AI for Business: Evolution V revolution

This is post #6 in our AI for Business series.

Whew, we covered quite a lot in those last few blogs about where AI can fit into your business processes. Some of you may be thinking “Awh right, I can see some easy wins right away!” while others might be feeling daunted by the size of the task if you’re considering pivoting your entire business model.

Both quick wins and bold changes are viable options for companies to explore. To find out whether your business should pursue an evolutionary or revolutionary approach to harnessing AI, let’s take a look at industry examples of both this week.

---

thoughtburgers 🍔

Let’s start by revisiting our delectable, AI-powered and entirely fictitious side-business, thoughtburgers (home to the tastiest and most thought-provoking burgers!).

To date in this series, we have looked at how thoughtburgers is planning to use AI to speed up deliveries to millions of hungry customers during rush hours (our goal). We looked at how automation, cognitive insights and cognitive engagement could be applied to make our current delivery processes better. This covered everything from automated and improved delivery driver scheduling and route optimisation to more effective triaging of delivery orders between different restaurant branches.

This incremental approach is evolutionary; we could add each enhancement independently and each would gradually help us achieve our goal of delivering food to customers more quickly.

But let’s say we were to zoom out from our rush hour problem. Ultimately, the goal is to get our delicious burgers to our customers quickly. We have been looking closely at optimising our existing delivery processes but, instead, is there an opportunity to completely reinvent them? Could we take a bold step to get food to our customers that would be revolutionary?

While teleportation would be the ultimate revolutionary solution, it doesn’t exist (yet…). So the next most revolutionary approach available is to use a fleet of delivery drones. Drones probably have the potential to slash our delivery times massively, maybe more than all our other efforts combined, in a very short space of time.

While this represents a major departure from our existing process, we could still take elements from the AI-powered scheduling and order triaging we explored previously, but apply them to our fleet of drones. So the two approaches can still work in tandem to some degree.

However, a pivot of this magnitude is risky and it will require a lot of investigation, research and testing. Some things we may need to consider are:

• What is the max flight duration?
• Will drones be able to support the weight of really large, family orders?
• Will customers be able to take a drone drop in a busy urban area?
• What if drone regulations change?
• What if they vary by county or region?
• Will sea gulls attack our burgers en route?
• Because of the thought-provoking nature of our burgers, will these well-fed seagulls get ideas above their station and attempt to rise against us?

This is without even mentioning the unique set of ethical, legal and societal challenges that would come with this approach (which we will address later in this blog series).

It is worth also remembering the Klarna example we looked at in the previous post. This case study illustrates the risks of going all-in on a revolutionary approach too soon.

---

Evolution - Netflix

One could argue that Netflix’s revolutionary move occurred in 2007 when they pivoted from posting DVDs to customers, to an online streaming platform.

In the age of AI, their utilisation of the technology has been more subtle and evolutionary. While they may be taking bold steps with AI for video production, their consumer application doesn’t show any noticeable changes. However, dig a little deeper and you will see AI at play.

For those who used the service way back when, you may remember that the cards for movies on your homepage used to show the actual movie poster as the thumbnail. But thanks to AI-powered cognitive insights and automation, this is no longer the case.

Without a lot of people realising it, Netflix’s personalisation capabilities have gradually gone off the chain. There are now multiple pieces of custom art work for each show or movie that can be displayed as thumbnails. Each thumbnail is tailored to what you like to watch.

For example, if you’re a fan of romantic movies, the thumbnail for “Gladiator” might show Russell Crowe in a field with soft Tuscan light and his little son running towards him in the distance, finally a family reunited 👶 🥲 😭. But, if you are more into action movies, it might show Russell Crowe looking buff in the arena and roaring “ARE YOU NOT ENTERTAINED?!?” 💪

Netflix can provide this extraordinary level of specificity because it tracks everything you do on the site; what you look at, what you click on, what you search for, what you watch, how long you watch it for.

This data has also allowed them to customise the homepage layout. Each row and the content within them is selected and arranged by algorithm. The content in the top rows are what they think you will like to watch the most while the bottom rows are for the content you are less likely to enjoy. And those “trending now” rows? They show content trending specifically for you, not a general global trend. Even the number of rows shown is customised to a specific user.

Netflix’s goal is to get quality engagements, i.e. for users to start and continue to watch a show or movie. This level of personalised discovery makes sense with that goal in mind and it can really move the needle to make it a reality.

On the surface, Netflix is the same as they were in 2007; an online streaming company. But, under the surface, a big shift has occurred. It was implemented gradually, incrementally layering on more powerful algorithms and customisations, resulting in a service that continuously improves and grows.

---

Revolution - FlyZoo hotel

Imagine you are on the road for work. You have spent a long day in meetings and want to blow off some steam. You decide to head to the hotel bar for a quick drink. If you were hoping to have a chat with the bartender about which local spots are worth checking out for the evening, you will be sorely disappointed. This is because you are staying at the FlyZoo Hotel and the bartender is a giant robot arm with no mouth. 🦾

Alibaba’s slightly weird and futuristic FlyZoo Hotel is located in Hangzhou and it is an attempt to revolutionise the hotel industry by harnessing technology and AI.

Rather than trying to reduce lines at check in by optimising concierge staffing schedules, FlyZoo circumvents the problem altogether. Check in can be done via their app. When it’s time to check out, there are no lines and no hassle; guests just pack and go.

You might be thinking, “Don’t I need to collect and return a keycard to access my room?” Nope. The hotel uses facial recognition software to grant you access to your room, elevators and other areas you have access to.

There are also no little cards with the Wifi details in each room. Instead, each room has a smart assistant, which responds to voice commands like “What is the Wifi password?” or “Can I get some more water please?”. Your water, like all food and drink in the hotel, will be delivered by a robot who looks like R2D2’s second cousin. And, if you are struggling to decide what to order, augmented reality can be used to showcase what your food selection will look like.

To be fair, Alibaba is not a hotel chain, it’s a massive technology company. They are developing the technology to sell it to hotel chains, with FlyZoo acting as a big testing ground. In that sense this is a much less risky project than if, say, the Marriott Hotel chain were to pivot completely to go all-in on a new approach like this.

It might also be an example where this sort of revolutionary approach is simply not worth it. It does seem a little ridiculous and gimmicky.

That being said, it still represents a sizable investment on Alibaba’s part so it is not entirely risk-free. It’s an interesting effort and, if it works, it has the potential to be a revolutionary application of AI and related technologies for the hotel industry.

---

Data capture

In both the evolutionary and revolutionary approaches, one thing is clear; you need good data (and lots of it) to really make use of AI.

Rather than treating data capture like a chore, a means to an end, it is worth considering if you could turn the process of data collection into something more valuable, like a business or product on its own.

Consider this; if you travelled back in time 15 years and decided to set up a self-driving car company, how would you do it? Many folks would think they need to found the next Tesla and start producing cars that can drive on the road, as a starting point.

However, what if you created reCAPTCHA instead? Yes, those annoying little verification prompts that make you identify a word or select all the buses or pedestrian crossings are gathering data for self-driving cars.

reCAPTCHA was a product that solved a real business need; it protected websites from spam, abuse, and bots. Originally it used hard-to-read text puzzles to ensure a user was a real person. However, the byproduct of this mass data collection of blurred words was able to help digitise the archives of The New York Times.

The next iteration, which again solved a real need (the same bot spam problem), had users identifying buses, bicycles and pedestrian crossings. This data collection contributes to refining the capabilities of Waymo’s self-driving vehicles.

While reCAPTCHA is an extreme (and ethically questionable) example, try to identify a good product you can create now that collects the data you need to create an exceptional product in the future. Keep data collection opportunities at the fore in your AI research, planning and strategy.

---

Bottom line

An evolutionary approach is less risky and will probably fit the majority of businesses. Many evolutions lead to revolution in the long run after all.

If you take this approach, you need to ask yourself questions like whether this will move the needle and generate enough Return on Investment (ROI), or do you risk falling behind the market by playing it safe? If a revolutionary option is available to you, you will need to consider how much risk you are willing to take.

Only you can really decide which approach will work best for your product or business. But if you stay focused on a clear goal and you analyse each step of your business processes, it should become clearer whether you need to enhance your existing processes or start from scratch.

---

💡 If you’re ready to start using AI to transform your business, thoughtbot would love to work with you. Let’s talk about making your AI initiative a success!

This blog post is part of a series based on the course Artificial Intelligence (AI) for Business run by University College Dublin (UCD). I took this course and found it so helpful that I’m recapping my top insights. thoughbot has no affiliation with UCD.