AI technologies are changing industries towards more efficient and personal businesses. Discover for yourself how 5 key industries are changing and let yourself be inspired in your own entrepreneurial journey.
Whether you’ve been an app founder for a month or a lifetime, the hype of AI technologies has surely reached you. Some think of AI as another service that resides in the box of your computer or in the cloud; just another point to check off on the specifications sheet.
But it’s more useful to think of AI technologies by considering what they will make easier and cheaper to process in our day-to-day lives. A16z’s guide on artificial intelligence mentions making predictions cheaper. But it can also make it more cost-efficient to:
- make things that move to drive or fly or sail themselves
- understand people and objects and how they relate in the real world
- optimise complex systems: things like driving patterns or resource consumption in a manufacturing centre
- create content, from written words to music, websites and movie trailers
- understand people’s interests, help people understand and use software, and help people understand each other
The truth of the matter is that artificial intelligence technologies are creating a path towards a brave new world, where we change the way computers help us in doing our work and living our lives. To get an understanding of what awaits us, we’ve put together a list of key industries where AI technologies are making a difference. We hope you’ll feel both amazed and inspired. Now let’s get started!
The fashion industry
Let’s step out of the clothes closet and go global first: the fashion industry is valued worldwide at approx. $3 trillion and represents 2% of the global Gross Domestic Product. With the rise of ecommerce, the industry has shifted from brick-and-mortar and home ordering services to a digital and social media-friendly customer experience.
This digitalisation is what allows for the potential of AI to take root. In the Business of Fashion-McKinsey Global Fashion survey, 20% believe AI can reinvent design, merchandising and marketing. Take for example Tommy Hilfiger’s work with IBM Research AI tools. They’ve worked to understand real-time fashion industry trends, customer sentiment around product and runway image, as well as resurfacing themes in patterns, silhouettes, colours and styles. With the help of AI, a massive amount of information is concentrated in key insights and provided to the human designer, who can make informed decisions around the next collection they’re designing.
Chris Palmer, the global cognitive offerings lead at IBM, stated that the purpose was to reduce brain clutter. “We wanted to know: How do we eliminate the repetitive tasks? Once you do, you can focus yourself elsewhere, instead of manually scanning Instagram and Pinterest. We’re pulling insights that brands can use again and again, from massive data pools. This is not the same as the creative process, and it’s not replacing it. It’s answering: How can data make us smarter?”
Similar applications of IBM’s AI technologies are visible in the Indian designer duo Falguni and Shane Peacock, who are combining analyses of over 600,000 images of runway shows and Indian couture to predict the future of Bollywood fashion. Stitch Fix uses algorithms to comb through millions of combinations of apparel attributes and narrows down decision making to just a few suggestions.
And these examples don’t even go into the potential changes AI technologies may bring into logistics and recommendation systems, which we are covering in-depth in the next sections.
Distribution and logistics
Everyday, food is getting in the supermarkets for you to buy and your local department stores are stocked up. Everyday, fuel is delivered to pumps, post and packages are sorted and delivered and commodities are stocked. Everyday, distribution and logistics companies are swarming more or less invisibly around you, making sure you and your community are well supplied.
In truth, with a global market size of over $8 trillion, keeping track of billions of products and commodities going through the global shipping network is already a task beyond human capabilities. That is why the industry is grabbing onto the potential of AI to optimise supply and demand, inventory, coordinated shipping networks, vehicle maintenance scheduling and more. The trouble lies with the sheer size of the industry. While companies who have shown a proactive attitude in including AI in their workflows show a 5% greater profit margin, only 21% of the transportation and logistics firms surveyed by McKinsey had moved beyond initial testing and started scaling their AI processes. Adopting new technologies and processes requires numerous changes and major capital investments. Nevertheless, the benefits are worth the effort. At this moment, there are two main trends driving the changes:
This technology is grounded on predictive algorithms running on big data. With the support of this technology, logistics professionals can improve the efficiency and quality of deliveries by estimating demand in advance – before the customer places an order.
Say the predictive algorithms identify a spike of demand for the latest mobile device model – the manufacturer will increase the production of the model accordingly. Moreso, transport companies will know to plan their transportation capabilities beforehand and better plan their routes. The retailer, in turn, will know to order appropriate stock, increase advertising efforts, prepare shelf displays and be ready for a jump in online shopping. An entire network, working in sync and with efficiency, starting from accurate predictions from AI.
Machine Learning Systems
Machine learning is giving computers the ability to learn without being explicitly programmed. In the context of the logistics industry, machine learning leverages data across multiple systems and data sets. This helps shippers more accurately predict demand, analyse trends in supply chains, monitor seasonal calendars, and track daily patterns on roads. Other popular use of machine learning technologies is in helping make better carrier selection, planning routes and assessing quality control processes. All these save money and improve efficiency.
Intelligent warehouses are another practical application of machine learning in logistics and distribution. These systems recognise trends and incidents, analyse repeated data and send insights to specific entities (including order information to customers). You can see them applied in warehouses of Amazon or DHL and be amazed by the combination of machine learning and robotics put to good use.
Sales and CRM
There’s a straightforward reason why AI technologies are making leaps and bounds in sales and customer relationship management: the more you know about a person and their circumstances, the better equipped you are to tailor your offer to their preference. The more you can personalise, the higher becomes the likelihood of a sales success. With the potential of intimate computing at our fingertips – capturing real-time information about our customers and from that, developing a full image of who they are and what their needs are – selling becomes a response to what happening to them, their company or their industry, instead of a generic pitch.
The drive to create an ideal world where the salesperson has a digital assistant that helps with research, admin work and follow-up is so strong, that the top 5 CRM companies (Salesforce, Oracle, SAP, Adobe Systems, Microsoft) are investing in AI. Markedly, Salesforce, the biggest player in the CRM industry, is laying their bets on AI. A study sponsored by them projects the use of AI technologies in CRM will boost global business revenue by $1.1 trillion by 2021.
By late 2016, Salesforce presented their AI tool, Einstein. With the goal of making AI accessible for most users, Einstein can offer account insights, lead prioritisation, automated data entry, ad personalisation, insights into social media conversations, product recommendations, image classification, and more.
It’s also worth remembering that the focus of AI technologies isn’t concentrated only on the sales stage of the customer journey. Customer education and support takes up just as many resources and time consuming tasks that could be freed up by the capabilities of AI.
IBM estimates that by 2020, 85% of all customer interactions will be handled without a human agent. AI, chatbots and automated, self-service technologies free up call centre employees from routine support requests, allowing them to focus on more complex tasks.
And more than the in-house efficiency, there’s a benefit to the user’s lifetime value as well: resolving customer service issues before they arise or escalate could significantly lower abandonment rates. Goes to show that at the end of the day, no matter the scale, keeping business personal strengthens user loyalty and increases their happiness.
Media and entertainment
If you’re thinking AI technologies can’t be fun when it comes to media and entertainment, you’re underestimating the quirks of human nature. Let’s first get the wow factor out of the way: companies are using AI technologies, specifically machine learning algorithms to help develop film trailers and design advertisements. Yes, you read that right.
In a collaboration between IBM Research (who else?) and 20th Century Fox, the suspense/horror film “Morgan” got the first-ever “cognitive movie trailer”. Specifically, the team fed an IBM system around 100 trailers, cut into moments, teaching it the patterns and types of emotions that resonate with each viewer. The purpose: to teach the system “what is scary”, and then create a trailer that would be considered “frightening and suspenseful” by most of the viewers.
Change the objective of machine learning from creating to categorising and you get large scale search optimisation. Producers are using AI software to improve the speed and efficiency of the media production process by giving the creative team the ability to quickly sort through and organise visual assets. Zorroa, for example, offers a platform for managing visual assets and allows users to search within large databases. You input documents into an “analysis pipeline” and the algorithms tag each file with components or people present. Afterwards, you can use this in-depth categorisation to look up specific content documents, within significantly less time.
The applicability of AI technologies doesn’t stop at digitally created trailers or intelligent search possibilities. Personalised recommendations are at the forefront of enhancing user experience all around media and entertainment outlets. Look no further than the recommendation systems of Netflix, Spotify and other streaming applications: with the help of machine learning algorithms, entertainment providers can recommend personalised content based on your activity and your behaviour. We’re used to getting recommendations while we’re online, but when you wrap your mind around the fact that every single user is getting personalised recommendations, that’s when the scale of the technology behind it becomes awe-inspiring
Language oriented AI technologies
There’s a whole section of AI research dedicated to the understanding and processing of human language. Specialists call it natural language processing and – to give you the most handy example – it’s the foundation of how Siri or Google Assistant talk to you. In fact, many consider teaching computers to understand human language as the next step in human-computer interaction.
The applications of computers understanding human language are broad and these are just some directions of development:
Think computer-assisted translation and your mind will probably jump to Google Translate. Not only is Google’s product at the forefront of machine translation, it’s also gotten an AI update at the end of 2016. The Google Neural Machine Translation uses an artificial neural network to increase fluency and accuracy in translations. The machine is set us to learn from millions of examples, translating directly from one language to another, and it becomes “smarter” the more it is used. Over time, Google Translate will create better, more natural translations – focusing on sentences as a whole, instead on translating piece by piece.
You’ve probably gotten spam in your inbox despite current spam filters, or lost that one important email in the spam folder. Getting to the bottom of filtering spam without false positives or negatives is at the heart of Natural Language Processing. The hope is that the better machines will get at gleaning meaning from text, the better they’ll tell the difference between spam and not spam.
Every day, millions of gigabytes of information are added to the world wide web. Our capacity to follow and understand that information has long been exceeded by the speed in which it is produced. That is why the capacity of machines to process language and glean meaning from it is becoming increasingly important. Think of the potential this may have for marketing or business development departments, where, by aggregating data from social media, companies can determine the general sentiment towards its latest offerings.
Watch any younger generation get stuck on a topic of conversation, and they’ll all swipe out their smartphones and search for answers. But not once, refining the search results to get the answer you need becomes a struggle. This is where Google’s focus on natural language processing comes in. After all, if the search engine gets better at understanding the meaning of our questions, it will get better at giving us answers. One great example to try out is Google’s new Talk to Books search engine, that combines language processing and a database of over 100,000 books. You can ask it any question you can think of, and it will give you relevant quotes from the books it has included in its database.
All in all, AI technologies are shaping industries across the world in all manner of processes. We’ve just uncovered the tip of the iceberg for you in this article, but we hope it was enough to get you excited. Find out for yourself what the world holds and what other companies are developing when it comes to AI technologies. But keep in mind it’s just as important to contribute yourself, in any way you can. AI technologies may become a cornerstone feature of your product, or they may become a means to help others run more efficient businesses. The fact of the matter is it’s becoming easier and easier to process large databases and large archives of content. Remember to ask yourself: how can I build a stronger startup with the technologies I have available? AI technologies are opening the way towards a brand new world of applications that are smarter and more personal than ever before. How will your product help shape it?