December 29, 2022
This year brought stunning new developments in the public use of artificial intelligence. One AI-generated image has even won an art prize while competing against human artists.
The fact is that artificial intelligence is all around us.
It’s used to help screen for cancer, fight poaching of endangered elephants, and to detect archeological ruins from space. But AI isn’t limited to the frontiers of science. In fact, artificial intelligence is everywhere.
You probably use AI in your daily life as well.
“AI technologies are penetrating nearly every aspect of our lives, almost imperceptibly,” said Guangjie Han, IEEE Senior Member. “It powers our devices while continuously improving by analyzing data we produce on those devices.”
Common uses include:
- Voice Assistants: The voice assistants on your smartphone or on your smart home device are backed by AI. Sometimes, requests can be processed on the phone itself. And sometimes requests are sent to a cloud server for processing.
- Smart Home Devices: With artificial intelligence, smart thermostats can automatically adjust your home’s HVAC system, while cameras can alert consumers to the presence of people, cars or packages.
- eCommerce: AI is everywhere in online shopping. Notable applications include product suggestions, but also chatbots that can help manage sales and returns, alongside customized shopping experiences.
- Trend Identification in Retail: Online stores aren’t just serving up recommendations when you shop, they’re actively using data on what’s selling – in their store and their competitors – to identify trends. They can use AI to design and manufacture so that the next big thing is ready and in stores when the public wants it.
- Content Recommendations: AI-powered content recommendation engines are trained using product catalogs and consumer data to deliver more personalized recommendations.
- Navigation and Travel: Paper maps are a thing of the past. AI-generated directions can optimize for travel time or fuel consumption.
- Drug Discovery: By identifying their hazardous potential and mechanisms of action, AI systems are finding novel medication applications. This technology helped establish several drug discovery platforms that allow businesses to reuse current medications and bioactive substances.
- Facial Recognition to Unlock Phones: You look at your phone, and it unlocks. These features are backed by an artificial intelligence that takes advantage of camera and sensor technology to accurately measure your face.
- Financial Fraud Detection: AI is extremely good at pattern analysis, and it turns out that most of us use credit cards in a fairly predictable way. That makes it well suited for the task of determining which credit card transactions may be illicit. If you’ve ever gotten a call from your bank asking if you made a transaction, that was likely the result of AI.
- Autocorrect: AI systems use machine learning, deep learning, and natural language processing to identify incorrect language usage in word processors, texting apps, and other textual media and provide corrections and suggestions.
For the foreseeable future, much of the effort in AI systems will focus on building larger, and less biased, data sets on which to train the models.
“The training of these systems is critically dependent on the data the AI model is trained on,” said IEEE Fellow Karen Panetta. “The more scenarios it can experience, the better it will get, but this requires well curated, annotated datasets. These must be diverse to reduce bias, and dynamically updated to reflect new conditions, and scenarios that arise in real-world applications.”
And cautioned against an over-reliance on AI systems in certain life-altering contexts.
“AI needs to be explainable,” Panetta said. “For example, many companies are selling products that evaluate people’s performance and determine compensation. And yet the decisions coming out of these systems cannot be explained as to how the AI determined that outcome. AI should not impact someone’s livelihood.”
Despite the prevalence of AI, there is plenty of room for it to grow and improve.
Learn more: Artificial intelligence is all around us, and that means designers of AI systems need to think about the impact it has on users. The IEEE Standards Association provides guidelines on the development of empathic and ethical artificial intelligence.