[DZone Research] AI for Personal vs. Professional Projects

This article is part of the Key Research Findings from the DZone Guide to Artificial Intelligence: Automating Decision-Making.

Introduction

As part of the research for our 2018 Guide to Artificial Intelligence, we surveyed 403 developers, data scientists, and technologists. From their responses, we've created a quick article on the way people are using AI in personal and professional projects. 

AI in Personal and Professional Projects

Given the complexity of AI, and the fact that large, enterprise-level companies like Google, Netflix, and Amazon all back AI and machine learning projects, one would expect the majority of developers who use AI to do so in the workplace. And, as 286 out of the 403 respondents to this survey work for organizations that employ over 100 people (i.e.enterprise-level organizations), one would, again, expect the data to slant toward AI for professional projects. But, in fact, 30% of respondents told us they have used AI in personal projects, whereas 20% have implemented it in professional projects. Something that might be skewing these numbers is that 55% of respondents have not implemented AI/ML in a project before. But of the 338 respondents who told us they have not implemented AI in their projects, 20% have at least toyed around with one or more of the following types of machine learning: supervised machine learning, regression, unsupervised machine learning, binary classification, and/or multi-class classification.

Machine Learning in Personal and Professional Projects

The five machine learning techniques listed above also turned out to be the most popular ways of working with machine learning among our respondents. 40% of survey takers reported having worked with supervised machine learning and 25% have used regression algorithms. The final three listed — unsupervised machine learning, binary classification, and multi-class classification — all received 18% of the vote in this category. Interestingly, the percentages of respondents who've worked with these types of machine learning both at home and the office are more or less identical. For example, 30% of respondents have used supervised machine learning in systems developed for work, whereas 33% have used supervised machine learning for personal projects. The only real deviation seems to be in the use of multi-class classification. Whereas 17% of those who use AI/machine learning at work have used multi-class classification, 13% have used this type of machine learning for personal projects. While this is not a large difference, it does, nonetheless, represent the biggest percentage gap in types of machine learning used on personal vs. professional projects.

Popular Use Cases and AI Adoption

Looking at how all respondents use AI and machine learning, regardless of personal or professional projects, the largest use case by far is for prediction. 43% of those who took the survey told us that they are using machine learning for its abilities with predictive analytics. The second most popular purpose for machine learning was classification, totaling 31% of respondents. While not nearly as popular as prediction and classification, some other popular machine learning use cases were recommendations (23%), detection (23%), and optimization (21%). 

Given that 55% of respondents told us they have never implemented AI or machine learning in the systems they've developed, let's quickly explore why some developers are hesitant to adopt AI. When we asked, "What issues prevent you from being interested in AI/machine learning?", 50% of the 221 respondents to this question answered that there are not enough use cases for applications built on this technology. Another 38% said there's not enough developer experience in the area, and 26% said time played a factor.

Conclusion

So, since this is titled as a versus article, who wins: AI for personal development or AI for professional development? Right now, given the limited number of jobs in AI development compared to other areas of software development, it seems AI for personal projects comes out on top. Working with various tools like TensorFlow allows developers to create interesting software from the comfort of their couch, and learn the skills they'll need to land a gig at a place like Google, Netflix, or IBM— i.e., organizations that are changing the nature of AI and ML as we know it. 

This article is part of the Key Research Findings from the DZone Guide to Artificial Intelligence: Automating Decision-Making.