Mark Tilbury
Synthetic social media creators can generate immediate net profit on a sub-$100 software budget, proving automated personas convert buyers even as organic algorithms delay peak audience growth.
AI-assisted summary of I Tried The LAZIEST Way to Make Money With AI — original source. Check the original for full context.
Source published:
Key ideas
- Mark Tilbury tested the AI influencer business model over seven days, generating $276 in gross sales and $157.48 in net profit selling a digital habit workbook created by a synthetic centenarian persona. (29:07)
- Social media users actively engaged with the generated influencers as if they were real humans—arguing over dating advice and asking for clothing links—despite profile bios explicitly disclosing their AI nature. (21:51)
- Although Amos (health) showed the strongest early engagement and was chosen for monetization, Vivian (wealth) surged past 250,000 views on the final day, illustrating the unpredictable latency of short-form recommendation algorithms. (29:07)
- Falling short of his self-imposed $1,000 profit benchmark, Tilbury forfeited $1,000 of his own money to a commenter while noting the model has substantial untapped upside if syndicated to TikTok and YouTube Shorts. (32:23)
Detailed notes
## Background Published on YouTube on September 21, 2026, by entrepreneur and creator Mark Tilbury, this video documents a 7-day experiment testing whether virtual AI influencers can generate real income. Tilbury outlines his end-to-end process using tools like Claude and Higgsfield to craft realistic avatars, publish short-form content, and monetize the resulting attention without ever using his own face or voice.
## TL;DR Mark Tilbury undertakes a strict seven-day challenge with a $99 budget cap to build and monetize three AI influencers across high-traffic niches: Amos (health), Grandma Vivian (wealth), and Sienna (relationships). Using Claude for concept generation and Higgsfield for photorealistic visuals and cinematic vertical videos, he launches dedicated Instagram accounts to test organic algorithmic reach. Amos emerges as the early front-runner and is monetized via a Claude-generated 30-day habit-tracking workbook sold on Stan Store. Although an unexpected algorithmic surge later propels Grandma Vivian to over 250,000 views, the experiment concludes with 23 product sales generating $157.48 in net profit—falling short of his $1,000 goal and prompting Tilbury to pay out his viewer forfeit while validating the business model's viability.
## Atmosphere & style The presentation blends fast-paced entrepreneurial experimentation with pragmatic, seasoned skepticism toward online 'get-rich-quick' claims. Tilbury balances rapid-fire software walkthroughs with lighthearted humor—quipping about AI characters looking younger than him and cheering for his extravagant 'bougie boss grandma' persona. Watching the video feels like an energetic, transparent business sprint punctuated by genuine surprise at audience reactions, comedic comment drama, and thoughtful reflections on how automated personas risk draining human connection from social media.
## Bulletin - [29:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1747s) Mark Tilbury tested the AI influencer business model over seven days, generating $276 in gross sales and $157.48 in net profit selling a digital habit workbook created by a synthetic centenarian persona. - [21:51](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1311s) Social media users actively engaged with the generated influencers as if they were real humans—arguing over dating advice and asking for clothing links—despite profile bios explicitly disclosing their AI nature. - [29:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1747s) Although Amos (health) showed the strongest early engagement and was chosen for monetization, Vivian (wealth) surged past 250,000 views on the final day, illustrating the unpredictable latency of short-form recommendation algorithms. - [32:23](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1943s) Falling short of his self-imposed $1,000 profit benchmark, Tilbury forfeited $1,000 of his own money to a commenter while noting the model has substantial untapped upside if syndicated to TikTok and YouTube Shorts.
## Key takeaways - [00:00](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=0s) The experiment tested claims of automated AI wealth under three strict rules: a $99 total budget cap, complete anonymity without personal brand leverage, and launching within 24 hours. - [01:14](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=74s) Auditing established virtual creators identified health, wealth, and relationships as top-performing niches, with personal finance standing out as an overlooked gap. - [04:23](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=263s) Claude brainstormed three contrarian personas: 101-year-old natural healer Amos, luxury matriarch Grandma Vivian, and modern dating commentator Sienna. - [06:36](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=396s) Integrating Higgsfield directly into Claude via a custom connector allowed continuous avatar and environment generation within a single chat workflow. - [11:58](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=718s) Pre-configured Claude skills automated the scripting process, outputting viral hooks and structured dialogue tailored to each character's target audience. - [15:22](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=922s) Higgsfield's Cinema Studio combined synthetic character voices, 9:16 vertical aspect ratios, and model references to maintain visual consistency across speaking clips. - [19:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1147s) Full disclosure in profile bios—such as labeling Amos 'Your AI grandpa'—had zero negative impact on viewer retention or engagement. - [21:51](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1311s) Early audience feedback proved hyper-realistic immersion, triggering genuine arguments between users defending or attacking Sienna's relationship takes. - [23:30](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1410s) Digital products proved to be the most practical monetization strategy within a seven-day window, bypassing the contract delays of brand sponsorships and physical fulfillment. - [25:31](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1531s) Claude authored Amos's 30-day longevity habit workbook and printable tracker, requiring manual editing to ensure practical value and remove AI filler. - [27:15](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1635s) Hosting the guide on Stan Store under a 14-day free trial enabled zero-cost sales fulfillment, generating 23 units sold at $12 each directly from Instagram bio traffic. - [29:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1747s) A delayed algorithmic spike pushed Grandma Vivian's content over 250,000 views on day seven, demonstrating that brief test windows can misidentify long-term winning content. - [29:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1747s) Total costs of $118.52 yielded a net profit of $157.48, proving the unit economics of the model despite missing Tilbury's $1,000 challenge threshold. - [32:23](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1943s) Syndicating identical short-form assets across TikTok Shop and YouTube Shorts represents the most immediate, low-friction pathway to scale AI influencer earnings.
## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=0s) The 7-Day AI Influencer Challenge and Rules
Claims have exploded across the internet of people secretly generating tens or even hundreds of thousands of dollars from the comfort of their own homes by using AI tools like Claude to build virtual influencers. Online creators boast numbers ranging from $2,000 to over $100,000 or $200,000 in monthly profit without ever putting their real face on camera. Many tout it as "the laziest AI side hustle," but Mark Tilbury—a self-made millionaire who built his wealth through traditional business—remains deeply skeptical about whether the business model genuinely works or is merely low-effort "AI slop."
To test whether virtual influencers can generate real cash, Mark gives himself a strict seven-day deadline to set up the system and see how much profit he can produce. He promises to document every step and share the raw, unfiltered financial results. To keep the stakes real, he adds a personal penalty: if he fails to make at least $1,000 USD within the week, he will pay $1,000 of his own money to a viewer in the comment section.
Before starting the clock, he lays down three non-negotiable rules for the experiment: - **Rule 1: Budget Cap** — The entire operation must be launched and operated for under $99 USD. - **Rule 2: Complete Anonymity** — He cannot use his own face, likeness, or voice, ensuring he gains zero advantage from his existing audience or personal brand. - **Rule 3: Rapid Execution** — The first AI influencer must be fully built and operational before the end of the first day.
```mermaid flowchart TD accTitle: AI Influencer Challenge Structure accDescr: Visual summary of the rules, three target niches, and setup workflow for the seven-day experiment.
subgraph Rules["Challenge Rules"] R1["Budget < $99 USD"] R2["No Real Face or Voice"] R3["First Influencer Built Day 1"] end
subgraph Niches["Three Selected Niches"] N1["Health: Amos (101 yo)"] N2["Wealth: Grandma Vivian"] N3["Relationships: Sienna (22 yo)"] end
subgraph Stack["AI Production Stack"] S1["Claude (Ideation & Connector)"] S2["Higgsfield (Avatars & Video Engine)"] end
Rules --> Niches Niches --> Stack ```
[01:14](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=74s) Step 1: Spying on the Competition
Mark kicks off the challenge by analyzing successful accounts across social media to reverse-engineer their positioning and monetization strategies. He starts with a household name in the virtual space: Lil Miquela. Boasting over two million followers, she appears at red carpet events like the Met Gala and the Grammys, positioning herself as a high-fashion supermodel. Her account pulls in an estimated $11 million a year primarily through major luxury brand partnerships with names like Prada and Calvin Klein.
Turning to more specialized creators, Mark finds Mother Satori, a US-based holistic health influencer claiming to be 108 years old ("I thought I was the old one around here," Mark quips). Health is historically one of the most lucrative niches online because audiences will readily spend money to feel healthier and live longer. Mother Satori's unique selling proposition is her extreme age and vitality, which she leverages at the end of every video to pitch a detox and gut-cleanse drink. The strategy works; thousands of daily viewers actively flood her comment section begging for purchase links.
Continuing his audit, Mark spots several other distinct virtual accounts: - **Podcast/Relationship Influencer**: A glamorous creator named Olivia sharing relationship advice and dating dilemmas in short-form podcast clips. - **The Patriotic Craftsman**: An AI-generated British woodworker shown meticulously carving wooden Union Jack flags. Mark calls out the deceptive ethics of manufacturing a persona to sell fake craftsmanship, though admitting it undoubtedly pulls in substantial revenue. - **The Viral AI Granny**: An eccentric, rebellious grandmother doing bizarre stunts who has amassed more than two million followers.
Analyzing these accounts reveals three classic high-ticket niches: health, wealth, and relationships. Curiously, none of the established AI influencers he found operated in the personal finance or wealth space, highlighting a clear gap in the market. To maximize his odds of success over the seven days, Mark decides to build three different characters spanning all three evergreen categories: one in health, one in wealth, and one in relationships.
[04:23](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=263s) Step 2: Generating Character Ideas with Claude
Committed to making the workflow as low-effort as possible, Mark turns to the free tier of Anthropic's Claude to brainstorm character profiles. He prompts the model to develop distinct personas for each niche, asking for a name, backstory, visual appearance, and unique selling proposition (USP).
For the health niche, Claude generates **Amos**: a 101-year-old traditionalist with a shaved head and a long grey beard who relies on ancestral herbal remedies passed down through generations. The persona effectively functions as a male counterpart to Mother Satori, giving him an immediate aura of ancient wisdom to promote wellness products.
For the wealth niche, Claude invents **Grandma Vivian**: a self-made, ostentatious matriarch who started from humble beginnings, outlived three husbands, and now flies in private jets clad head-to-toe in luxury designer fashion. Mark dubs her a "bougie boss grandma," noting that she represents the exact antithesis of the typical wealth influencer—young men posing in rented Lamborghinis. Her novelty provides a natural edge in a crowded industry.
For relationships, Claude devises **Sienna**: a stylish 22-year-old chronicling her modern dating struggles, red flags, and "situationship drama." Her profile mirrors the format used by the influencer Olivia, positioning her perfectly for conversational, podcast-style short-form videos.
[06:36](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=396s) Step 3: Generating Character Photos via Higgsfield and Claude
With character concepts locked in, Mark moves to image generation, selecting Higgsfield as his visual engine. He registers via email, setting his onboarding profile to personal avatar creation at an intermediate skill level. Rather than accepting the default onboarding offer, Mark applies a custom discount code (`TILBURY`) for 20% off. Because video volume will be critical for testing content, he subscribes to Higgsfield's Ultra tier—providing over 100 video credits—while keeping total software expenses safely below the $99 budget limit.
To streamline the workflow, Mark connects Higgsfield directly to Claude using a custom connector link in Claude's settings menu (`Customize` > `Connectors` > `Add Custom Connector`). This allows him to generate photorealistic images directly within the Claude conversation without tab-switching.
```mermaid flowchart LR accTitle: Claude and Higgsfield Integration Flow accDescr: Sequence showing how Claude connects directly to Higgsfield via custom connectors to output character imagery.
A["Claude Free Plan"] -->|"Custom Connector"| B["Higgsfield Engine (Ultra Plan)"] B --> C["Prompt Engineering in Chat"] C --> D["Amos Character Render"] ```
Mark asks Claude to compose detailed generation prompts for Amos, specifying: - Multiple camera perspectives: wide full-body shots and close-up portraits. - A clean, seamless white studio backdrop. - Authentic, weathered, un-airbrushed skin texture appropriate for an outdoorsman living off the land.
When the tool renders the first generation of Amos, the photographic quality is impressively sharp, but an immediate flaw surfaces: Amos looks far too young and vibrant for his concept. Mark remarks that the rendered man looks younger than himself rather than an authentic 101-year-old centenarian. Without skipping a beat, he prompts Claude to adjust the output and dramatically age Amos's visual features.
[10:32](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=632s) Generating Character Environments and Finalizing Visuals
After dialing in the character appearance for Amos—joking that the elderly farmer "still looks younger than me though, but I'll accept it"—the next objective is creating realistic, cohesive backdrops for each virtual persona. For Amos, the environments center entirely around a classic rural American aesthetic: a front porch, a rustic farmhouse kitchen, a lush vegetable garden, and the yard. The generated scenes turn out convincingly lifelike, capturing an authentic rural setting where Amos naturally fits in.
Moving swiftly to the other two influencers, Vivian's assets are rendered with high-society glamour: sunglasses, statement jewelry, and an opulent fur coat. The striking look immediately brings Cruella de Vil to mind, prompting the reaction: "Now that has got to be the coolest grandma you've ever seen. It looks like that woman from 101 Dalmatians, doesn't it?" Her scene backdrops match her lavish lifestyle, featuring: - A private jet cabin - A sleek sports car - An upscale classical drawing room - A manicured terrace suited for afternoon tea and scones
Finally, visuals and home environments are generated for Sienna. She comes across as effortlessly stylish and uncannily realistic to the point where an unsuspecting viewer wouldn't realize she was AI. Her generated backdrops mirror the everyday, aesthetic living spaces expected of a modern lifestyle creator.
---
[11:58](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=718s) Step 4: Mapping Out Viral Content with Claude Skills
With character models and environments locked down, the production moves to scripting viral content. Rather than drafting scripts by hand, the workflow relies on custom Claude skills—pre-built modules loaded with detailed system instructions designed to output formatted, high-converting scripts on demand without lengthy prompting.
```mermaid flowchart TD accTitle: Claude Skill Scripting Workflow accDescr: Process of importing a custom Claude skill, prompting the Content Engine, and outputting video scripts.
subgraph Setup["Tool Configuration"] A["Claude Customize Menu"] --> B["Upload & Install Skill"] end
subgraph Generation["Content Generation"] B --> C["Trigger /content engine skill"] C --> D["Request Ideas for Character"] D --> E["Select Top Concepts"] E --> F["Generate Full Scripts"] end ```
The process begins by navigating to the customize menu inside Claude, uploading the custom skill file, and activating the tool inside the chat window by typing `/` to select the content engine skill. Because Claude already retains the context of Sienna from earlier character ideation, it instantly produces a tailored list of video concepts:
- **Concept 2 (Vulnerability & Friendship):** A heartfelt script discussing toxic friendships where someone vanishes once they get into a relationship, creating deep emotional relatability among female viewers. - **Concept 1 (Actionable Advice):** Direct-to-camera lifestyle tips. - **Concept 7 (Psychological Secret / Podcast Clip):** A hook revealing a hidden psychological trick delivered in a popular podcast interview format.
Full scripts are generated for those three concepts, followed by a repeat workflow for Amos's longevity tips and Vivian's blunt wealth-building monologues.
---
[14:08](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=848s) Step 5: Giving the Characters a Voice in Higgsfield
Giving the AI influencers authentic speech requires pairing text with voice synthesis. Inside Higgsfield's audio dashboard, reference character portraits are uploaded along with a test script: *"Hi, my name is Vivian. Use this sound so the character speaks like this."*
To obtain an accurate vocal texture, Claude is prompted to write a descriptive voice profile matching Vivian's persona. The resulting description is pasted directly into Higgsfield's voice parameters. When generated, the voice comes out polished, refined, and authoritative, drawing the humorous remark that she sounds just like a strict former schoolteacher. The identical voice cloning and matching process is repeated for Amos’s folksy, aged cadence and Sienna’s natural, modern tone.
---
[15:22](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=922s) Step 6: Bringing the Influencers to Life in Cinema Studio
The project enters Higgsfield's Cinema Studio to merge the generated stills, environments, voice tracks, and scripts into full vertical videos. A project named "Sienna" is initialized with specific export settings: - **Aspect Ratio:** 9:16 (vertical format designed for mobile short-form feeds) - **Resolution:** 1080p for optimal clarity
The initial test begins with non-talking B-roll footage. A companion Claude video-prompt skill translates a core concept into a detailed cinematic prompt referencing the character model sheet (`Image 1`) to preserve visual consistency across takes. The resulting footage shows Sienna moving naturally, ideal for overlaying relatable text prompts.
Next, full speaking clips are generated across all three accounts, demonstrating impressive realism: - **Amos on Daily Habits:** A misty morning scene showing coffee steam rising as he recites, *"Early to bed and early to rise makes a man healthy, wealthy, and wise."* In another, he preaches the benefits of his top longevity food: *"Beans. It's the cornerstone of every long-lived culture out there."* - **Sienna's Relatable Hooks:** Flirty pickup lines (*"I personally can't turn water into wine, but I can turn you into mine"*) and sharp relationship call-outs about fair-weather friends (*"Babe, she was using you"*). - **Grandma Vivian's Direct Advice:** A street-interview style clip where she delivers ruthless property investment advice: *"Buy the ugliest property on the nicest street, fix it up properly, rent it or sell it. Take that money, do it again and again for 40 years. Don't touch it. Let it buy the next one."*
---
[19:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1147s) Step 7: Launching the Instagram Accounts
After adding dynamic on-screen subtitles using simple mobile editing software, the rollout shifts to publishing. To keep the test focused and manageable under time constraints, distribution is limited to Instagram, leaving TikTok and YouTube Shorts as potential expansions later on.
Amos's account setup serves as the primary template: - **Display Name:** Amos | Health and Wellness - **Handle:** `@amos.oldways` - **Profile Picture:** A warm, realistic portrait of Amos holding a young lamb - **Bio Copy:** *"101 years old. Simple habits for peak energy, focus and longevity. Your AI grandpa."*
The label "Your AI grandpa" is intentionally included in the profile bio to ensure full transparency and prevent followers from feeling deceived. Identical profile setup and bios are rolled out for Sienna and Vivian.
The strategy is straightforward: post an initial batch of 10 videos on all three accounts, track performance over several days, crown whichever persona generates the highest engagement as the winner, and direct monetization efforts into that winning profile.
---
[21:37](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1297s) Unplanned Check-in: Early Audience Reactions
Before the scheduled check-in to evaluate metrics and declare a winner, unexpected user engagement forces an impromptu update. Monitoring the incoming notifications reveals that viewers are already reacting passionately to the content. The comment sections are filling up with intense, genuine reactions—including critical comments from users actively arguing with and bashing Sienna's relationship takes as if she were a real, living person.
[21:51](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1311s) Comment Drama and Crowning the Winning Account
Reviewing the initial engagement on Sienna's Instagram reveals how convincing the AI persona has become. One hater mocks Sienna in the comments, claiming she is "wearing a Shein dress and drinking a $3 USD Frappuccino." Another woman quickly jumps to Sienna's defense, arguing that she has every right to choose whatever standard she wants in a partner because she is in such high demand. Tilbury finds the back-and-forth hilarious: both commenters are so absorbed in arguing that neither realizes Sienna is not an actual human being. Other users leave comments inquiring where she bought her clothes, further proving the hyper-realistic quality of the AI generation.
After intentionally stepping away from the accounts to allow the algorithms time to work, Tilbury reviews the metrics across all three characters to determine which one will move forward into the monetization phase. The standings are revealed:
```mermaid flowchart TD accTitle: Account Performance Ranking accDescr: Ranking of the three AI influencer accounts tested during the challenge.
A["Account Performance Test"] --> B["3rd Place: Vivian (Grandma Viv)"] A --> C["2nd Place: Sienna"] A --> D["1st Place: Amos (Winner)"]
B --- B1["Underperformed early on; stung because Mark rooted for her"] C --- C1["Strong numbers for a brand new account"] D --- D1["Massive lead; crowned winner to monetize"] ```
Vivian comes in third place, which stings because Tilbury openly rooted for her from the start. However, this discrepancy validates the initial strategy of launching three accounts simultaneously rather than pinning all hopes on a single persona. Sienna secures second place with genuinely impressive numbers for a brand-new page. Amos takes first place with an undeniable margin. Because a seven-day timeline leaves no room to build out multiple revenue funnels, Amos is officially crowned the winner.
[23:30](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1410s) Choosing the Monetization Model
With Amos chosen, Tilbury evaluates the potential revenue models commonly employed by AI influencers to find the most practical method:
* **Brand Deals:** Unfeasible for a short timeframe, as pitching companies, waiting for email replies, and negotiating contracts takes weeks. * **Physical Products:** Creating a dedicated Shopify store and managing fulfillment is an extensive project of its own (though Tilbury offers to tackle it in a future video if viewers request it). * **Ad Revenue:** Instagram Reels does not currently offer direct ad revenue payouts for short-form video views. * **Affiliate Income:** TikTok Shop is recognized as one of the fastest ways to monetize an AI creator by taking a cut of product sales, but Tilbury had not launched TikTok accounts for the challenge—a decision acknowledged in hindsight as an oversight. * **Digital Products:** The winning candidate. Educational and lifestyle figures like Amos match digital products seamlessly, and the entire offering can be conceptualized and generated directly inside Claude.
Recognizing that generic ebooks are largely obsolete because audiences can easily prompt ChatGPT or search Google for basic facts, Tilbury emphasizes that a viable digital product must be structured, interactive, and guide the customer through an actionable transformation. Amos's focus on simple, grounded living makes an interactive habit-tracking workbook ideal.
[25:31](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1531s) Building Amos's Habit Workbook in Claude
Tilbury accesses Claude, pulls up the dedicated digital product prompt module, and enters a request to design a workbook for Amos. Claude returns three separate concept outlines with market rationale explaining why a consumer would purchase each one. Tilbury selects the first option: a 30-day challenge grounded in Amos's traditional lifestyle, introducing one specific daily habit at a time.
Instructing Claude to build out the full product produces a comprehensive digital guide: - A structured 30-day program featuring dedicated daily rules with checkboxes. - A printable habit tracker placed at the back, suitable for sticking on a refrigerator. - Bonus supplementary material, including his mother's traditional dinner recipes.
Tilbury adds a generated portrait of Amos to the front cover, remarking on the surreal reality that an individual who did not exist five days prior is now the featured author of his own published book. Before taking it to market, Tilbury thoroughly inspects the text to verify that the advice is safe, logically coherent, and stripped of unedited "AI slop."
```mermaid flowchart LR accTitle: Digital Product Creation and Setup accDescr: The workflow used to generate, polish, and list Amos's digital product.
subgraph Claude ["Claude Generation"] A["Digital Product Skill Prompt"] --> B["Select 30-Day Challenge Concept"] B --> C["Generate Workbook & Recipe Bonuses"] C --> D["Manual Review & Edit"] end
subgraph StanStore ["Stan Store Setup"] E["Upload Cover & Product PDF"] --> F["Set Pricing: $15.99 to $12 USD"] F --> G["Paste Store URL in Instagram Bio"] end
Claude --> StanStore ```
[27:15](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1635s) Storefront Setup on Stan Store
To sell and deliver the digital download, Tilbury utilizes Stan Store, taking advantage of their 14-day free trial to keep startup expenses at zero.
Setting up the storefront involves: - Navigating to the product creation dashboard and selecting a "callout" layout design. - Uploading a profile thumbnail of Amos with a sheep to maintain consistency with his Instagram avatar. - Setting the display title to "Get my 30-day energy reset" and labeling the call-to-action button "Start the old ways challenge." - Pricing the workbook at an original price of $15.99 USD, discounted to a promotional price of $12 USD. - Uploading the PDF file, grabbing the hosted landing page URL, and pasting it directly into the Instagram profile bio. - Privately linking a payout bank account off-camera before verifying that the live landing page functions properly.
[29:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1747s) A Twist in the Algorithm and Final Profit Breakdown
On the final day of the challenge, the accounts experience an unexpected algorithmic shift. Despite being abandoned in third place, Vivian's account suddenly goes viral: one video crosses 10,000 views, another reaches 20,000, and a third hits over 250,000 views. Tilbury laments not trusting his instincts on "Grandma Viv," noting that while seven-day experiments are great for testing concepts quickly, organic social media algorithms often take longer to reveal which persona will truly dominate.
Reflecting on the broader test, Tilbury points out that despite transparently disclosing that the influencers were synthetic personas, not a single viewer left a negative comment complaining about AI. Even so, Tilbury shares personal reservations: while AI influencers are clearly profitable and straightforward to launch, he cautions against an internet overrun by automated personas, observing that "it sort of kind of sucks the soul out" of social media, where human connection across diverse walks of life is the true appeal.
The financial results of the experiment demonstrate clear profitability for the $12 USD workbook:
| Metric / Expense Item | Amount (USD) | Notes | |:--- |:--- |:--- | | **Total Revenue** | **$276.00** | 23 workbook sales at $12.00 each | | Claude Subscription | $0.00 | Free plan utilized (waiting on credit refreshes) | | Instagram Edits | $0.00 | Native tools used for video captioning | | Stan Store Hosting | $0.00 | Utilized 14-day free trial | | PayPal Processing Fees | $19.52 | Deducted directly from gross sales revenue | | Higgsfield Subscription | $99.00 | Accounted at standard retail cost without affiliate discounts | | **Total Costs** | **$118.52** | Startup tooling budget plus payment fees | | **Net Profit** | **$157.48** | Pocketed profit over the 7-day challenge |
Tilbury highlights that while $157.48 USD in profit is not life-altering on its own, generating pure profit from an initial $99 USD budget with minimal manual labor proves the viability of the AI influencer business model—especially when considering that the content could easily be repurposed and syndicated across TikTok and YouTube Shorts.
[32:23](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1943s) Scaling Potential and the $1,000 Challenge Giveaway
Reflecting on the broader potential of the business model, Mark points out that this experiment only scratched the surface. A creator could easily cross-post and syndicate the exact same video assets across competing short-form platforms like TikTok and YouTube Shorts to instantly multiply views, audience attention, and revenue "without really lifting a finger." Furthermore, of all the diverse monetization avenues mapped out at the start of the experiment, only a single method was tested on a single character. Rolling out every monetization pathway across all three virtual personalities would yield significantly higher returns.
Even with that untapped upside, Mark admits that "a deal is a deal." Having imposed a strict seven-day deadline that he failed to satisfy, he follows through on his forfeit pledge:
> "And as I clearly failed the challenge, I'm going to be giving one of you in the comments $1,000 USD of my own money."
To enter the giveaway, viewers simply have to leave a comment detailing who their favorite AI influencer was and why—with Mark joking that he promises not to be biased toward Grandma Vivian. Wrapping up the tally of his profits, he directs viewers to his next video attempting a seven-day AI dropshipping challenge, urging everyone to subscribe if they are serious about building wealth.
- **Untapped Distribution Channels**: Repurposing short-form video assets onto TikTok and YouTube Shorts for zero-effort audience multiplication. - **Giveaway Terms**: $1,000 USD awarded to a commenter sharing their favorite AI influencer and reason. - **Next Challenge**: The 7-day AI dropshipping challenge.
## Notable quotes > [01:14](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=74s) I thought I was the old one around here — Mark Tilbury
> [10:32](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=632s) still looks younger than me though, but I'll accept it — Mark Tilbury
> [10:32](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=632s) Now that has got to be the coolest grandma you've ever seen. It looks like that woman from 101 Dalmatians, doesn't it? — Mark Tilbury
> [14:08](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=848s) Hi, my name is Vivian. Use this sound so the character speaks like this.
> [15:22](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=922s) Early to bed and early to rise makes a man healthy, wealthy, and wise. — Amos
> [15:22](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=922s) Beans. It's the cornerstone of every long-lived culture out there. — Amos
> [15:22](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=922s) I personally can't turn water into wine, but I can turn you into mine — Sienna
> [15:22](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=922s) Babe, she was using you — Sienna
> [15:22](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=922s) Buy the ugliest property on the nicest street, fix it up properly, rent it or sell it. Take that money, do it again and again for 40 years. Don't touch it. Let it buy the next one. — Grandma Vivian
> [19:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1147s) 101 years old. Simple habits for peak energy, focus and longevity. Your AI grandpa.
> [29:07](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1747s) it sort of kind of sucks the soul out — Mark Tilbury
> [32:23](https://www.youtube.com/watch?v=LlhTEttKcwQ&t=1943s) And as I clearly failed the challenge, I'm going to be giving one of you in the comments $1,000 USD of my own money. — Mark Tilbury
## Mentioned Mark Tilbury, Claude, Anthropic, Higgsfield, Lil Miquela, Mother Satori, Amos, Grandma Vivian, Sienna, Olivia, Instagram, TikTok, TikTok Shop, YouTube Shorts, Stan Store, Shopify, PayPal, ChatGPT, Google, Prada, Calvin Klein, Shein
## Video description Get an EXTRA 20% OFF Higgsfield with code TILBURY: https://higgsfield.ai/s/marktilbury-lwzqFm
*This is an affilliate link. Offer only available for the first 5,000 people who use the link & redeem the code.
Get the FREE cheat sheet here: https://www.dropbox.com/scl/fo/r4v0e3a75sygqhh0dl1yj/ADa4_5eA14PYBaSkuCPHr0A?rlkey=rq5asdmab9vhurd7zeyl2gari&st=bd5uq4fh&dl=0 ________________________________________________ TIME STAMPS: 00:00 Intro 01:14 Step 1. Spy On The Competition 04:23 Step 2. Generate Character Ideas 06:36 Step 3. Generate The Photos 11:57 Step 4. Map Out Viral Content 14:08 Step 5. Give Them A Voice 15:22 Step 6. Bring Them To Life 19:07 Step 7. Launch Our Accounts 21:26 Step 8: Crown The Winning Account 21:28 Unplanned Check-in 22:37 Step 9. Start Making Money 29:05 Step 10. Count The Profits ________________________________________________ In this video I tried one of the biggest AI side hustles right now... creating AI influencers from scratch and seeing if they could actually make money.
Using AI tools like Claude and Higgsfield, I built three virtual influencers, created AI photos and videos, launched Instagram accounts, tested viral content and tried to turn the attention into real income... all without using my real face or voice.
In this 7-day AI influencer challenge, I show exactly how to create an AI influencer, make realistic AI videos, grow an AI Instagram account and monetise it with digital products.
If you’re interested in AI side hustles, making money with AI, virtual influencers, AI content creation, Claude, Higgsfield, Instagram Reels, YouTube Shorts or passive income ideas, this is a real-world test of whether the AI influencer business model actually works. ________________________________________________ GET IN TOUCH: For business inquires only, please use this email: mark@marktilbury.com
[Source](https://www.youtube.com/watch?v=LlhTEttKcwQ)