August 2, 2019

Warning: AI replicates more than our intelligence

Bias. You’ve heard about it, witnessed it, and let’s be honest, you’ve probably even been it. Whether we think so or not, we’re all biased. But what’s a bias and how does it come about? 

The conventional view of bias is favouring something over another without real or fair justification. In other words, it’s the human brain taking a shortcut in its thinking by grabbing information from a previous experience and using it for future decision-making. 

It’s your brain’s natural way to save energy to make the correct survival decision if the time comes. Without this essential skill, we probably wouldn’t be here today. 

If this all seems a bit fuzzy, here’s an example of bias in action. Imagine you’re walking along and you see a lion. At that moment, you don’t need to consciously analyse the situation to know that the lion could eat you. No, you immediately, and without thinking, react in a way that would hopefully save your life. You take a shortcut to perceive the situation to respond quickly. 

However, biases have evolved to more than just a way to survive. In fact, biases come in many different forms and are easy to overlook unless you understand them: 

  • Have you ever favoured information that confirms your existing beliefs, and disregarded information that doesn’t? That’s called confirmation bias. With this bias, we tend to accept ideas that support our views because being right feels good, and we don’t want anything to challenge that feeling. 
  • How about crediting yourself when things go right and blaming external factors when things go wrong? That’s self-serving bias at work. In this situation, we tend to think we’re the cause of our triumphs and abandon our responsibility when mistakes happen.

These examples represent only two of the many forms biases can take in daily life. So, why does recognising and understanding bias matter in our world today?

Bias in AI

As we start to translate our intelligence into artificial “beings,” we see the potential to use technology to create something that talks, moves and even thinks like us. But as we create this ideal being, we need to teach them using the experiences and knowledge of real humans. 

Now, here’s where things get interesting. 

AI learns by getting data from humans and processing that information by finding patterns. Just as a child uses patterns to make sense of the world, so does AI to find connections in the data to interpret their environment and make future decisions. But because the data AI received comes from real human experiences, it’s likely that information has some level of bias within it. 

The reality is the knowledge humans feed into AI isn’t “clean data,” or unbiased knowledge. In fact, there’s a high likelihood that AI is gathering biased information and mimicking negative behaviour on a greater scale.This issue is not a problem for the future, but a real issue happening today. 

Take Tay, the “Thinking About You” bot released by Microsoft in 2016.

Based on analysis and interactions on Twitter, Tay learned in the ins and outs of Twitter and then began to send out tweets. After 96,000 tweets and 16 hours, the bot was taken offline. Why? Because Tay started posting racist and sexually-charged messages. The bot learned from politically incorrect phrases – the ones getting the most attention and engagement – to create inflammatory tweets using its “repeat after me” capability. It wasn’t the bot’s fault, (remember it’s not conscious), it merely reflected the conversations people were having online.

Another example of bias in AI is the recruiting tool that Amazon was secretly using to filter through job applications. The algorithm behind the tool was scoring resumés on a scale, but ended up systematically scoring men higher than women for technical jobs. The algorithm learned to create this bias after reviewing historical data of applicants hired by the company, which showed many of the technical roles were filled by men. That’s human bias at work again.

From these examples, we can see how transferring our own unseen biases into AI algorithms is detrimental. But what can we do to improve the learning process and curb the biases that seem so ingrained in us?

How do we stop AI from becoming a terrible artificial human?

Let’s think back to why biases form in the first place. Biases occur because it’s an energy-saving method for the human brain. But AI doesn’t have a brain, or at least one that we refer to as human. With a constant source of electricity, AI can never tire as we do. If we think about it this way, then bias really has no place in AI at all. 

Now, that’s a nice theory but it tends to overlook the fact that as long as humans are in charge, biases will continue to sneak back into the algorithms. While there are many ideas floating around what is the best approach to overcome bias in AI, we believe a good first step is to introduce more transparency

Think about it: if the average person were to learn more about what AI is, how it works and where it’s used, it’s more likely they can be part of the conversation and add their insight to make the data more comprehensive and quell the bias. 

Imagine how transparency could have influenced the Amazon recruitment tool. If Amazon was open about the tool and how it works, applicants could have sent in their input or feedback, or at least better understand how the algorithm judges an application. 

In this case, reducing bias becomes this community-driven effort where people can collectively offer their perspective, and in the process, neutralise the space from the opinion of one data scientist. 

The future of bias in AI may be uncertain, but through increased awareness and transparency, we have an opportunity to control this great technology and use it to bring humankind a step closer to a more ideal society.

June 28, 2019

AI is coming for your brand

Every technological advancement that has wide reach provides opportunities for brands to communicate with people on a deeper level, but it also means these same brands need to adjust their tactics and strategy in order to do so. From old-time radio commercials to the introduction of social media, brands have had to evolve from broadcasting to interacting. Today, artificial intelligence (AI) plans to take this communication a step further, and in the process, change branding as we know it forever.

Flipping the interface on its head

We believe one of the most radical shifts AI will bring to branding is voice technology. Let’s take the smart speaker– think Amazon Alexa or Google Home. Instead of clicking, swiping or scrolling, people can have an actual conversation with these devices using their voice and language– an inherently human feature. 

This natural form of communication strikes a chord as smart speaker adoption is forecasted to be higher than mobile phone adoption was¹. Although most people see smart speakers as a way to get immediate answers to their questions or to play their favourite song, the technology of this device has the potential to achieve so much more.

Like Spotify for products

We believe this shift to voice technology will cause users to trust smart speakers not just as a source of information but as a way to make decisions for them. It’s possible that someday a smart speaker will have enough knowledge to predict when a user needs something and perform an action without any prompt. 

Let’s say a smart speaker is aware a person is running low on toothpaste. It could place an order ahead of time and have a new pack delivered to a person’s home just before they run out. At first, the device might choose a toothpaste based on the person’s preferred brand. But then, as it gets to know the person better, it will pick up on values and preferences that may influence it to choose a different brand it believes is a better fit for them. 

How does AI know what I want? 

Every time a user talks to a smart speaker, the device is collecting new information and using that input to feed its machine-learning algorithms. With all the data it has about you it can craft a profile about you. As AI starts to get to know you better, it will get a better understanding of your personal preferences. It will start knowing to what extent you are price sensitive, how much you value quality over price, and how that preference can differ between product categories. On a basic level, Amazon’s recommended products section already does this today. However, a smart device in your home will be able to aggregate way more data than a webpage, since it will be integrated in your home and interact with you on a daily basis. This data will make it smart, and allow it to make a prediction of how satisfied you might be with a particular product.

AI taking over buying decisions

Just as Spotify uses algorithms to give personalised song recommendations, the smart speaker can use machine learning to recommend certain products or brands. But if a device is capable of making choices based on user needs, what does that mean for brands? 

As AI takes over purchasing decisions, doing branding the old-fashioned way will not be enough. Marketers will need to brand their products on a human level and AI level. The human aspect uses emotions and creativity, while the AI bases itself on logic, preferences and characteristics. This dual perspective of brand experiences provides new challenges for advertising agencies, creatives, companies and brands, where integrity between the emotional and the rational becomes more important than ever.  

In this world, if a brand is going to be successful, it will need to present a solid understanding of what makes it unique (UVP), its target audience, and how to effectively translate its values and mission into a complete AI brand experience. 

For a toothpaste brand, for example, on the emotional side, they could get so entwined in our lives that they manage to get us to say “Alexa, order brand X of toothpaste.” Or, on the rational, algorithmic side, they combine company, brand and product characteristics with our own preferences to decide, they are the brand for us. One thing is for sure, a crystal-clear brand definition will be more important than ever.

Why we’re excited

As technology advances and provides new layers of connectivity between brands and customers, it creates opportunities for brands to integrate deeper into the lives of their customers. Since the whole idea behind AI is to make technology more human, it means your brand could become more human too. Which means marketers need to ask themselves, if my brand could think, talk and had a personality, what would it say and do? AI is changing the way the advertising industry looks at branding, but it’s opening up a new door for establishing closer brand affinity and closer marketing relationships.

January 3, 2019

Make your chatbot personality count

Star Wars taught us why we have to build personas for bots. Because who doesn’t love R2-D2? You have to love his moral righteousness. His internal conflict whenever Han and Luke go rogue. His everlasting politeness and his shiny golden armor.

Just kidding.

That’s C-3PO in the gif, obviously. We’re just talking about two clumsy robots, but everyone still seems to be able to tell them apart from a few characteristics. Not only do people know the difference, they go crazy for the pair. And that’s why your Conversational User Interface (CUI) needs a persona.

Don’t worry
If you have a CUI, it already has a personality. From the first message in a one on one conversation, we unconsciously start to attach a personality to whatever entity we’re talking to. The bad news: if you don’t put in the time to build a rich persona, your CUI will suffer from a schizophrenic, off-brand and off-purpose personality.

Think about the children
Let’s take a step forward. A big one. In a few years your children will be talking to a lot of chatbots about a lot of different topics. Every brand will either have one big CUI or multiple small ones to tackle everyday problems. That’s why you have to start thinking about persona coherency across future CUIs today. It’s about standing out while staying true to yourself. Starting from your brand values, you can decide what personality traits will be fixed across CUIs. Is your persona the nutty professor or a quiet wallflower? A caring mom or an experimenting 16-year-old?

At the end is the pot of goals
Once you have defined your fixed personality traits, you can start spicing it up. On the individual level, the purpose of your CUI defines everything. Whether that purpose is solving customer support issues or taking over the world, you need people to get to the end of the flow to reach your goal. You need the perfect guide to get people there. If you want to convince people rabbits used to be fluorescent, a bunny professor might be a good idea. If you have to tell people their power will be cut, he’s probably not the man for the job.

It’s not a conversation
It can be challenging to keep people interested until the end because as much as we would like that, it’s not really a conversation. Until we can create a real C-3PO for every brand, CUIs will have clear scenarios in which brands do the talking. Unfortunately, only nerdy copywriters are interested in reading brands talk.

So, you need to break up the text.

Show a video.

Share a photo.

And keep it interesting. This might look like a practical thing, but it has a great influence on the persona. Everything shared by a bot, is designed specifically for the bot and is part of its language and persona. A bot has as many tools to build a persona and engage users as you can think of. Use them.

How are you feeling today?
Your persona is not only a guide to the pot of gold. It’s also a vital touchpoint with your audience. If not your major touchpoint, it’ll definitely be the most personal one. Your brand will have one on one conversations with different people in different emotional states. Depending on the purpose, your CUI might be talking to angry 50-year-old men or to jolly 8-year-old girls. Your persona should be equipped to talk to any audience, even if it’s all of them.

The talk
It’s time for the talk. The talk between you and your customers. Let’s start building strong personas for every brand. And remember that we don’t really have a choice because every bot has a persona. A good one will help to make things easily digestible and reach business goals. There are many conversations nobody wants to sit through. Make sure the conversation with your CUI doesn’t become one of them.

January 3, 2019

Five tips to build a conversational UI

Conversational UI (CUI) is nothing new, but it is making a big comeback in the mainstream market because of advancements in AI. With the number of chatbots and voice assistants that exist today, CUI is an active medium for organisations and brands. But what should you know before jumping on the bandwagon? We rounded up our top five tips to consider to get your CUI just right.

Trigger and channel
First, you must pinpoint an incentive as to why your audience should engage with a CUI. Will it help users complete a task, solve an issue or exist for entertainment? Once you understand why there is a need for a conversational interface, it’s time to think about how you can trigger the user to engage and map out the channels to carry the conversation. For example, will the entry point start on a website before switching to another platform? Will it incorporate text or voice? 

Data access
One of the advantages of using this medium is its ability to create a truly personalised experience for each user. But to do this, it will need some context beforehand, and the best way to get this information is through data. Allowing data access is necessary because it creates a framework for a seamless, one-on-one conversation for every user. Recording data also gives the CUI a reference point for future discussions. 

Fall back plan
AI technology has come a long way, but it’s inevitable there is a hiccup now and then. To avoid frustrating conversations with continuous “I don’t know” responses, you need to think of a plan B. Offering a link to a contact centre, redirecting the conversation back to its original context or offloading the chat to a real human are all viable alternatives. 

We like to think of developing a personality as more of a characterisation. What we mean by that is it’s essential to use copy and media together to not only ensure the conversation completes the right action, but it does so in a way that’s fun, interactive and consistent, to ensure the CUI maintains a distinct tone of voice in every interaction. 

User testing
User testing is crucial to make to ensure your project meets expected outcomes and, above all, is usable. Defining parameters helps set the bar to measure results and determine which elements need further development. In most cases, this could mean optimising the NLP so that it not only recognises what a user is saying, but also the variety of expressions to say it.

Got a project?

We’re more about the walk than talk, but if you have an idea for your customer experience project — we want to hear it!


©2021 Springbok. All rights reserved.