After my recent article in which I discussed the future of work, and how AI technology will be used to disrupt once safe traditional roles, I felt that an article explaining the difference between Artificial Intelligence (AI) and Machine Learning (ML) was needed. Unfortunately, I realise that many people in the tech industry often use these hot buzzwords interchangeably. So here is my understanding of these terms and I hope it helps.
Artificial Intelligence. You've probably already seen AI being incorrectly referenced on social media and in the news. Then thinking about AI think of it as a broad set of different technologies in which a computer is able to answer a question without being programmed to do so.
Machine Learning is an application of AI which uses a large set of data and advanced statistical analysis in order for machines to determine the answer from previous similar question and answers.
The way I see it, AI is more of a vision, a direction of travel with only a very few organisations actually doing true AI. Unfortunately, you never see or hear of Machine Learning in the movies or on the news, so many people in the tech industry incorrectly saying that their products and services are using AI when in reality they're actually using some form of Machine Learning. Machine Learning is still a very impressive technology and has given us so much already. Take Google photos, for example, I love being able to find pictures of friends, locations and even activities all done thanks to Machine Learning.
There are very few real-world examples of AI, in fact, the only one I can think of was when Google used Machine Learning to translate words between multiple languages. After a while, the program created its own “interlingua” to translate more efficiently. For more information see https://techcrunch.com/2016/11/22/googles-ai-translation-tool-seems-to-have-invented-its-own-secret-internal-language/. If you're interested in learning more about Google’s Multilingual Neural Machine Translation System the research notes are available at https://arxiv.org/pdf/1611.04558v1.pdf
Hopefully, this makes things a little clear, and to my friends in marketing please stop staying that your products are powered by AI when they actually using Machine Learning.
Artificial Intelligence. You've probably already seen AI being incorrectly referenced on social media and in the news. Then thinking about AI think of it as a broad set of different technologies in which a computer is able to answer a question without being programmed to do so.
Machine Learning is an application of AI which uses a large set of data and advanced statistical analysis in order for machines to determine the answer from previous similar question and answers.
The way I see it, AI is more of a vision, a direction of travel with only a very few organisations actually doing true AI. Unfortunately, you never see or hear of Machine Learning in the movies or on the news, so many people in the tech industry incorrectly saying that their products and services are using AI when in reality they're actually using some form of Machine Learning. Machine Learning is still a very impressive technology and has given us so much already. Take Google photos, for example, I love being able to find pictures of friends, locations and even activities all done thanks to Machine Learning.
There are very few real-world examples of AI, in fact, the only one I can think of was when Google used Machine Learning to translate words between multiple languages. After a while, the program created its own “interlingua” to translate more efficiently. For more information see https://techcrunch.com/2016/11/22/googles-ai-translation-tool-seems-to-have-invented-its-own-secret-internal-language/. If you're interested in learning more about Google’s Multilingual Neural Machine Translation System the research notes are available at https://arxiv.org/pdf/1611.04558v1.pdf
Hopefully, this makes things a little clear, and to my friends in marketing please stop staying that your products are powered by AI when they actually using Machine Learning.
Comments
Post a Comment