buying signals and draft follow-ups

• Why outbound email requires different safeguards than newsletters and transactional email

• How Nitro Send handles customer data and reduces the risk of inaccurate results

• What goal-based marketing could look like when agents optimize campaigns autonomously

• Why George believes much of traditional SaaS will be pushed behind the AI assistant

Episode highlights

• 00:31 George's background and the launch of Nitro Send

• 01:51 Why existing email marketing tools were painful to use

• 03:52 Building software where the AI agent is the primary user

• 04:16 Why 98% of Nitro Send activity now happens through agents

• 05:41 The problems with AI-powered prospecting and follow-up

• 06:58 Building complete customer lifecycle campaigns with an agent

• 09:47 Creating and redesigning newsletters through a single prompt

• 10:38 The move toward autonomous, goal-based marketing

• 11:43 Will AI agents replace traditional CRMs?

• 14:10 Why expensive CRM dashboards may become unnecessary

• 17:00 Lead enrichment, lookalike audiences and automated prospecting

• 18:57 How Nitro Send approaches AI hallucinations and data accuracy

• 22:01 Product availability, outbound beta and the roadmap

• 23:19 How to try Nitro Send

• 24:39 Connecting your business domain and email

About George Hartley

George Hartley is a repeat technology founder and co-founder of Nitro Send, an agent-first email marketing platform designed to run through tools such as Claude and Codex. He previously helped build Australia's largest online art marketplace and co-founded SmartrMail, an ecommerce email automation company that was acquired in 2022.

Try Nitro Send

Sign up free for NitroSend via this link or ask your AI agent to install Nitro Send. The platform offers a free account that includes up to 5,000 emails.

George is also on X at @GTHartley.

Get AgentShock

AgentShock: Surviving and Profiting from the Rise of AI Agents and Robots, by Jas Dhillon and Mark Fidelman, explains what happens as AI moves from answering questions to completing work, directing workflows and running entire business functions.

Learn more and get the book at Agentized.com/book.

Episode Show Notes: AI in Real Estate Leasing with Ran Fridman

Episode: AI Marketing Podcast — AI in Real Estate Leasing Host: Mark E. Fidelman Guest: Ran Fridman, founder of MyRemotely Website: myremotely.ai

Episode overview

Ran Fridman joins Mark to discuss how the challenges of managing rental properties across multiple states led him to build an AI-powered leasing and property-management workflow. The conversation covers self-guided property tours, identity verification, connected lockboxes, automated lead follow-up, and how human staff and landlords remain involved in application decisions.

What they discuss

• Ran's real estate journey [05:06–07:00]: Originally from Ukraine, Ran moved to Chicago after living in China for eight years. He began investing by purchasing and renovating six foreclosures on Chicago's South Side, then expanded his personal portfolio to 22 units across Illinois, Indiana, and Wisconsin.
• The challenge of remote showings [07:13–09:54]: With properties in three states, Ran couldn't personally handle every showing. After struggling to find a property manager suited to the logistics, he and his wife started their own technology-focused property management company.
• How the leasing system works [10:02–11:04]: Ran describes an AI agent that responds to leads from platforms such as Zillow and Apartments.com, reportedly in under 30 seconds, and helps prospects schedule self-guided tours.
• Identity verification and property access [08:00–08:41; 11:42–12:29]: Before a tour, prospective renters verify their identity using identification and face recognition. Ran says the system uses a connected digital lockbox to provide time-limited access, and that visitors are notified that tours are monitored by video and audio.
• After-tour follow-up [13:44–15:08]: The system can follow up by text, email, or phone. Interested prospects can receive an application link; those who aren't interested can be asked for feedback about the property.
• Human review and landlord decisions [15:08–15:59]: Ran says staff monitor the process and review submitted applications. The landlord retains the final decision on applicants.
• From internal tools to a public product [16:36–17:56]: Ran explains that the software and lockbox were built to solve problems in his own business and are now being offered to other landlords and property managers. He directs interested listeners to myremotely.ai.

Key takeaways

• Remote leasing can help property owners and managers serve prospective tenants across different locations and schedules.
• Ran's workflow combines AI lead response and follow-up with identity checks, self-guided tours, and a connected lockbox.
• Automation supports the leasing process, but people remain involved: staff review applications and landlords make the final approval decision.
• MyRemotely's tools grew out of operational challenges in Ran's own rental portfolio and property management business.

Guest: Nikhil Vimal Host: Mark Fidelman

Starting an e-commerce company used to mean assembling a small army: researchers, marketers, content creators, media buyers, customer-support reps, and operations people. AI has changed that equation.

In this episode, Mark Fidelman sits down with AI strategist, entrepreneur, and author Nikhil Vimal to map out how they would build an e-commerce business today. Their playbook starts with AI-powered market research, moves through content creation and campaign execution, and closes the loop with monitoring systems that learn from every result.

The bigger opportunity is a connected system of AI agents that can watch competitors, spot cultural moments, create and repurpose content, re-engage customers, and improve over time—with a human responsible for every important approval.

Mark's take: Founders who use AI for isolated tasks will save time. Founders who build a learning system around it will build an advantage.

## What You'll Learn

- Why AI agents can replace much of the early infrastructure once required to launch a direct-to-consumer brand - How to automate competitive research across promotions, ads, creators, partnerships, and adjacent industries - How AI can turn market intelligence into seasonal campaigns, product images, video, copy, and content calendars - Why "clipping" and content repurposing can turn 20 original assets into 100 distribution-ready pieces - How to react quickly to viral cultural moments without forcing a founder to live on social media - Why every agent still needs a human owner who approves consequential work - How observability and monitoring help the system learn from wins, misses, and customer behavior - Where AI voice agents fit into 24/7 customer support - Why Shopify, Perplexity, Claude, ChatGPT, HubSpot, Buffer, and Zapier are practical places to begin - What Zapier still does well: visual workflow design, collaboration, connection monitoring, and operational visibility

## Episode Chapters

- **00:00** — Welcome and Nikhil Vimal's background - **01:42** — The challenge: build an e-commerce company with today's AI tools - **02:12** — AI agents as an on-demand marketing and growth team - **04:34** — Step 1: continuous research and market analysis - **06:02** — Learning from winning ads and adjacent industries - **07:16** — Step 2: campaign preparation and generative content - **08:22** — Why a human must own every AI agent - **09:08** — Step 3: execution, outreach, UGC, and content calendars - **09:52** — Scaling content through clipping and repurposing - **10:53** — Using AI to react to viral moments in real time - **13:32** — Step 4: observability, monitoring, and continuous improvement - **14:54** — The central lesson: build systems, not scattered automations - **15:54** — AI voice agents and the customer-support opportunity - **17:35** — Where a new e-commerce founder should start - **19:19** — The practical starter stack - **20:25** — Why Zapier still matters in the era of connected LLMs - **23:00** — Final lessons and a preview of agentic AI at enterprise scale - **24:18** — Where to find Nikhil and his forthcoming book

## The Operating Model

1. **Research:** Track competitors, adjacent categories, creators, ads, promotions, and market shifts. 2. **Prepare:** Convert that intelligence into campaign concepts, product visuals, video, copy, and calendars. 3. **Execute:** Publish, distribute, follow up with customers, recruit creators, and repurpose the best material. 4. **Observe:** Measure results, identify why something worked or failed, and feed those lessons back into the next cycle. 5. **Approve:** Keep a named human accountable for the decisions that can affect customers, pricing, reputation, or revenue.

## Tools Mentioned

- Perplexity for research and market analysis - ChatGPT and Claude for preparation, execution, and connected workflows - Shopify for commerce infrastructure and educational resources - Zapier for visual automation, workflow orchestration, and monitoring - HubSpot for marketing and customer operations - Buffer for content calendars and publishing workflows - Notion, ClickUp, and Asana for project and workflow management - Midjourney, Higgsfield, and Seedance for image and video creation - Meta Business Suite and Marketo for marketing operations and analytics

## About Nikhil Vimal

Nikhil Vimal is an AI strategist, entrepreneur, investor, advisor, and author who has spent roughly 15 years working across startups, e-commerce brands, agencies, and large enterprises. His forthcoming book, *The Essential Pivot*, focuses on navigating change and separating signal from noise.

Connect with Nikhil on [LinkedIn](https://www.linkedin.com/in/nikhilvimal) and visit [nvimal.com](https://nvimal.com).

## Closing Thought

AI can help one founder operate with the reach of a much larger company. But the advantage doesn't come from collecting more tools. It comes from connecting the work, measuring what happens, and making the system smarter every time it runs.

That's the operating model to build around.

------------------------------ Want all of my insights in one place, buy our book at www.agentized.com/book (Release Nov 2026)

Host: Mark Fidelman Guest: Aksel Bedikyan, marketing research and analytics practitioner based in Montreal, Canada

Episode description

How is AI changing marketing analytics—and what remains valuable when AI can write code, analyze data, and automate reports? Mark Fidelman talks with Axel Benikian about AI agents, the changing role of data science, and practical ways marketers can get more insight from the data they already have.

Axel explains how his team uses AI to automate recurring analysis and presentation work, while emphasizing that useful insights still depend on access to good data. His advice to businesses: map your existing data first, use AI to explore what it can tell you, and turn to custom research when important questions can't be answered with public or existing data.

Key takeaways

• AI is broadening access to data science. Coding may become less central to some roles, while more people use AI to build apps, models, and tools for business questions.
• AI can automate repetitive analytics work. Aksel's team uses Claude to help generate analyses and recurring PowerPoint reports, including brand trackers.
• Start with the data you already have. Map your internal data and connect it to the questions you want to answer before investing in new data collection.
• AI has limits when it comes to proprietary information. LLMs can help analyze available data, but they don't automatically have access to a company's private customer or business data.
• Custom research still matters. Interviews, focus groups, ethnography, and surveys can fill gaps when existing data can't answer new questions.
• Learn by doing. Aksel recommends experimenting with AI tools and sharing what you learn with others.

Episode timestamps

• 00:27 —Aksel's background in aerospace engineering, consulting, and market research
• 02:36 — Will AI replace data scientists?
• 05:25 — AI agents: risks to jobs and opportunities for innovation
• 07:05 — How Aksel's team applies AI to marketing analytics
• 09:54 — Automating recurring analyses and PowerPoint reports
• 11:46 — How marketers can make better use of existing data
• 15:13 — When AI needs to be supplemented with custom research
• 15:47 — Staying curious and learning through hands-on experimentation

Links

• Mark Fidelman's book, Agent Shock: agentize.com/book
• Aksel Bedikyan: OG Insights · Find Axel on LinkedIn

Stop Chasing AI Tools: What Marketers Should Actually Do With AI

AI marketing is moving fast and the constant stream of new tools can leave even experienced marketers feeling behind. In this episode, Mark Fidelman talks with marketing educator Will Francis about how to put AI to work without getting swept up in the hype.

Will shares why he recommends starting with one primary AI assistant, how desktop apps can help marketers work directly with their files, and why he prefers practical workflows over complicated agent hierarchies. He also walks through his own AI-powered dashboard, explains how he uses voice dictation throughout the day, and discusses where human judgment still matters.

The conversation also covers Will's AI for Marketing community, which he said was planned to launch in mid-October at the time of recording.

Key takeaways

• Start with one main AI workspace instead of trying every new tool.
• Look for useful outcomes—like organizing documents or creating a daily briefing—before building a complex AI setup.
• Treat AI agents as assistants that need direction and review; left unchecked, they can produce "work slop."
• Connect AI to the systems and data you actually use if you want a more complete picture.
• Make room for human judgment, strategy, and leadership—and step away from AI during the day.

In this episode

• 00:04:20 Will's background and his practical approach to AI education
• 00:07:29 How to avoid getting overwhelmed by AI tools
• 00:10:30 Will's tool recommendations for marketers
• 00:12:30 Why he's skeptical of complex, always-on agent systems
• 00:18:02 Practical ways CMOs can get started with AI
• 00:22:09 AI as a sounding board—and the continuing role of human judgment
• 00:25:15 Will's daily AI workflow and custom dashboard
• 00:30:30 The AI for Marketing community and workshops Sign up for my book AgentShock

About Will Francis

Will Francis is a marketing educator and creator with 25 years of experience, including work at MySpace, advertising agencies, and his own agency. He creates practical, no-hype content to help people use AI for marketing.

Learn more: willfrancis.com

AI Marketing Podcast – AEO (Answer Engine Optimization) with Elina Panteleyeva

Guest: Elina Panteleyeva, Founder of ShowUpWithAI Host: Mark Fidelman Topic: How to rank in AI answers (AEO) and capture high-intent leads from ChatGPT, Claude, Gemini, etc.

Episode Overview

In this episode, Mark and Elina break down Answer Engine Optimization (AEO)—how to become the brand that AI agents and chatbots recommend first.

Elina shares:

• Why AI-driven leads convert 4–12x better
• How to optimize your website, content, and off-site signals for AI
• What's changing from traditional SEO to AEO/GEO
• Real-world case studies of brands exploding inbound leads in 15–60 days
• Why agentic commerce is the next big wave—and how to get ahead of it

Timestamps & Segments [0:02:46–0:04:52] Warm-up & context

• Light banter about calls and busy schedules.
• Mark mentions bringing a company public (Exascale Labs) and dealing with heavy legal sign-offs on marketing.
• Sets the stage that this is an AI + marketing focused audience.

[0:04:52–0:08:59] Guest intro & origin story

Your setup and question:

• 0:04:41 Mark: "All right, so we're going to talk about the AEO… I'm going to be the voice of the listener and walk through it as if you were talking to me for the first time."

Key points:

• 0:07:39 Elina's background:
  • Founder of ShowUpWithAI
  • Worked at an AI startup, got laid off in 2023.
  • Bootstrapped an e‑commerce brand to seven figures.
  • Noticed people using ChatGPT to choose doctors, therapists, services.
• Realization:
  • "There's no page two of chat."
  • She spent hundreds of hours studying documentation, signals, and how LLMs evaluate brands.
  • That research became ShowUpWithAI

[0:08:59–0:12:13] Why AEO matters now

Your question:

• 0:08:59 Mark: "Okay, so why should anyone care about AEO?"

Key points:

• People coming from AI/chat are 4–12x more likely to convert.
  • They've already discussed their problem with an AI.
  • They trust the model and are now ready for a solution.
• AI is always recommending someone—you or your competitor.
• It's still early:
  • Most businesses aren't optimizing for AI yet.
  • Results can come within 30–60 days.
• Elina: AEO is "the biggest ROI you can get in 2026."

Your follow-up question:

• 0:10:13 Mark: "What if 1,000 plumbers are doing this? How does ChatGPT know which one to recommend?"

Key points:

• Right now, 1,000 plumbers are not doing it, which is why starting now matters.
• LLMs care deeply about recommendation quality:
  • If they recommend bad options, users switch models (Claude, Perplexity, etc.).
• Models use specific signals different from traditional SEO:
  • On-site: technical optimization, schema, phrasing, blogs, structure.
  • Off-site: Reddit threads, YouTube, reviews, third‑party mentions.

[0:12:13–0:15:05] Signals, platforms & AI-optimized content

Your question:

• 0:12:13 Mark: "Is that primarily how you rank… or is there something else you have to do to make sure you're top of mind, especially locally, for ChatGPT, Gemini, or Claude?"

Key points:

• It's industry-dependent:
  • Some spaces are Reddit-heavy.
  • Others lean on YouTube, Trustpilot, or specific review sites.
• Local AEO:
  • Google Business Profile + reviews are critical.
• Website & content structure matter a lot:
  • Blogs should be AI-optimized, not "typical" blogs.
  • Key elements:
    • Summary at the top
    • FAQs at the bottom
    • External citations to credible sources
    • Internal linking: blogs → service pages; services → blogs
• There are "100+ more" technical and structural factors behind the scenes.

Your next question:

• 0:13:51 Mark: "So it's no longer the link-sharing game with SEO or trying to get links in high-traffic sites?"

Key points:

• Big difference vs. traditional SEO:
  • Backlinks still help, but…
  • AI can infer trust from unlinked brand mentions.
• Example:
  • Someone mentions your brand on Reddit without a link.
  • If your website is properly set up, AI can:
    • Recognize the entity
    • Attribute that mention back to your site and brand
    • Treat it as a credibility signal

[0:15:05–0:18:07] Search behavior shift & Google's AI overview

Your question:

• 0:15:05 Mark: "Do you have any breakdown of how many people are doing Google searches vs AI now… it used to be like 99/1?"

Key points:

• Tracking is tricky:
  • People find you via chat/AI then go Google your brand.
  • In Google Search Console, you'll see a spike in branded searches if AEO is done right.
• Behavior changes:
  • People still use Google, but:
    • They rarely scroll far.
    • They definitely don't hit page two.
  • The new game:
    • "How do you get into that Google AI overview box?"
    • That's where the real battle is now.

Your commentary:

• You share your own behavior:
  • You now start most product/service discovery via ChatGPT/Claude/Gemini, not Google.
  • You trust AI to synthesize reviews, source data, and filter out "gamed" SEO.
• Strategic takeaway:
  • If you're paying 30% to Amazon vs 0% on your own site, AEO is crucial.
  • Do "whatever you can" to rank high in AI results.

Your next question:

• 0:17:06 Mark: "Is there any software that tells you where you rank in these AI systems? With SEO there's a lot—what about AEO?"

Key points:

• There are tools for AI visibility tracking.
• The real bottleneck:
  • Implementation is like a 30-hour/week job:
    • Ongoing optimization
    • Technical/content updates
• You must define the prompts/queries you want to rank for:
  • Tools need a list of target prompts.
  • If you choose the wrong prompts, the data is misleading.

[0:18:07–0:20:30] Prompt strategy & why "just use an agent" doesn't work

Your clarification/question:

• 0:18:07 Mark: Paraphrased – So you basically have to reverse engineer prompts back into FAQs, summaries, ranking signals, etc. And if you don't, it gets hard unless you focus on a few phrases.

Key points:

• Elina's tip:
  • Scrape Reddit for your industry.
  • Use AI (e.g., Claude) to extract pain points and questions.
  • Those questions are likely the same queries people ask AI models.
• This gives you a starting list of prompts to optimize for.

Your provocative question:

• 0:18:46 Mark: "Why shouldn't I just have an AI agent do all this for me? Why wouldn't I get 100 agents to post as fake customers on Reddit?"

Key points:

• Reddit hates self‑promotion—especially from new accounts.

• There's a case study:

  • A casino tried something similar.
  • Result: their entire domain got banned.

• Platforms now aggressively detect:

  • Bot-like behavior
  • Coordinated fake posts

• Beyond Reddit:

  • You must know:
    • How to script YouTube videos so AI will cite them.
    • What to say, when to say it, and how to structure them.
    • How to format blogs and web pages exactly for AI.

• If you simply tell an LLM:

  "Optimize me for AEO," you'll get a basic, shallow implementation that everyone else has— not the nuanced, multi-signal strategy that actually moves the needle.

[0:20:30–0:23:27] Case studies & ROI

Your question:

• 0:20:30 Mark: "Do you have any examples of companies or individuals you've helped and what the results were?"

Key case studies:

1. Eight-figure franchise brand
   • Start: ~9% visibility for their target prompts.
   • After ~15 days: 55% visibility.
   • Lead impact:
     • Went from 1–2 inbound leads per week
     • To 3 inbound leads on a single Saturday (within ~20 days).
2. 20‑year-old marketing agency
   • Was getting ~1 lead every other day.
   • After AEO work:
     • 4–5 inbound leads per day.
     • And from industries they'd never seen inbound from before.

Your follow-up question:

• 0:21:56 Mark: "How long did it take these companies to set things up so leads started showing up?"

Key timing details:

• Franchise:
  • Started ~20 days before the results she cites.
  • Saw big movement in 15–20 days.
• Agency:
  • Saw strong inbound increases in ~35–40 days.
• Elina's positioning:
  • She typically says 30–60 days to see increases.
  • If your customer LTV is > $10k, AEO is a "no‑brainer."
  • AEO work builds durable assets:
    • YouTube videos
    • Content
    • Brand signals
  • Those assets pay off beyond AI search itself.

[0:23:27–0:24:28] AEO vs SEO: where should budgets go?

Your question:

• 0:23:27 Mark: "Should people stop spending money on SEO and focus everything on AEO? What do you tell companies that hire you?"

Key points:

• Elina:
  • Her company focuses on AEO, and good SEO is a byproduct.
  • Contrast: traditional SEO shops slap on "AEO" at the end.
• Her thesis:
  • AI is here to stay and expanding.
  • Agentic commerce is coming:
    • Personal agents that know you better than you know yourself.
    • They'll make purchasing decisions automatically.
  • If AI/agents can't find you, you're not even in consideration.
• Therefore, primary focus should be:
  • Optimizing to show up in AI search as the recommended option.

[0:24:28–0:26:41] Agentic commerce & future state

Your commentary:

• You agree agentic shopping is inevitable.
• Likely path:
  • Starts from LLMs → evolves into persistent agents.
  • They'll look at:
    • What people like you have bought.
    • High‑rating clusters.
• Implications for sellers:
  • Need to provide great products and service.
  • Need strong public signals: reviews, mentions, satisfaction.
• You tie this into your upcoming book:
  • "Agent Shock" – coming out next month.
  • This AEO focus is one of the core recommendations.

Your question to close the AEO topic:

• 0:25:27 Mark: "Is there anything we didn't cover that we should have?"

Key points:

• Elina reiterates:
  • Best time to start was 3 months ago, second best time is now.
  • This isn't fake scarcity; someone in your industry is already being recommended by AI.
  • ROI is currently very high and very fast, like SEO in the early 2000s.

[0:26:19–0:27:21] Future episode idea: ChatGPT ads vs AEO

Your suggestion / question:

• 0:26:19 Mark: You propose a future episode on:
  • ChatGPT ads
  • How to take advantage of them
  • How they interact with organic AEO
• Note that you're just starting to see ads and know nothing about pricing or mechanics yet.

Elina: Open to coming back for that topic.

[0:27:21–0:29:07] Post-show wrap-up

• You explain your goals with the podcast:
  • Learn what's happening in your industry.
  • Meet potential customers, especially people needing compute / data centers.
  • Talk about:
    • AI infrastructure
    • Data center boom vs actual deployments
    • Security topics like the Hugging Face hack
• Quick tangent on:
  • Novel data center ideas (e.g., ship in the water using the ocean for cooling).
  • The gap between announced and actually built data centers.

What happens when an AI agent stops answering questions and starts running the business?

In this episode of AI Marketing, Mark Fidelman talks with veteran marketing automation expert Gary Henderson about Luna, the AI agent his company built to handle podcast outreach, customer service, community management, content creation, social media, direct messages and marketing funnels.

Luna has contacted approximately 700 podcasts, booked Gary on more than 20 shows and opened conversations with dozens more. She also replaced a seven-person concierge team, now resolves roughly 95% of customer-service issues without human intervention and communicates with customers around the clock in their preferred language.

Gary explains how he went from managing a 12-person operation to running the business largely by himself, with AI handling much of the execution.

In this episode:

• How an AI agent booked more than 20 podcast appearances • Why AI agents are becoming operators, not just assistants • Which marketing and customer-service tasks can already be automated • How to give agents enough freedom to act without losing control • Why the perfect marketing funnel may soon disappear • How AI will personalize every customer journey • Why niche creators may outperform traditional advertising • How professionals can stay employable as agents and robots become more capable

Mark and Gary also discuss the larger shift ahead: businesses will soon move from individual AI tools to coordinated fleets of agents capable of completing entire workflows. The people and companies that learn to direct those systems will have an enormous advantage. Those that wait may discover the tide has already reached them.

About AgentShock

AgentShock: Surviving and Profiting from the Rise of AI Agents and Robots, by Jas Dhillon and Mark Fidelman, is a practical guide to staying one rung ahead as AI advances from simple tasks to workflows, agent fleets and physical robots. It explains what is coming, what it means for your career and business, and how to turn disruption into opportunity.

Learn more and get the book at Agentized.com.

Episode Description

In this episode of AI Marketing Today, host Mark Fidelman sits down with legendary copywriter and AI entrepreneur Jon Benson (benson.ai) to unpack what it really takes to:

• Make AI sound human
• Clone your brain into a 24/7 digital worker
• Use specialty AI to out-market and out-sell your competitors

Jon shares how he went from fitness author to inventor of the Video Sales Letter (VSL), why generic LLMs will never be enough for serious marketers, and how his "You Cloned" system builds AI versions of yourself that can consult, write, and think like you—about 95% as well as the real thing.

We also dive into mirror language, specialty AI vs. consumer AI, and why your job isn't at risk from AI—but from the person who knows how to use AI better than you.

Guest

Jon Benson

• Get his program here: https://bnsn.ai/pod
• Founder of benson.ai – specialty AI for copywriting and marketing
• Creator of "You Cloned" – digital brain clones for consulting & copy
• Long-time direct response copywriter and creator of the Video Sales Letter (VSL)

Resources from Jon:

• Website: benson.ai
• Cloning PDF & tools: youcloned.ai (free resource mentioned in the episode)

Host

Mark Fidelman AI & marketing strategist, podcast host of AI Marketing Today, co-author of Agentized.

Key Topics & Timestamps

[00:02:25] – Jon's Origin Story

• From fitness author (Fit Over 40) to copywriter
• Accidentally inventing the Video Sales Letter (VSL)
• Early experiments with AI copy in 2010 and why 2017–2018 AI felt "harder than putting a rover on Mars"

[00:05:11] – Why AI Still Struggles to Sound Like You

• The gap between "decent AI copy" and a true voice clone
• Why current models still have tells and feel slightly off
• The distinction between sounding human vs. sounding like you

[00:05:31] – You Cloned: Digital Brain, Not Just Deepfake Video

• Jon's "You Cloned" offer: a digital worker that thinks and talks like you
• Training on:
  • 20+ years of emails
  • Courses, funnels, and consulting calls
• Why this requires offline / private models (Claude Code, Codex, etc.) pulling directly from your own files

[00:07:38] – Voice vs. "Voice" (Literal and Metaphorical)

• Spoken voice: 11Labs voice cloning, tricks to make it sound more human
• Strategic voice: how you think, answer questions, and advise
• Why Jon focuses on behavior and reasoning more than perfect audio mimicry

[00:08:35] – Recording Your Life vs. Recording Your Expertise

• Wearables like Limitless that record everything 24/7
• Jon's more practical approach:
  • Clone the consultant part of himself (his $5,000/hour brain)
  • Focus on marketing and copy instead of his entire personality

[00:11:32] – Imperfect Humans, Imperfect AI, and Unrealistic Expectations

• Why expecting AI to be perfect on the first draft is delusional
• Human copywriters also revise their first drafts—AI should be held to the same realistic standard
• The "microwave generation" mindset vs. real creative work

[00:15:08] – You Won't Lose Your Job to AI…

• You'll lose it to someone who knows AI better than you
• Importance of:
  • Staying on top of new models and tools
  • Understanding specialty AI, not just big general LLMs

[00:16:18] – Agentic AI & Complex Workflows

• Jon's system is fully agentic and has been for over a year
• Over 70 agents inside Benson for:
  • Big ideas
  • Viral hooks (e.g., Firecrawl searches)
  • Copy variants and consulting flows
• Difference between "has agents" and truly orchestrating specialized agents with tools and knowledge bases

[00:20:09] – Should You Even Clone Your Voice?

• When you shouldn't clone your literal voice
• Why voiceover talent (or 11Labs presets) can outperform many founders
• The real gold: your buyer's voice, not yours

[00:20:52] – Mirror Language: The Secret Weapon in Copy

• Mirror language = what your buyers secretly say to themselves in the mirror
• How Jon discovered his Fit Over 40 buyers were:
  • Average age: 67
  • 72% female
• Why that mismatch in avatar and language still sold—but could've sold 10x more with accurate mirror language

[00:24:35] – Repel Most People, Attract the Right Ones

• Jon's contrarian philosophy:
  • Don't sell to everyone
  • Repel bad-fit buyers
  • Create rabid fans who share you organically
• Speak your values clearly so the wrong people self-select out

[00:28:06] – Segmenting by Sophistication (Beginner vs. Expert)

• Newbie copywriters vs. advanced copywriters
• Why jargon like "hypnotic flow" or "reverse NLP" alienates beginners
• Analogy: talking differently to your 10-year-old vs 18-year-old without changing who you are

[00:29:51] – How the Clone Adapts to the User

• Layered questioning to determine:
  • Who the user is
  • What level they're at
  • How to pitch or advise them accordingly
• Specialty AI as a curated expert system, not "go read the whole internet"

[00:31:44] – Specialty AI vs. Consumer AI (Clive vs. Dr. Expert)

• The "Clive in his mom's basement" analogy vs. a 20-year fasting expert
• Why specialty AI with true IP will beat generic models in high-stakes niches
• Jon's stealth move:
  • Creating training only for his AI, never sold publicly
  • LLMs can't scrape it—it only lives inside his models

[00:32:41] – The Future: Your IP, Inside Your Own AI

• Corporations and experts locking their proprietary knowledge into closed systems
• Why your unique data + your unique process = the real moat

[00:32:58] – Wrap-Up & Next Steps

• Mark invites Jon back to demo the next version of Benson
• Jon invites listeners to:
  • Visit benson.ai
  • Download the cloning PDF at youcloned.ai
• Agreement to do a follow-up episode after Mark test-drives the new release

How to Reach Jon Benson

• Website - https://bnsn.ai/
• Instagram - https://www.instagram.com/itsjonbenson/
• Twitter - https://x.com/itsjonbenson
• Facebook - https://www.facebook.com/itsjonbenson
• YouTube - https://www.youtube.com/channel/UCPDvmgurd-rym4pyPjviIuw?sub_confirmation=1
• LinkedIn - https://www.linkedin.com/in/jonbenson

(He shares a personal email off-record in the conversation, but it's not in the transcript, so we'll keep it to his public sites.)

My Book

If you're serious about agentic AI, workflows, and turning AI into an actual operator in your business, sign up for our book:

👉 Get "Agentized" at Agentized.com

Learn how to:

• Design AI agents that think in multi-step workflows
• Plug AI into marketing, ops, and sales
• Stay ahead of the people who think "prompting ChatGPT" is a strategy

In this episode, host Mark talks with James Thornton, CEO of Daz 3D / Tafi, about what it really means to build an AI-centric business. James shares how his company evolved from a 3D content and avatar business into a key AI data provider for some of the world's largest tech and gaming brands.

They discuss:

• The shift from scraping web data to bespoke, rights-clean AI training data
• Using AI across product, marketing, and customer service workflows
• Real-world applications in VR, robotics, gaming, and product visualization
• How AI is transforming email marketing and localization
• What the next 3 years of AI adoption will look like inside typical organizations

Key Topics & Timestamps

[0:00:00] – Introduction & Guest Background

• Mark opens the show and frames the topic: AI-centric businesses with a marketing slant.
• James introduces himself and his company Daz 3D / Tafi.
• His background:
  • CEO at Daz for over a decade
  • Previous role growing a creative product brand from near zero to a $1B retail brand in 3.5–4 years
  • How Daz has continually pivoted into AI trends

[0:01:18] – What It Means to Be an AI-Centric Business

• James defines AI-centric as "AI-first" thinking across the business, especially in marketing.
• Emphasis on balancing AI with people, expertise, and intuition, rather than AI replacing human judgment.

[0:01:37] – Daz / Tafi's Evolution into AI

• Company roots:
  • 25-year-old business with a strong direct-to-consumer freemium model
  • Deep community of 3D digital content creators (avatars, products for gaming, etc.)
• Past pivots: work in Web3 and NFTs, then parlaying that into today's AI data and services.
• They now both:
  • Support other companies' AI initiatives
  • Use AI internally to streamline content creation and business workflows

[0:03:37] – How They Help Other Companies with AI

• Core specialty: 3D digital content, avatar systems, and digital assets.
• Use AI and procedural tools to:
  • Scale their content library
  • Produce rights-clean, structured, annotated AI training data
• Customers include top tech, gaming, and enterprise brands training foundation models for:
  • Text-to-character
  • Product creation & visualization (including physical product development)
  • Robotics and other AI applications

[0:04:51] – Concrete Benefits of AI for Clients and Daz

• Central thesis: Good AI requires high-quality training data.
• For clients:
  • Bespoke, properly licensed data for training a wide range of models
• For Daz's own business:
  • AI embedded in product workflows and content production
  • AI in product search, customer service, QA, and other functions to drive efficiency and effectiveness

[0:06:07] – AI Agents and Localization

• Use of AI agents in customer service to handle technical tickets and customer questions.
• Heavy focus on localization:
  • AI to adapt website and e-commerce for global markets
  • Integrating translation tools directly into their software so customers can use products in their native language
  • James calls this a "huge lift" for the business.

[0:07:45] – Tools: Open "Claw" and Enterprise AI Options

• They use Open "Claw" (contextually, likely a reference to a general-purpose LLM) through virtual machines across business units, including marketing.
• Acknowledge recent changes in that ecosystem and the need to adapt.
• Considering more enterprise-focused alternatives, such as the "Nemo Claw" and other corporate-friendly solutions, with an eye on governance and parameters.

[0:08:54] – Their Process for Helping Enterprise Clients

• First step: deeply understand voice of the customer and specific AI data needs.
• Industry shift:
  • Moving away from scraping the internet or generic licensing for training data
  • Toward custom, bespoke training sets
• Unique position of Daz / Tafi:
  • Own the full rights to their 3D content library (avatars, scenes, props)
  • Avoid ethical and legal issues around unauthorized training data
• Process:
  1. Scope the client's needs.
  2. Use AI and internal workflows to generate customized datasets.
• Example:
  • For a major tech company building VR headset experiences, they created tens of thousands of unique characters, avatars, clothing, and assets to train their models.
  • James notes they might be the only company able to deliver this level of tailored solution at that scale.

[0:11:34] – The Next 3 Years: How AI Changes Organizations

• Based on discussions with CEOs and senior leaders:
  • Debate continues on whether AI creates or eliminates jobs; James personally believes it will create jobs.
  • Near-unanimous view: every company must integrate AI into workflows.
  • Employees and companies with a clear AI strategy will win versus those without one.
• Expectation:
  • Continued job creation plus a strong push for AI-centric efforts in every part of the business
  • Focus on efficiency, output, and more bespoke customer solutions

[0:13:25] – Advice to Marketers & Final Takeaways

• James emphasizes "just get started" and experiment with AI.
• For marketers:
  • AI enables hyper-specific adaptation to customer needs.
  • Allows new ways of marketing that aren't feasible manually today.
• Example from their own marketing stack:
  • Their largest marketing channel is email
  • They manage a seven-figure, vetted email list mailed daily
  • AI is used to:
    • Optimize send times
    • Customize products and offers by demographic and other signals
• Overall message:
  • Embrace AI to become more effective and deliver better, more tailored customer experiences.

[0:14:58] – How to Contact James

• Website: Daz3D – daz3d.com
• Direct email: [email protected]
• James invites listeners to contact him directly with questions.

In this episode of the AI Marketing Podcast, host Mark Fidelman sits down with David Amar, founder of Makina (a new conference dedicated to physical AI), to explore how robots and humanoids will change the future of marketing.

They discuss why robots are such powerful brand activations, when we might see in‑home humanoid housekeepers, how China is leading on hardware while the West leads on software, and why 2025–2026 feels like the "GPT moment" for physical AI. David also shares what to expect at Makina in Paris on July 7 and why marketers should get ahead of this trend now.

Guest

David Amar

• Background in computer science and neuroscience (UCL)
• Formerly worked in prosthetics
• Founder of Makina, a conference that brings together the fragmented physical AI ecosystem: humanoid builders, robot "brain"/OS providers, capital, industrial partners, and talent

Key Topics & Timestamps 1. Why Robots Are Marketing Gold

[0:00:00 – 0:02:24]

• David's background and the launch of Makina
• Why robots are "premium marketing material":
  • Robots tap into deep cultural fascination (e.g., Star Wars, Star Trek)
  • Simply announcing "Robot X/Y will be on site" can materially boost event attendance
• Humanoids as especially compelling because of their uncanny, human-like form

"I don't think I've ever met somebody that says this isn't interesting… It's just premium marketing material." – David [0:01:31]

2. Timeline: When Physical AI Hits Everyday Life

[0:02:24 – 0:03:25]

• David's long‑range outlook:
  • Short term: impressive demos, but still lots of technical bottlenecks
  • ~10–15 years: expect robots/humanoids in places we never imagined, with deep dependence on them
• Contrast with digital AI:
  • We're already "slaves" to ChatGPT and cloud AI for knowledge work
  • Physical dependence on robots will follow later

3. How Robots Show Up in Marketing (Beyond a Robot at a Desk)

[0:03:25 – 0:06:27]

• Robots won't replace marketers by typing at a desk—that's the realm of LLMs and digital AI
• Instead, robots will act as:
  • Brand avatars and mascots (e.g., "the Amazon robot," "the Walmart robot")
  • Physical activations at events, retail, and public spaces
  • Product demo agents in stores, on the street, or wherever target audiences gather
• Comparison to today's street activations (e.g., sign spinners) but in a far more advanced, interactive form
• Emotional/branding angle:
  • A charming C‑3PO‑style humanoid pitching products can be more captivating than a celebrity

"There's just something more charming about a C‑3PO showing the new Coca‑Cola than just a regular old Joe… even if it's George Clooney." – David [0:05:33]

4. Humanoids vs. "Robots" – What's the Difference?

[0:06:27 – 0:07:55]

• Humanoid:
  • Robot with human‑like physiology and form (height, posture, movement)
  • Tends to get anthropomorphic traits projected onto it
• Robot:
  • Any robotic form, e.g. a single robotic arm, a robot dog, or R2‑D2‑style platforms
• Long‑term: Mark expects humanoids to become increasingly indistinguishable from humans in 20+ years

5. In‑Home Humanoids: How Close Are We Really?

[0:07:55 – 0:11:29]

• West vs. Asia split:
  • West: stronger on software and AI models
  • Asia (especially China): stronger on hardware and shipping units at scale
    • Today you can already order multiple Chinese robot models online and have them delivered within a month
• Current leading players mentioned:
  • 1X – focused on household/housekeeping tasks
  • Sanctuary (Sunday Robotics) and others delivering early trial units
• Reality check on timelines:
  • No one truly knows, but David's informed estimate:
    • 5–7 years to order functional in‑home humanoids online
    • Dependent on breakthroughs in:
      • Fine manipulation of small objects
      • Robust computer vision
      • Autonomous navigation in unmapped environments
• Many "impressive" demos are partly marketing:
  • Used to raise capital, build momentum, and buy time while teams fight through technical bottlenecks

6. Data, Compute, and How These Robots Actually Learn

[0:10:42 – 0:13:16]

• Today's deployed robots are often trial models used primarily to:
  • Collect huge amounts of real‑world data
  • Train the next generation of more capable robots
• Data and compute needs:
  • Humanoids need even more data than LLMs:
    • Touch, force feedback, vision, balance, navigation, etc.
  • Massive compute, similar or greater than what's used for digital AI
• Where the compute lives:
  • Training: in large data centers, often the same infrastructure used for AI
  • On‑device inference:
    • Onboard boards like NVIDIA Jetson inside the robot's "chest"
    • Local models run on-device, optionally connected via Wi‑Fi for streaming data and updates
• Most robots in the wild are still tightly constrained and far from general-purpose autonomy

7. The Makina Physical AI Event in Paris

[0:13:59 – 0:19:11]

• Date: July 7
• Location: Station F, 13th district of Paris (central), the world's largest startup campus
• Format:
  • One‑day dedicated physical AI conference
  • Paired with the RAISE Summit (July 8–9), a broader AI conference
• Makina's mission:
  • Fix the current ecosystem problem where events are:
    • "Tech geeks talking to tech geeks"
    • "Commercial to commercial" with limited cross‑pollination
  • Bring together the full vertical stack of physical AI:
    • Humanoid builders
    • Robot brain / OS providers
    • Investors and capital
    • Industrial partners and adopters
    • Talent and researchers
• Expected scale & hardware:
  • Targeting 1,500+ attendees
  • David's goal: 18–20 robots on site, split between stage demos and exhibitor robots
• Notable participants mentioned:
  • Boston Dynamics (CEO Amanda)
  • 1X (CEO)
  • Google DeepMind Robotics leadership
  • Other leading US, European, and Asian robotics companies

"We really are trying to regroup the very fragmented ecosystem that is physical AI… the vertical stack of physical AI in terms of ecosystems." – David [0:13:59]

8. Why Marketers Should Care (Now, Not Later)

[0:16:43 – 0:18:07] & throughout

• Humanoids and robots as future marketing must‑haves:
  • Likely every major brand will have a robot avatar/mascot within ~10 years
  • Use cases:
    • Product demos & in‑store experiences
    • Public activations and stunts
    • Content creation, fail/reaction videos, and social media hooks
• Strategic advantage:
  • Marketers who understand physical AI early will:
    • Shape the first killer use cases
    • Align brand positioning with new capabilities
    • Avoid being late adopters in a fast‑moving bull run
• David's framing:
  • Physical AI is in a "bull run", possibly a GPT‑moment equivalent for the physical world
  • Huge capital flows, many new robotics startups, and intense industrial interest

9. Robotics Reality Check: Hype vs. Capabilities

[0:19:11 – 0:21:59]

• GTC/NVIDIA event anecdote:
  • Few humanoids present, many robots were wired or constrained
  • Mostly robot dog style units; limited truly autonomous humanoids
• Chinese hardware advantage:
  • Companies like Unitree and others can ship robot dogs today, often ahead of US peers on commercialization
• Software and usefulness still lag:
  • Hardware works, but what you can actually make them do is still narrow
  • Many current units are about data collection more than true deployment
• Industrial partnerships:
  • Examples: hexagon robotics & Mercedes, Figure & BMW, Boston Dynamics & Hyundai
  • Present robots can perform very specific, tightly defined tasks with minimal uncertainty
  • Outside of that, most can still just walk, dance, wave, and pose for marketing

"We're safe, and the Terminator is not coming next year." – David [0:21:59]