Do you enjoy sending invoices, paying bills, arranging insurance, shopping, booking flights and hotels and hire cars, and so on? I don’t. In a perfect world, I would do only the tasks at work that I love, and enjoy my non-working time with family and friends. I’m excited by the thought that my personal assistant, for example Cortana, will be able to take care of all of these chores for me.
It will happen. Already I rely on Outlook’s AI to learn which emails I need to prioritize and which I can ignore. I look forward to seeing this functionality expand to help me decide, for instance, what I’d enjoy watching on TV. Sooner or later, I will be able to say “Cortana, buy me a new shirt” or “Cortana, find me a hotel in Frankfurt”. I will trust Cortana to know my preferences and make the same decisions I would have done if I’d put the time into researching them.
You may feel uneasy about how much data Cortana – or Alexa, or Google, or Siri – would need to know about you to be able to do this. That’s understandable, but personally I’m very relaxed about it. Working at Microsoft, I see every day how seriously the company takes its principles for developing AI: fairness, accountability, transparency, ethics (you can read more here).
I also understand how far-reaching will be the implications of the EU’s new General Data Protection Regulation. The GDPR gives us, as individuals, the right to ask any company to tell us and everything they know about us, and to demand that they delete all data they may have on us. It’s a huge change and other countries will likely follow with similar regulation. Microsoft has published an eBook for governments with policy recommendations to better protect the data of their citizens here.
So I’m not too concerned about the potential downsides of trusting an AI assistant to know me intimately and make decisions for me. And by reading a recent Harvard Business Review Article, Marketing in the age of Alexa, I realized that the upsides will be even greater than I had imagined.
Consider how we often buy the same products because we don’t want to spend the time to explore other options. Would you bother to research alternative brands of – say – laundry detergent, unless you were dissatisfied with your current brand? I’m sure you have better uses for your time. But an AI assistant could routinely research alternatives to everything we buy.
The implications are enormous, and not immediately obvious: when companies market to our AI assistants, and not to us, they will have to market differently. AI assistants won’t be swayed, as we are, by adverts trying to create positive emotional associations with a brand of laundry detergent. They’ll want to see data – for example, ingredients, pricing and reviews.
Cortana will know my preferences – for example, if I’m willing to pay extra for a product that has more eco-friendly ingredients or is produced locally. Cortana won’t favor a certain brand unless I tell her to. This will motivate top brands to not sit on loyalty earned through their past successes, but to continuously improve and keep earning the good opinion of their customers.
It will become much easier for startups to disrupt leading brands. If they have persuasive data that their product is superior, that will be all my AI assistant needs to recommend that I give them a try. And it should be cheaper and easier for underdogs to get data in front of AI assistants than it is now to invest in advertising campaigns to create brand awareness.
Currently, companies leverage our data to target ads towards us. In future, AI personal assistants will leverage our data to continuously check whether we could be getting a better deal on everything from laundry detergents to mobile phone contracts. You could almost say with an AI personal assistant my data works on my behalf and not on behalf of the companies selling me something. This won’t just free up our time – it will force companies to focus more on improving their products and services than fine-tuning their advertising. That sounds like an attractive future to me.