You do NOT need a product for AI to change your business

Hey Innovator,

We've talked about the two divergent paths for AI enthusiasts – one where you create a consumer-facing product, and another where you create an internal AI process that supercharges your existing, non-AI business.

You can do both – especially once you learn the art of rapid prototyping. But what was most exciting to me was all the feedback we got from readers who’d never really considered building an AI process, and were newly excited and inspired when they realized there were ways to take advantage of AI without jumping into the fray of the 1,000s of consumer-facing AI tools.

The stories that really get to me, though, are the ones where readers tell me that they feel like they can’t do anything with AI because they don’t have the time, energy or desire to create a consumer product.

Let’s make this clear:

You do NOT need to create a product to succeed in AI.

If you get caught in the “I need to build a product” trap, you end up wasting your potential because there is SO MUCH you can do with an AI process.

Today I’ll dive deeper into why creating an AI process seems more-difficult and less-obvious, and why people naturally (and often incorrectly) gravitate toward products first.

If you’re feeling iffy about creating a product – but still want to harness AI to grow your business or career – today’s email is for you.

AI Processes require your unique subject-matter expertise

Earlier this week I shared my experience taking a bunch of $10 courses with titles like “23 ways to make money with AI.” (Yes, I really bought them and spent a long weekend sifting through each one.)

The common thread through all the “ways to make money” they present is that anyone can do them.

And that’s exactly the problem – if everyone can do it, it is inherently not valuable. That’s where all the lists of “1,001 ChatGPT prompts” and “327 AI business ideas” fail. They are so generic that they are effectively worthless.

Instead of generic ideas, you need to incorporate your unique subject-matter expertise into everything you do with AI.

After all, you’ve spent your entire career – many decades in some cases – building expertise in your specific domain. That domain wasn’t directly related to AI, since ChatGPT just launched in 2022. But now that AI exists, you have a unique opportunity to combine a lifetime of expertise (which is really hard for anyone else to replicate) with all the new possibilities that come from easy, no-code access to AI.

Think about some of the examples IWAI students shared in our Demo Day (free replay here).

1 – Erich has a lifetime of expertise in television production and won an Emmy in 2016.

So when he created an AI process, it wasn’t something generic like “use ChatGPT to write a Google ad.” Instead, he created a 10-step process that allows him to take a cryptic request from a TV network (they call these “mandates” and they’re always vague) and turn it into dozens of new TV show ideas, complete with imagery and scripts for promotional sizzle reels, within minutes.

There are probably less than 10,000 people on the planet who have the subject-matter expertise to do what Erich did, and probably less than 10 who have the subject-matter expertise AND the AI expertise.

That’s what’s so exciting about building AI processes right now. Because you are an early adopter of AI, you are already in the 99th percentile of all humans in terms of AI knowledge. And when you combine that with your established expertise in your non-AI field, you are automatically one of the elite few who can create an AI process to supercharge the specific type of work that you do.

That’s a big, potentially life-changing claim. So let’s look at the data that backs it up.

First, you need career experience and subject-matter expertise. If you start college at 18, it will take four to six years to get a bachelor’s degree and start your first full-time job or business venture. That puts you around age 23.

Of course, these days new grads average four different jobs in the first 10 years of their career, so it’s hard for them to settle in and achieve expertise in a field before 30. Even as someone who got an early start (I was coding at 12 and running small side hustles in my teens), this applies to me – there was a very noticeable shift around age 30 where I felt like (and was viewed by others as) a seasoned pro in my field.

So, we know it’s very rare for someone to be a subject-matter expert before they turn 30. There’s just not enough time to accrue 10,000 hours of deliberate practice and hands-on experience before then.

But then you hear statistics like “100 million people use ChatGPT every week.” Someone is using AI – but guess who?

It’s mostly people age 18-29 – the exact cohort that hasn’t had time to develop subject-matter expertise in their profession. In fact, most of them haven’t even decided on their profession.

In December, YouGov asked 13,000 Americans about their use of AI. In the 18-29 age bracket, 50% said they used AI at least weekly. This dropped to 42% for those aged 30-44 and just 18% for those aged 45-64.

In other words, the more experienced a person is in their career, the less likely they are to have a working knowledge of AI.

That’s why Erich is crushing it – he spent 20 years building his real-world industry expertise.

Then, he caught up to the tech-savvy teenagers in terms of AI skills in a matter of weeks.

Compared to 18-29 year olds, he’s in the upper echelons of career success. Compared to 30-64 year olds, he’s in the upper echelons of AI skills. He’s right in the sweet spot for AI success, and there’s a good chance that you are too.

2 – Matt has a lifetime of experience in health and wellness.

He runs a wellness consulting company already, and his clients give him a wealth of data via detailed questionnaires and real-life sessions. And, like Erich, he created a process to turn what used to be a slog into something that can be done in a matter of minutes with AI.

The first step in working with Matt is filling out an intake questionnaire, which goes into a lot of depth about your current health status, challenges and goals. In the past, Matt would spend ten hours analyzing that questionnaire and writing a bespoke report for each client. Now, his AI process (which you can see in the Demo Day replay) does 90% of it for him in a matter of minutes.

That takes his intake time per client down from ~10 hours to ~20 minutes – a 30x efficiency improvement in the most important leading performance indicator of his business.

But it’s important to note that a non-expert like you or me could not build a wellness-report-generating AI process right now. Because you and I do not have the subject matter expertise to know how the AI should analyze the client’s answers, our version would just turn out to be generic and elementary. Matt’s expertise makes all the difference.

Likewise, you and I don’t have decades of expertise in TV production, so we couldn’t just say, “Hey ChatGPT, write me a TV show idea” and get anything of value back. Erich’s expertise is the secret sauce.

A lot of people like to ask the question “what’s your moat,” which comes from the tech product world. (The moat is the metaphor for the reason it’s hard for competitors to steal your idea, just as a moat protects a medieval castle.) I often think this question is kind of silly, but in this case, there is a good answer.

Erich and Matt have a moat because very few other people can match their subject-matter expertise. And among those who can, almost none of them can match their AI expertise.

They also have a moat because building an AI process is inherently a private endeavor. There is no big launch day or press release.

Nobody can sign up for a subscription and reverse-engineer your whole system like they might try to do with a product.

You control access, and thus it’s nearly impossible for a sketchy competitor to copy you.

That’s the magic combination that makes your AI process so powerful, and why you can dramatically increase your income without ever selling an AI product to an outside customer.

AI Processes are multi-step and often require a human touch

AI processes also confuse people because they don’t fit the mold of “complete, instantaneous automation.”

For example, if you look at the popular AI products out there, they promise things like the ability to instantly “bring your digital twin alive” with AI video or “turn a single brief or piece of content into a multichannel campaign in seconds” with an AI writing assistant.

Notice the common thread here – AI products are often sold as “instant magic.”

Your AI process specifically does not promise instant and complete automation. In fact, it’s almost always going to be a multi-step process (like Erich’s 10 GPTs or Matt’s dozens of Zapier tasks) and often will require you, as the expert, to do some light work to move data between steps. You can see why that’s far less sexy than “bring your digital twin alive.”

But the outcome is incredibly powerful, even without full automation.

Erich, for example, used to max out at 4 sizzle-reel pitches a year. Now his AI process can generate dozens of show ideas a week while requiring very little of Erich’s mental energy. Since these shows are 6-figure and 7-figure deals when they get picked up, even going from 4 a year to 16 a year has the potential to dramatically increase Erich’s income.

Matt used to require 10 hours to onboard a new client – now he can do it in 20 minutes. (He’s also getting a lot of new leads since he was featured to 50k+ people during Demo Day, so if you’re interested in his services, you should join the beta while there’s still room.)

Look at Matt’s 30X math: In the past, Matt would need 40 hours to bring in 4 clients. Now, those 40 hours could be used to onboard up to 120 clients – a 30x increase with nearly zero new costs.

AI Processes don’t get venture capital funding or TechCrunch headlines

Alas, Matt’s 30x increase in client traffic probably won’t make headlines any time soon.

The reality of tech and media is that people love rapidly-growing consumer products and bad-boy founders. The rise of Jeff Bezos is a compelling story for the front page – the fact that Matt can 10x his income is not.

As a result, all AI process creators make money in the shadows.

They probably won’t get featured in TechCrunch or Mashable any time soon, which means they’ll need to reconfigure their perception of success.

The flip side of this is that building AI processes is the hidden gem in the world of AI right now. Everyone is racing toward products – which certainly can be fun and lucrative – but the general public has no idea about the growth potential in building an AI process for your existing business or career.

It’s pretty rare to encounter an inflection point where a new technology allows you to quietly 10x or 30x your business – without hiring anyone and without learning to code.

But that’s where we are right now with AI processes, and we’re already seeing students go from zero to launch on a new AI process in 30 to 60 days.

On Tuesday I'm hosting our third and final Demo Day, where you'll see four more Incubator students present the AI products and processes they created in the past 60 days.

​Add it to your calendar and be sure to attend live because it's also our AI Incubator Opening Day event, where I'll be sharing a special bonus ONLY for those of you who join our next cohort on Day One.

See you next time,

– Rob
​

PS. Tuesday, April 16, is our third and final Demo Day. Watch 4 Incubator students show their brand new AI ideas to the world and get access to our Opening Day bonuses (we're unlocking the Incubator during this live event).

​Add to calendar >​