Not everyone is a technology early adopter, but in the case of artificial intelligence (AI), it may take people even longer than usual to embrace the possibilities. AI has exciting potential to create systems that can discern, see, smell — even think — in ways that previously could be done only by humans. But it doesn’t take long before suspicion sets in that AI could replace humans or track their every move. Hollywood has explored these themes at length, and conspiracy theorists have expounded upon them, all feeding the fear, uncertainty and doubt (FUD) around AI.
It’s Time for a New Perspective
It’s time to replace the FUD around artificial intelligence with a more realistic view of the technology. AI isn’t ominous science fiction like Hollywood would have us believe. It’s actually based on three down-to-earth principles:
AI systems need data to learn. Data scientists or machine learning engineers use data sets to train AI systems. The system only “knows” what people tell it, and its capabilities and performance rely on the data it uses. AI systems can use unstructured data such as images, video and voice to produce results or trigger responses.
AI systems also need people to tell them how to use data. They follow an algorithm — comparable to a recipe — that gives the AI system a set of rules or processes that enables it to discover patterns, solve problems and deliver the desired outcomes.
AI systems, specifically machine learning and deep learning systems, can adapt and improve the way they solve problems. They can’t, however, learn anything and everything — their learning capabilities are limited to the domain for which they are programmed.
Real-Life, Not Hollywood, Examples of AI
Even software developers, data scientists and robotics engineers admit that AI sometimes conjures images reminiscent of the infamous HAL 9000, but AI applications are coming to market in much more practical ways:
- Intelligent transportation systems that predict traffic congestion and reroute fleet vehicles or alert commuters via an app
- Spam filters that recognize unwanted and, perhaps, malicious phishing email and sends them to a junk mail folder or deletes them
- Tools that assist language teachers by checking grammar and grading essays
- Mobile check deposit functionality for banking
- Network security systems that can spot zero-day exploits based on anomalies
- Voice-to-text translators and foreign language translators
- Personal assistants that carry out tasks from voice commands
- Machine vision systems that ensure quality control on a manufacturing production line
- Predictive maintenance systems that use sensor data to address problems proactively to avoid machine failure and downtime
- Home care robots that understand voice commands, accomplish simple tasks and monitor a patient’s vital signs
AI: Taking the Right Direction
AI applications have emerged as tools to help people, not replace them or surveil them in “Big Brother” fashion. Leveraging AI systems, people can save time and work more efficiently and safely. AI also gives people the advantage of new capabilities that can discover new patterns, see things from a new perspective and pave the way for innovation.
Granted, the potential exists to cross the line and invade privacy or put personal information at risk. AI systems that use consumer data, for example, may use images of shoppers’ faces and follow their activity based on payment transactions, storing sensitive information that hackers could monetize.
The tech and regulatory communities are well aware of concerns over the safe and ethical use of data, and they’re taking a cautious approach. Some companies
have formed advisory panels or ethics review boards to guide the responsible development and use of AI. And legislators are enacting laws to protect people’s privacy and secure their personal data. The General Data Protection Regulation (GDPR), for example, empowers EU residents
to control how their data is collected and used — and to have it erased upon their request — and the California Consumer Privacy Act
gives consumers control over how their data is used, as well as the right to deny permission to sell it.
This atmosphere of checks and balances will help keep AI technology advancements heading in a direction that not only will result in significant benefits, but also a level of comfort that people can live with. Although there may be a little more fear, uncertainty and doubt about how life could change with AI compared to the FUD related to other new technologies, with the right balance of innovation and ethics, it can become a common, generally accepted part of our daily lives. Someday we’ll wonder how we ever did without it.
about how ADLINK is supporting AI advancements in industrial environments.