Three initiatives. Three different responsibilities. One philosophy.

I’ve spent the better part of 20 years building WatchMojo, a media company that has had to adapt to virtually every technological disruption imaginable.

Search. Social media. YouTube. Mobile. Streaming. Algorithms. Programmatic advertising. Platform policy changes.

And now, artificial intelligence.

Over the years, I’ve been called everything from a visionary to a dinosaur, sometimes by the same people, depending on whether I embraced a particular technological shift quickly enough—or questioned its implications.

My position on AI is neither blindly enthusiastic nor reflexively hostile.

AI should democratize creation, not become an excuse to exploit creators. It should eliminate unnecessary barriers to producing work, not eliminate the humans who give that work its value.

But here are two major nuances that seems to get lost in the increasingly polarized AI debate:

First: The main factor that shapes how people judge AI: who is using it, who benefits, and what the public perceives as the greater good. At WatchMojo, I may still see myself as the scrappy outsider or nobody who started with $250,000 and went $500,000 into debt to avoid layoffs, but after 20 years, we’re an incumbent—with employees, an established brand, and responsibilities that come with success. I have explained why incumbents can benefit most from AI, but have less leeway.

At SoundMojo, we were starting something entirely new, using AI to experiment with music. But because the immediate public benefit wasn’t obvious—it was initially a passion project, even if audiences might eventually enjoy the music—we became an easy target for internet drama and commentary channels looking for their next controversy.

Contrast that with the Peanut Project, where AI is helping one individual pursue something that could bring joy to millions of Expos fans. People are understandably more inclined to give that use of AI the benefit of the doubt. Yet the moment we floated the possibility of merchandise—not even primarily to make money, but to generate momentum—questions about intellectual property naturally emerged. And rightly so. The lesson? Ethics aren’t determined by popularity or good intentions, but perceived benefit, existing obligations, and the greater good absolutely influence how society evaluates the same behavior in different contexts.

Second: The ethical implications of using AI depend enormously on the context.

Consider three initiatives I’m involved in:

1. WatchMojo: An established media business with a brand, employees, contractors, customers, and two decades of institutional history.
2. SoundMojo: A creative initiative exploring how AI can help develop musical ideas and demos before committing resources to commercial productions.
3. The Peanut Project: My personal initiative to help restore Major League Baseball to Montreal, currently pursued without a formal franchise ownership vehicle, a salary, or a dedicated workforce.

I use AI across all three.

But I don’t believe the same ethical standard applies in precisely the same way to each.

To understand why, we need to distinguish three questions:

What is legally permissible? What is ethically defensible? And what is strategically wise?

Those are not always the same thing.

---

1. Three ventures, three different responsibilities

	WatchMojo	SoundMojo	Peanut Project
Nature	Established media business	Emerging creative venture	Personal sports restoration initiative
Existing workforce	Yes	Not necessarily	No dedicated workforce
Primary AI opportunity	Productivity and creative augmentation	Lower-cost experimentation	Research, planning and execution
Main ethical concern	Displacing or devaluing existing staff	Replacing or imitating musicians without consent	Misrepresentation and manufactured credibility
Human contribution	Editorial judgment, creativity, accountability	Composition, performance, artistry	Leadership, relationships, judgment
Core principle	Augment before replacing	Prototype before producing	Research before mobilizing

The biggest distinction is that WatchMojo has existing obligations to people that SoundMojo and the Peanut Project do not have in the same way.

That doesn’t mean AI should be restricted to protect every existing job or process.

It means the transition deserves more care.

An entrepreneur developing something from scratch has different responsibilities from an employer introducing automation into an organization whose people helped build the business.

Neither situation is inherently more virtuous.

They’re simply different.

And context matters: I have never laid off staff the way others have intermittent blood letting exercises. It even takes a lot for me to approve a dismissal when an executive suggests it’s time. I am idealistic, and fundamentally believe a corporation has an obligation to create and provide employment within a given community (within reason).

2. WatchMojo: AI as a force multiplier, not an invisible replacement workforce

WatchMojo has spent two decades building an editorial operation, a brand, a library, a distribution network, and a global audience.

We started with $250,000 in seed capital and outlasted competitors that raised hundreds of millions.

Part of that survival came from adapting to technology rather than fighting it.

But the people who contributed to WatchMojo’s success have legitimate expectations about how new technology will be introduced.

Good and ethically defensible uses of AI

• Research assistance, transcription, translation, summarization, and fact-checking support.
• Brainstorming titles, formats, outlines, thumbnails, and distribution strategies.
• Identifying patterns across our enormous back catalog to inform programming decisions.
• Automating repetitive tasks such as metadata preparation, captions, and content organization.
• Helping writers and editors explore more ambitious concepts within existing production schedules.
• Enabling smaller teams to produce new formats that otherwise wouldn’t be economically viable.

These are all examples of AI expanding human capability.

A researcher who previously spent three hours compiling preliminary information might now spend 30 minutes gathering it and devote the remaining time to verifying facts, developing arguments, and improving the final product.

That’s potentially a better use of human intelligence.

But there’s an important distinction between AI generating a draft that an editor meaningfully reviews and AI generating an entire video that nobody takes responsibility for.

The editor must remain accountable for the finished product.

Bad or ethically questionable uses

• Quietly replacing writers, researchers, or editors while presenting the output as their work.
• Imitating the voices or likenesses of WatchMojo talent without explicit permission.
• Publishing AI-generated factual claims without adequate human verification.
• Using freelancers’ work to train proprietary systems in ways that violate contractual rights or reasonable expectations.
• Using productivity gains exclusively to reduce headcount while demanding ever-greater output from the remaining staff.

That last point is complicated.

I’m an entrepreneur. I’m not going to pretend technology never eliminates jobs.

History suggests otherwise.

Sometimes automation genuinely changes the economics of a business. Sometimes positions become redundant. Sometimes a company must restructure to survive.

But there is a difference between managing technological change transparently and using AI to quietly discard people who helped build your business.

Where practical, I favor retraining, redeployment, and natural attrition.

I don’t believe a business can promise to preserve every job forever.

I do believe it owes its people honesty, fairness, and a serious effort to help them adapt.

For WatchMojo, my principle is straightforward:

Augment before replacing.

3. SoundMojo: AI as the demo studio, humans as the artists

Now consider SoundMojo.

Historically, someone with a musical idea might need a composer, musicians, a producer, studio time, recording equipment, and significant financial resources just to determine whether the concept had potential.

What if the song wasn’t very good?

What if the melody worked but the arrangement didn’t?

What if the idea sounded promising in your head but failed in execution?

You could spend considerable time and money discovering that.

AI changes the economics of experimentation.

Today, someone can explore dozens of musical directions, test arrangements, experiment with instrumentation, and generate demos at a fraction of the historical cost.

That is a remarkable democratization of creative development.

And I think it can be a very good thing.

The SoundMojo model

Stage	Approach
1. Concept	Human idea, lyrics, theme, and creative direction
2. Experimentation	AI-generated demos, arrangements, and variations
3. Selection	Human judgment determines which ideas deserve further investment
4. Production	Musicians, singers, and producers collaborate on commercial versions where appropriate
5. Distribution	Commercial release with clear rights, credits, and economics

The idea isn’t necessarily to eliminate musicians.

It’s to lower the cost of discovering which musical ideas deserve investment in musicians.

Think about the traditional Hollywood process.

A writer develops a script. A studio commissions concept art. A filmmaker creates a storyboard. An animator produces a rough sequence.

Nobody confuses the storyboard with the finished film.

In my mind, AI-generated musical demos can serve a similar purpose.

They allow us to experiment before committing substantial resources.

This is why WatchMojo Studios, AceMojo Productions, was able to move a handful to a dozen of my creative ideas from development to production, where humans will be employed and whose creativity can be unleashed (at a time when Hollywood is shrinking).

Where does it become unethical?

There are obvious concerns.

Using AI to imitate a living singer’s recognizable voice without permission? Problematic.

Passing off synthetic performances as recordings by real musicians? Deceptive.

Appropriating another artist’s work in violation of applicable rights? Potentially unlawful and unethical.

Failing to compensate collaborators whose contributions materially shape the commercial recording? Unfair.

But here’s another nuance.

I don’t believe every commercial musical release must involve human musicians to be ethical.

An openly AI-generated instrumental, produced using appropriately licensed tools and inputs, could be perfectly defensible.

Conversely, hiring a singer to reproduce an AI-generated demo doesn’t automatically make the underlying process ethical or legally safe.

The critical questions are consent, rights, attribution, compensation, and honesty with the audience.

My preferred SoundMojo philosophy is:

Prototype before producing.

AI allows us to discover which ideas are worth pursuing. Humans decide what those ideas should become.

4. The Peanut Project: AI as a one-person research department

Now let’s turn to something completely different.

For nearly a year, I’ve been pursuing an improbable objective: helping restore the Montreal Expos.

I’ve called it the Peanut Project.

It’s a massive undertaking involving professional sports economics, private equity, franchise ownership structures, stadium development, municipal planning, infrastructure, media rights, commercial partnerships, and community engagement.

Traditionally, pursuing something of this magnitude would require an army of investment bankers, lawyers, architects, consultants, analysts, public relations professionals, and researchers.

And enormous financial resources.

But here’s the thing.

I’m not currently operating an MLB franchise. I’m not being paid to run an ownership group. I’m not even necessarily going to be part of the eventual ownership structure.

I’m an individual pursuing an initiative I believe could benefit Montreal, baseball fans, and the broader community.

I have no dedicated Peanut Project workforce to displace.

And thanks to AI, I can investigate questions and develop preliminary work that might previously have required a small organization.

What AI allows me to do

• Research MLB expansion economics and ownership structures.
• Model franchise fees, stadium budgets, and potential financial returns.
• Compare potential stadium sites and development scenarios.
• Generate conceptual stadium renderings.
• Draft investor outreach, articles, letters, and presentations.
• Explore transportation, zoning, and infrastructure considerations.
• Compare potential media-rights and commercial models.
• Organize survey findings and evaluate public feedback.
• Challenge my own assumptions before approaching professionals and decision-makers.

This is an extraordinary force multiplier. False modesty aside, I have done the job of entire accounting, consulting, civic planning and public relations firms alone.

And unlike WatchMojo, where introducing AI may affect existing workflows and jobs, the Peanut Project is largely an example of AI making work possible that otherwise might never have happened at all.

Without AI, I probably wouldn’t have pursued the initiative with the same breadth or speed.

Not because I lack the interest or willingness to work.

But because I would have lacked the time and resources to investigate so many dimensions simultaneously.

But there is an ethical line

Use	My assessment
Research MLB expansion economics	Good, provided sources are verified
Model franchise fees and stadium financing	Good, with assumptions disclosed
Generate hypothetical stadium renderings	Good, clearly labeled as conceptual
Draft letters, articles, and investor outreach	Good, provided I take responsibility
Compare potential development sites	Good for preliminary analysis, not professional certification
Invent investor commitments that don’t exist	Unethical
Present an AI rendering as an approved municipal plan	Unethical
Fabricate endorsements from MLB, celebrities, or government officials	Unethical
Generate fake fan testimonials or survey responses	Unethical

The ethical concern here isn’t job displacement.

It’s the risk that persuasive presentation gets ahead of underlying reality.

A beautiful stadium rendering isn’t a zoning approval.

A financial model isn’t committed financing.

An AI-generated investment thesis isn’t investor due diligence.

And a compelling article isn’t evidence that MLB agrees with my conclusions.

Those distinctions matter enormously.

But they don’t make the tools themselves inappropriate.

If anything, AI allows someone without institutional backing to develop a more informed case before asking others to commit time, money, or credibility.

My Peanut Project principle is:

Research before mobilizing.

5. Five ethical principles that apply everywhere

Although the circumstances differ, I believe five principles should govern AI use across all three initiatives.

Principle 1: Accountability cannot be outsourced

AI can write, calculate, illustrate, summarize, and suggest.

But whoever publishes or acts on the result must own the consequences.

If WatchMojo publishes an inaccurate video, we cannot blame the chatbot.

If SoundMojo releases a recording that violates someone’s rights, we cannot blame the music generator.

If I publish a flawed financial analysis about the Expos, I cannot blame the model.

AI is a tool. Accountability remains human.

Principle 2: Consent matters more than technical possibility

Just because AI can clone a voice doesn’t mean you should.

Just because it can recreate someone’s likeness doesn’t mean you have permission.

Just because it can mimic an artist’s recognizable style doesn’t mean every resulting use is appropriate.

Technology expands what is possible.

Ethics helps determine what is acceptable.

Principle 3: Transparency should be proportionate to the risk of deception

I don’t believe every email needs a disclaimer saying AI helped improve the grammar.

Nor does every preliminary financial model need a warning that software assisted with the calculations.

But if a realistic video depicts an event that never happened, viewers should know.

If a stadium rendering is purely conceptual, it should be labeled accordingly.

If a synthetic singer could reasonably be mistaken for a particular real artist, that raises a much more serious disclosure and consent issue.

The greater the risk of misleading people, the greater the obligation to be transparent.

Principle 4: Productivity gains should create opportunities, not merely extract savings

At WatchMojo, AI efficiencies can help fund better programming, new intellectual property, employee development, and more ambitious productions.

At SoundMojo, inexpensive experimentation can free up resources to hire musicians and performers for promising projects.

At the Peanut Project, AI can allow a grassroots initiative to advance without a multimillion-dollar administrative structure.

The ideal outcome isn’t simply doing the same thing with fewer people.

It’s accomplishing things that previously weren’t economically possible.

Principle 5: Human relationships remain the scarce resource

AI can draft an investor pitch.

It cannot genuinely commit $300 million.

AI can generate a stadium design.

It cannot obtain consent from residents.

AI can create a musical demo.

It cannot automatically reproduce the trust, judgment, lived experience, and chemistry that make a great creative collaboration work.

And AI can help me research the Expos.

It cannot build the relationships necessary to restore a franchise.

The more abundant information and content become, the more valuable trust, character, judgment, and relationships may become.

6. The paradox: You can embrace AI and still protect human creativity

Some people seem to believe there are only two positions.

Either you’re pro-AI and therefore indifferent to human creators.

Or you’re pro-human and therefore opposed to AI.

I reject that framing.

You can enthusiastically embrace AI and still believe in fair compensation, creative ownership, meaningful employment, and the dignity of human work.

But I would also challenge the argument that because WatchMojo has employees, AI should be heavily constrained there, whereas because the Peanut Project has no employees, anything goes.

That’s not my position.

The ethical differences are real, but the underlying principles remain consistent.

Using AI to create an original stadium illustration? Defensible.

Using AI to imitate an architect’s work and falsely claim the architect supports the project? Not defensible.

Using AI to brainstorm a SoundMojo song? Defensible.

Using it to impersonate a singer without permission? Not defensible.

Using AI to help WatchMojo researchers produce more thoughtful scripts? Defensible.

Using it to fabricate sources? Not defensible.

The ethical test isn’t simply whether AI replaces labor. It’s whether AI creates value responsibly, respects rights, preserves accountability, and avoids deception.

And even that requires nuance.

Because technology has always displaced some forms of labor while creating others.

The printing press. The camera. The automobile. The computer. The internet.

AI is not the first technology to challenge existing economic arrangements.

It may, however, be among the fastest and broadest.

Which makes how we manage the transition particularly important.

7. Steward, entrepreneur, individual: three different relationships with AI

When I look at my own use of AI, I see three distinct roles.

At WatchMojo, I’m a steward.

I have responsibilities to employees, contractors, shareholders, partners, audiences, and a brand built over two decades.

AI should strengthen that institution while treating the people behind it fairly.

At SoundMojo, I’m an entrepreneur.

I’m exploring how new technology can reduce the cost of creative experimentation, make more ideas economically viable, and establish new partnerships between human creativity and machines.

At the Peanut Project, I’m an individual.

I’m using technology to investigate an ambitious undertaking that would otherwise require resources far beyond those of one person.

These aren’t contradictions.

They’re three applications of the same underlying philosophy:

Use technology to expand what people can accomplish.

The real question is what we do with that leverage.

Do we create more value, more opportunities, and better work?

Or do we simply appropriate value created by others?

There’s another distinction worth making: authorship versus execution.

A person can originate an idea, exercise meaningful creative judgment, direct AI-assisted development, and remain the driving force behind a project.

But that doesn’t automatically mean every AI-generated element is copyrightable, nor does it erase the contributions of collaborators.

As AceMojo Productions and SoundMojo move toward commercial releases, those questions will become increasingly important.

The goal should be to establish clear ownership, rights, compensation, and creative responsibilities from the outset.

8. The biggest opportunity: Giving individuals the capabilities of institutions

I keep coming back to something that I think is getting lost in the debate.

Most discussions about AI focus on how established businesses can reduce costs.

How many jobs can be automated?

How many hours can be saved?

How much productivity can be extracted?

Those questions matter.

But I wonder whether the greatest opportunity lies somewhere else entirely.

What happens when people who lack capital, connections, or institutional support suddenly gain access to capabilities that were once reserved for large organizations?

A young filmmaker can visualize a concept before securing financing.

A songwriter can produce a demo without renting a studio.

An entrepreneur can evaluate an opportunity without immediately hiring consultants.

And one determined Montrealer can spend a year researching how to restore a Major League Baseball franchise without first assembling an investment bank.

Of course, AI doesn’t guarantee any of those ideas will succeed.

It doesn’t eliminate the need for capital, expertise, relationships, or hard work.

But it dramatically lowers the cost of trying.

And that matters.

Because historically, many ideas never failed in the marketplace.

They never reached the marketplace.

They died because their creators couldn’t afford to test them.

AI has the potential to change that.

Conclusion: Reduce the cost of trying, not the value of being human

If I had to summarize my philosophy in three lines, it would be:

At WatchMojo: Augment before replacing.

At SoundMojo: Prototype before producing.

At the Peanut Project: Research before mobilizing.

And across all three:

Empower before employing.

That last line isn’t an argument against employment.

It’s an argument for giving individuals the tools to accomplish more before they need to assemble expensive organizations.

Sometimes those experiments will fail.

Sometimes they’ll become businesses.

Sometimes they’ll create jobs.

And occasionally, perhaps, they’ll help bring back a baseball team.

Twenty years ago, I built WatchMojo around the idea that technology could democratize media production and distribution.

Today, AI is taking that democratization to another level.

But the underlying principle hasn’t changed.

Technology should expand human potential, not diminish human worth.

Or, more simply:

AI should reduce the cost of trying, not the value of being human.