Look, I get it. You’ve spent 2024 watching AI take over your LinkedIn feed faster than Taylor Swift takes over, well, everything. Everyone’s either “doing AI” or lying about doing AI. And here you are, wondering if you should jump in or if you’ve already missed the party.
Here’s the thing: I’ve spent the last year watching companies try to crack this AI thing, and let me tell you… it’s been like watching a cat try to understand TikTok. Some get it surprisingly right, most are hilariously wrong, and everyone’s a little confused.
The Real Talk About 2024
Let’s start with a fun fact: 74% of companies struggled to scale AI value in 2024. That’s not just a stat – that’s basically everyone who tried to “do AI” discovering that it’s harder than making New Year’s resolutions stick. Around July through August I read more articles about the AI bubble than how to monetize AI.
But here’s where it gets interesting: while most were struggling, the finance folks were quietly crushing it. Capital One casually dropped 314 AI patents this year. Banks – yes, the same people who still sometimes make you fax things – are now AI powerhouses. If that doesn’t tell you something about where this is all headed, I don’t know what will.
Why I’m Weirdly Obsessed with Embedded Finance + AI
Look, I spend an uncomfortable amount of time thinking about two things: embedded finance and AI. It’s like my version of fantasy football, but with more spreadsheets and fewer injuries.
Here’s why you should care:
1. The Money People Are All In: When banks start acting like tech companies, pay attention. It’s like seeing your dad join TikTok – surprising, but probably important.
2. It’s Actually Getting Easier: Remember how hard it used to be to add payments to your app? Now it’s basically Lego blocks for finance. The same thing’s happening with AI.
3. We Know What Works: We’ve moved past the “throw AI at everything and see what sticks” phase. Thank god.
What’s Actually Working (No, Really)
Let’s break this down into two categories:
The “Why Didn’t We Do This Sooner” Stuff
- Risk assessment that doesn’t require sacrificing your firstborn
- Payment flows that don’t make users want to cry
- Customer service that actually… serves customers
The “This Is Actually Pretty Cool” Stuff
- Predictive analytics that feel like having a crystal ball (but with math)
- Financial products that don’t treat everyone like they’re the same person
- Compliance automation that doesn’t make your legal team break out in hives
Your 2025 Game Plan (Because You Need One)
Step 1: Stop Starting with AI
I know, I know. It’s like telling someone not to think about pink elephants. But seriously – start with what you’re actually trying to fix. AI is a how, not a why.
Step 2: Copy Smartly
See what’s working for others and shamelessly adapt it. It’s not stealing if you make it better. Think of it as product evolution with a dash of survival of the fittest.
Step 3: Get Your Data House in Order
Your AI is only as good as your data, and right now, most companies’ data is about as organized as my garage after Christmas (3 kids). Fix that first.
How to Mess This Up (Trust Me, I’ve Seen It All)
Want to know what not to do? Here’s my favorite fails from 2024:
- Trying to build the Iron Man suit when you need a bicycle
- Thinking AI will magically fix bad data (it won’t)
- Not measuring anything because “AI is magic” (it’s not)
Where to Focus (If You Don’t Want to Look Silly)
Just Starting Out?
1. Figure out what data you actually have (and if it’s any good)
2. Find the low-hanging fruit (it’s usually lower than you think)
3. Map out where money touches your product (it’s probably everywhere)
Already in the Game?
1. Look for the boring stuff that’s eating up time
2. Add financial features that don’t require a PhD to understand
3. Start thinking about patents (yes, really)
The 2025 Shopping List
Hot Areas (That Aren’t Just Hype)
- Making compliance suck less
- Automated money decisions that don’t terrify users
- Financial stuff that feels personal (in a good way)
- Green finance (because the planet’s not going to save itself)
Markets Nobody’s Looking At (Yet)
- EdTech (because education is still stuck in 1995)
- Tourism (down 28% in AI patents – opportunity knocks)
- Consumer goods (everyone needs/buys stuff)
- Power industry (they’re literally asking for help)
