You're Probably Paying Too Much for AI

Have you ever looked at all these new AI tools and wondered...

How much is all of this eventually going to cost me?

Because every week, there's something new.

A better model.

Another AI tool.

A new subscription.

Another company promising to change everything.

And if you're trying to use AI in your business, it can start to feel like keeping up means constantly spending more money.

I get it.

But something happened in AI this week that made me think we might be heading in a very different direction.

What if AI actually gets cheaper as it gets better?

And before you think I've lost my mind, let me explain. 😂

First, something interesting happened.

Facebook (Meta) is making a big bet on wearable AI.

We're talking about devices that can listen, understand context, connect to your information, and potentially help with things like calendars, emails, and even ordering things.

Now, I'll be honest.

The idea of something always listening to me?

I'm not entirely comfortable with that. 😂

But what caught my attention wasn't really the device.

It was the business behind it.

Meta is exploring different ways to make money from AI, including models that could put more of the cost on merchants rather than consumers.

And they're doing something else interesting.

They're investing heavily in their own infrastructure.

Why does that matter?

Because the cost of running AI is a HUGE part of the equation.

If companies can change who pays, how they charge, and how efficiently they deliver AI, the economics start to look very different.

But that wasn't even the most interesting thing I came across.

Then came something that really got my attention.

A new AI model called Jev.

Now, here's what's interesting.

Jev isn't designed to write emails.

It isn't designed to write blog posts.

It isn't trying to have long conversations with you.

It makes decisions.

That's it.

Give it a situation, some parameters, and it gives you a decision.

Yes or no.

This or that.

One category or another.

And you might be thinking...

Okay, Atiba. Why should I care?

Because think about how many little decisions happen inside your business every single day.

An email comes in.

Is this an existing customer who needs help?

Or is this a potential customer who wants to buy something?

That's a decision.

One email goes to customer service.

The other goes to sales.

Simple, right?

But right now, a lot of us are using powerful language models to handle tasks like that.

We're asking AI to read something, think about it, and generate a bunch of text when all we really needed was:

Customer service.

Or:

Sales.

That's it.

We're using something incredibly powerful to do something incredibly simple.

And we're paying for that extra work.

Here's where it gets interesting.

According to the pricing discussed in the announcement, Jev costs around $42 for one BILLION input tokens, with no charge for output tokens.

One billion.

I do a LOT with AI.

And I don't think I've personally used a billion tokens this year.

The model is also claimed to be significantly faster for the kinds of decisions it's designed to make.

Now, that doesn't mean it's going to replace Claude or ChatGPT.

And I wouldn't use a specialized decision model for every business decision, especially one that requires careful judgment.

But for simple, repetitive routing decisions?

This could be a big deal.

Because imagine thousands of small decisions happening in your business without needing an expensive, general-purpose AI model for every single one.

You stop paying for more intelligence than the task actually needs.

That's the part that excites me.

And it reminded me of something.

I've seen something like this before.

I was heavily involved in technology during the dot-com boom.

And back then, we had some of the same conversations we're having about AI today.

Computing power was expensive.

Storage was expensive.

Bandwidth was expensive.

Everything had a cost attached to it.

You needed more storage?

Pay for it.

More processing power?

Pay for it.

More bandwidth?

You guessed it.

Pay for it.

And people worried about how expensive it would all become.

Sound familiar?

But look at what happened over time.

Storage became dramatically cheaper.

Computing became more accessible.

Bandwidth became something most of us barely think about.

Things that once felt expensive and scarce became ordinary parts of running a business.

Competition and innovation changed the economics.

And I'm starting to wonder whether we're watching something similar happen with AI.

Not overnight.

Not necessarily in the same way.

But the pattern feels familiar.

And then another interesting thing happened.

In the same week, Anthropic announced Opus 5.5.

According to the announcement I was discussing, the new model offered stronger performance at a lower price.

That caught my attention.

Because for a while, the pattern seemed to be:

New model.

More capability.

Higher price.

New model.

More capability.

Higher price.

But now we're seeing signs that the direction might change.

Better AI doesn't necessarily have to mean more expensive AI.

And if one major company starts offering more capability for less money, what happens to everyone else?

They have to compete.

That's good for you.

That's good for me.

And that's especially good for businesses that want to use AI without building a massive technology budget.

Of course, one price reduction doesn't prove that every AI service will get cheaper.

But it does raise an interesting possibility.

What happens when specialized models start handling simple work, while the bigger models compete to offer better capabilities at lower prices?

I think that's worth paying attention to.

Here's what I really want you to take away.

I don't think the future of AI is necessarily one giant, expensive model doing absolutely everything.

I think we're going to see more specialized AI.

Some models will write.

Some will reason.

Some will make simple decisions.

Some will handle very specific jobs extremely well.

And as that happens, businesses may have more choices about what they're paying for.

Which brings me back to something I've been saying:

The best AI isn't necessarily the most powerful one. It's the one that does the job you actually need done.

Sometimes you need something incredibly intelligent to solve a complicated problem.

Other times?

You just need something to decide whether an email belongs in sales or customer service.

Why pay for the first when you only need the second?

That's the shift I'm watching.

And if history teaches us anything, it's that today's expensive technology doesn't always stay expensive.

We're still early.

There's plenty we don't know.

But I'm excited about where this could go.

Because the more accessible AI becomes, the more businesses can actually put it to work.

And that's when things get really interesting.

Can I ask you something? 🤔

I'm curious whether the cost of AI is already affecting how you use it.

For me, I'm constantly looking at which models can do the job well without paying more than necessary.

Are AI costs already becoming a concern in your business?

Hit reply with just one sentence.

Even “Yes, we're spending more than expected” or “No, cost hasn't been an issue yet” is perfect.

I'd genuinely love to hear where you stand.

Want the full breakdown? 👇

I recorded a video walking through the AI announcements that caught my attention, why specialized decision models are interesting, and what my experience during the dot-com boom tells me about where AI pricing might be heading.

For the physicians reading this 👨‍⚕️

Our AiM (AI & Marketing) Rounds community has been growing, and the conversations are getting better because more doctors are sharing what they’re actually trying, questioning, and figuring out in real time.

We’re talking about practical AI, marketing, what’s working, what isn’t, and how all of this actually fits into a medical practice.

No pressure to be an expert. No need to have the perfect answer.

Just doctors learning from doctors.

If you want in, reply “JOIN AIM” and I’ll send you the WhatsApp link.

A few things worth checking out 👀

Just sharing a few things I've been exploring lately:

• 🤝 Let’s connect: I share more of my AI experiments, failures, and what we're learning on LinkedIn.

• 📘 Book I keep thinking about: The Delegation Trap — exploring what happens when leaders stop being the bottleneck.

• 🤖 What we’re building with AI: A look at what we’re working on with AI

• 📖 Currently reading: The book on my nightstand right now

• 🛠 Tool I’ve been using: GoHighLevel (GHL) — keeps the important stuff in one place.

Was this email helpful? 💬

I'd love your quick feedback.

❤️ Yep — this gave me a new perspective.
💛 Interesting — I'm curious where AI pricing goes.
💜 I'm still figuring out how AI fits into my business.

Just hit reply with ❤️,💛, or 💜.

I read every response, and it helps me know what I should write about next.

— Atiba