Demystifying AI
Google's Deepmind trains AI to cut its data centers' electricity bills by 40%.
Berg Health, a biotechnology company, uses AI to research cancer; has been able to create 14 trillion data points from a single cell.
As of 2017, around 44% (out of 835 companies surveyed by TCS) were using AI to detect security attacks.
Gartner Consultancy firm predicts that, by 2020, at least 75% of security tools will be using some form of AI or ML capabilities. (Just 3 months to go for this!)
For those who are new to this concept, Artificial Intelligence is the theory and development of computer systems and programs which are able to perform tasks which normally require the human or manual intervention of intelligence, like, image recognition, speech recognition, complex decision making, language translation, understanding emotions.
Are the numbers above a good enough validation of the penetration of Artifical Intelligence into our business world in the present era? For me, they sure are. Artificial Intelligence was a very vague term just a few years ago, and now it has evolved itself into a plethora of applications that are seeping their way into the most complex and also the most simple of tasks. And this fact is very well known by all the top executives around the world. A report by NewVantage Partners earlier this year found that 96.4% of top executives around the world's companies reported that AI was the number one disruptive technology they are investing in, 80% of which admitted that it was also the most impactful technology.
But, even after all this rapid emergence of efforts into AI, how does an organization actually utilize its full potential to create business value? Let us dive into a small conceptual zone with something called 'The AI Ladder' (derived by Rob Thomas, GM of IBM Watson). This concept atomically describes a series of phases that an organization must evolve through to actually become AI-enabled and extract measurable business value from it. This requires an iterative approach by organizations to understand the stage of AI maturity they are in. This helps them in crafting the ladder steps to be on the way to actualize AI value. A core message of this concept is that the road to being AI-ready is heavily dependent on effective data management. Data preparation efforts need to be aligned for this. The AI Ladder also highlights the problems companies face while getting themselves ready for AI. Cloud Computing accelerates AI Adoption by providing migration capability of assets to the cloud, which acts as an alternative step in the AI ladder.
AI is not just about executing a single business driven project, it is more about changing an entire business culture. It, actually, is about creating a culture of iteration and experimentation.
Much of the concepts, the most we all as businesses are concerned with is money, right? Is AI just a money-churning investment or is its value seeps through penetratively across industries? You should read what a PWC 2017 report says in the below paragraph!
"Global GDP will be higher than as much as 14% because of AI. That is an equivalent of an additional $15.7 Trillion." This statistically qualifies AI as the biggest commercial opportunity in the present generation.
This rapid AI growth comes with a drawback on human employment. Whilst companies are constantly trying to cut costs to provide quality as well as increase profits. Someone is bound to take a hit with the AI coming into the picture. A report by Wells Fargo says that by 2030, about 200,000 banking industry employees in the U.S. could lose their jobs to AI. This makes it the biggest headcount drop in history.
60% of consumers stopped using a brand after one poor customer experience incident (2016 data). Inconsistent customer delivery processes can be corrected and streamlined with AI in place.
Since I myself am a product and a marketing guy, let us briefly look at all the avenues where AI is currently being used in the industry by various companies around the globe:
- Social Listening or Sentiment Analysis: This is an application of NLP (Natural Language Processing) for brands to oversee conversations around their brand to gauge customer satisfaction.
- Search Engines: Understanding similar product searches, auto-correct mistakes, finding links between product titles, are just some of the applications of AI in search.
- Data analysis and filtering: Who does not know about personalisation and tailored marketing efforts based on the categorization of data in such a way! See how AI helps in Account Based Marketing here.
- Product Recommendations: Clustering techniques which help in pairing demographics with profile information to gauge the needs based on user's activity across the internet. You would love to see what IDIO (an AI tool) does to generate personalized product recommendations.
- Visual Search and Image Recognition: We all know about Google Lens in this regard, don't we? One of Visual Search's major application lies in improving shopping experience on the internet by improving merchandising.
- Product Categorization and Pricing: Heaps of data is classified by various AI techniques so that they can be used in some form. Data comes in very heterogeneous forms from retailers, seemingly different words need to be categorized into one bucket for providing a more humane experience to the end user. Same goes for pricing techniques where the integrated AI takes a lot of seemingly unrelated but very much related factors to build up a complex but very comprehensive pricing system for an product. AirBnB's pricing system is one good example in this regard where the algorithm suggest users of pricing tips and based on user's acceptance or rejection the algorithm adjusts its recommendations moving forward.
- Predictive Analytics: Determining future trends by extracting consumable information from a wide variety of data sets is a direct path to improving customer experience and customer stickiness to your product or platform.
- Chatbots and Conversational AI: Chatbots are a direct application of NLP (Natural Language Processing) to provide uninterrupted service to the users at the most unusual hours. Conversational AI's have gained a larger traction across cohorts of user base, with devices like Google Home, Amazon Alexa, providing go-to-home and everyday-digital-routine solutions.
- Augmented Reality and Computer Vision: AR Advertising is big application of AI. Providing unobtrusive elements in the natural surrounding can open advertiser's to new possibilities. For a very clear example, Home Depot launched its own AR app 'Project Color' to let users visualize how a certain paint color would look on their walls in real time (Isn't this cool?)
- Copy-writing, Sales Forecasting, Speech recognition, Programmatic ad-targeting, deep learning programs, data filtering, cluster combination, lead generation, and the applications of AI just do not stop here.
The applications of AI, are backed by the impact it is translating onto the end customers. The below numbers are reported in a report by Salesforce (Title : New Research on the State of Customer Experience) can't be more self explanatory.
Closing this, AI is bound to have both its advantages and disadvantages on the business front, where efficiency and costs get reduced, but human jobs become redundant leading to lay-offs. How does one compare the two on a scale!
Meanwhile, is your firm AI-enabled already?