The 4 Types of AI Applications
Artificial intelligence consists of 4 primary waves of applications: Internet AI, Business AI, Perceptive AI, and Autonomous AI. Here are someexamples of each wave:
1 . Internet AI (e.g., Recommendation Services)
The use of the internet is becoming an individual experience due to algorithms that help tailor content according to a user’s preferences. Amazon, YouTube, Netflix all offer recommendations based on previous user history and preferences.
This technology has implications like filter bubble - wherein a user only encounters information that fits their bias and interests; troll factory – a barrage of posts or messages from questionable sources designed to influence political opinion or cause trouble; echo-chamber – much the same way as a filter bubble. Others include fake news etc.
2 . Business AI (e.g., Chatbots)
Chatbots already support many businesses by helping to deal with customer
service issues. By responding to customer complaints and inquiries in real-time, chatbots suggest solutions to issues, recommend products or services, or otherwise, facilitate transfer to a human customer care agent if required. Call waiting times can be reduced through more wide-scale use of chatbots. Also, customer satisfaction can be greatly improved.
3 . Perceptive AI (e.g., Computer Vision)
Research related to computer vision has been ongoing since the 1950s. Computer vision is an application of AI that uses deep learning to recognize, classify, and react to digital images and videos. The recent improvements in computer vision have been boosted by improved computing power, smartphones, the ease in capturing images and videos, and the greater availability of the hardware required for computer vision. In today’s world, computer vision is used in healthcare for examining radiological images and making a diagnosis; in the infrastructure sector for predictive maintenance of infrastructure, detecting construction site safety violations, identifying cargo contraband, etc.; and also, in the retail industry for filing out-of-stock shelves.
4 . Autonomous AI (e.g., Self-Driving Cars)
Self-driving cars are not widely used because the available infrastructure (in terms of roads) and regulations are not yet up to the required standard for the wide- scale use of self-driving cars. For safety reasons, these areas must catch up to the advancements that are ongoing in the creation of autonomous vehicles. The issue of safety is important because it is exactly because of safety concerns that work is ongoing on self-driving cars. A lot of accidents happen every year and more than 90% of them are the result of human errors like drunk driving, texting while driving, etc. A fully automated car would not suffer from the same problems. Automobile giant, Ford, projects they would resume commercial sales for ride- sharing cars by 2021. The cars would not have steering wheels or pedals.