Myth-busting Artificial Intelligence: 7 Facts You Should Know
Artificial Intelligence, Machine Learning and Deep Learning researcher addresses common questions and myths about the emerging technology on Reddit AMA
By Molly Gluck
Only 33% of consumers think they use Artificial Intelligence (AI) technology, when in reality 77% actually use an AI-fueled service or device. As AI continues to infiltrate into different industries and aspects of our daily lives research surrounding its use is more critical than ever before. Kate Saenko, associate professor at the Boston University Department of Computer Science, is researching AI, Machine Learning (ML) and Deep Learning (DL) to explain how the technology understands language and vision, adapts to novel environments, and makes decisions. Kate took to Reddit AMA to discuss her ground-breaking research, and bring to light the future of AI by myth-busting common AI misconceptions and answering questions within her areas of expertise. Check out the top seven takeaways from her discussion below.
1) AI technology struggles to dig deeper in identification functions
Kate explains how AI machines struggle to understand writing if it changes from its original appearance (for example, if the font or color changes).
2) The most exciting impact of AI research today is letting machines adapt to the real-world
The field of simulation-to-reality domain adaptation has been a highlight in Kate’s research.
3) Society is both overestimating and underestimating AI at the same time
According to Kate, people think that AI is “smarter than it is,” while also underestimating its effect on different information processing tasks.
4) AI “going insane” and its implications for society
Major problems could occur if a “bug” appears in AI code.
5) There are numerous exciting opportunities for women to enter the AI field in today’s world
“AI will change the world, Who will change AI?”
6) AI is learning to mimic human behavior
Customer service agent or chatbot? Can you tell the difference?
7) Machines do not yet understand the “meaning” of language
Kate uses deep learning to help machines interpret language as a set of numbers.
For additional commentary by Boston University experts, follow us on Twitter at @BUexperts. Follow Boston University Department of Computer Science on Twitter at @BUCompSci.
