Intro

There are two hot buzzwords for startups right now: AI and blockchain. As the time of this writing, people are drowning in uses and ideas for AI, and desperately trying to think of uses for blockchain.

How about you? Are you getting cool ideas on how this can help you and your business? Or are you letting the technology happen and joining in when it’s simple?

It’s too late to be the first to have a website or a mobile app, but it’s the right time to add AI to your business. Let’s talk about lots of great ways businesses can use, leverage, and benefit from AI. I would love to spark some ideas for your business.

AI Today

Data-driven training belongs to AI and Machine Learning. As I put it in my “Join the Revolution” keynote:

We see this consistently. From removing backgrounds from photos, re-colorizing black and white film, checking grammar, finishing emails, composing music, captioning images, and more, large example datasets make for near-human performing machines.

How can we use this technology today?

Exercises for your Business

Let’s do some exercises together, and apply AI to businesses! We’ll go through the following scenarios:

1. I sell something on my website
2. I have a physical location/event/office
3. I have an app/app idea
4. I have user content or cross-communication
5. I’m looking to disrupt a market

While all the above situations could also abuse AI to increase profit, we’ll avoid those sticky ethical concerns for this educational foray.

I sell something on my website

Recommendation engines, analytics, and more

While this isn’t the easiest to develop, it’s surely the easiest to suggest. Any business that sells something, could use a recommendation engine. There’s a reason the small, cheap, high-sugar items are near the checkout at the grocery. It’s a great opportunity to up-sell. Showing someone something they feel they want is the basic premise for all sales. It’s also the fundamental formula for monoliths like Google ads!

It’s hard to compare yourself to Google, but as we often see, algorithms that are groundbreaking one year, become common practice shortly after. Netflix cared so much about improving their prediction of film ratings, they paid One Million USD back in 2009.

If you had tons of time, you could look at where your sales were coming from, figure out how best to suggest purchases, and identify how to present that to your users… but that’s exactly what AI would be good at.

It’s gone beyond the “what” and into the “how”; as UIs slightly adjust pixels, colors, and images to push recommended products. As a matter of fact, AI is starting to be able to intelligently chat with users who have questions. While this currently doesn’t replace human interaction, it can augment your sales with site intelligent conversational interfaces that make frequently asked questions friendly! Scalable concierge services for the little guy? WOW! Now you can become even more customer-centric.

This really only scratches the surface, as AI can be involved in any aspect of your sales process. What’s important to your sales?

• Dynamic pricing
• Predictive customer service
• Lead scoring
• Predictive analysis
• Something else?

If you sell something, take some time to frame the question, “How could I automate, improve, or study my sales pipeline with AI optimization?”

I have a physical location/event/office

Image recognition, inventory, and behavior

Welcome to the world of the visual. Everything you see, a computer could see. Right now, AI is being used to identify damaged products, handle inventory, and identify issues. Some people are even using AI to identify scammers, robbers, and people cheating self-checkout lines!

If you wish you could hire and train a person to stare into a camera all day and report back when something happens, you’re really looking for AI. It’s so cheap and easy, you can even set up a camera to help you find parking spots.

We’ve all noticed doorbell cameras are starting to use facial or person recognition. You can make up your very own smart camera idea, and turn it into a reality today.

• For your office, you can help people communicate when they are missing a meeting or even out to lunch.
• For a physical store, you can manage inventory, supply logistics, streamlining, notice when a customer is without service, or even map how items sell.
• For an event, and even the above situations, you could use sentiment analysis. Are people smiling and happy? If so, what was the cause?

We all enjoy people watching, and with AI in a physical location, that’s great data you could be learning from. Hard drive space is cheap, and lessons from people are key.

I have an app/app idea

User input gives you one-of-a-kind experiences and data

This is the world of augmenting your UX. Since this is a pretty big domain, let’s do a hypothetical situation to surprise our app users.

Let’s say we deliver packages, and when we leave a package, we take a photo proving we delivered it. That photo is then sent to the user to show them that the package was actually delivered. No AI needed, right?

Now let’s say, one day we deliver a package, we take the photo, and suddenly our app says “Looks like something might be off. Please click OK to verify this package was delivered to XYZ Drive.” Wow… you’ve never seen that message before, and sure enough, you notice you’re at the house next door! But, how did the app know? The GPS is too close to call, so something amazing just happened.

Well, this house has had 50 packages delivered, and that means there are 50 existing images of that house’s front door. Your mobile app has been trained to somewhat recognize the visuals as a person would, and when something was significantly off, it let you know. This is a very doable thing for AI.

Odds are, your business doesn’t have that exact scenario, but your customer data works the same way. How many long-standing customers were upset or lost in a mistake that was never caught? Or maybe the bigger question, how many potential customers could you have gotten by letting AI be part of your product from the beginning!

If you have an app, you’ve got customer data. Anomaly detection, experience augmentation, and quality maintenance are all features you could build into your product.

The question is, what does your app need in order to be better?

• Personalized experiences based on user history
• Simplify UX — “take a photo or use your video, etc.”
• Team up with edge devices like wearable tech, automation, etc.
• Mistake Detection and Quality Assurance

All of these are possible with AI.

I have user content or cross-communication

Decency, sentiment, and abuse checking

Are your users happy? Are they uploading lewd photos? Are they sending rude comments or pornographic material to others? You could be a harbor for illegal or indecent day-ruining content without knowing it.

Google can employ thousands of moderators, and actually does, to check user content. How many can you hire?

While most AI isn’t as good as humans, it still cuts down on the amount of work a human would have to do. False positives, if expectations are set up correctly, are fine. A user can always override, and a moderator can always remove content. It’s about quantity while AI will continue to improve.

Sure, you can just check for a list of dirty words, but friends curse in fun, so it will help with public areas, but won’t help in private messages. AI could actually detect the sentiment those words caused with your users, helping their claims of direct message abuse. And as for pornographic images, you could warn a user that the AI flagged something as possibly indecent. If the user decides to view it anyhow, that’s a choice of your site and your customer.

AI can help curate content in any way you can think!

• creativity
• category
• sensitivity
• quality
• and more!

I’m looking to disrupt a market

AI is doing things that have never been done before

DELL computers made over 2 Million dollars in sales a DAY after only two months of creating their online site. This helped them battle with existing juggernauts like IBM and Apple.

We’re living in a time where AI will spring-board businesses into their future.

Example: Are you late to the sticker game? Well, how about you automatically use AI to remove backgrounds and create proofs in seconds, rather than using a team of designers?

Now… these models aren’t useful to anyone but your business, but that’s the beauty of it. That’s your leg up on the competition.

It’s important to note, AI is doing our mundane tasks. AI can’t come up with those tasks, because that’s a creative challenge, and AI can’t do that. It’s your challenge to get creative for your business and we can help. We at Infinite Red are excited to see if we can help you come up with your AI advantage. Find out more at our website or email hello@infinite.red to chat.

Feel free to take our FREE — 5 Day Mini-course on AI / ML

https://academy.infinite.red/p/ai-demystified-free-5-day-mini-course

If you’re looking for a good list of existing uses of ML and DS in the industry to compare against your own, here’s a good resource.

Interested in learning how to bring AI and Machine Learning to web with TensorFlow.js?

JavaScript lets you create front-end websites that can leverage the power of AI directly on the browser. Learn from scratch with this book.

Reserve your copy on Amazon now

Gant Laborde is Chief Innovation Officer at Infinite Red, published author, adjunct professor, worldwide public speaker, and mad scientist in training. Clap/follow/tweet or visit him at a conference.

I also help run the FunMachineLearn Twitter account.

Have a minute? Read some more.