Anaconda has released its 2022 annual State of Data Science report, which reveals the widespread trends, opportunities, and perceived roadblocks facing the data science, machine learning (ML), and artificial intelligence (AI) industries. The global study targeted the open source community through three cohorts of academics, industry professionals and students.
While open source software was created by and for developers, today it is an integral part of commercial software development and the backbone for continuous enterprise innovation. 20% of respondents cited the speed of innovation and affordability of open source as the most valued benefits of its use.
When asked about the top threats to continued innovation and advancement within the open source community, respondents focused on several areas:
Concerns about open source security are growing
With the incidents that have troubled the industry over the last year, including the Log4j breach and the rise of Protestware, open source security remains at the forefront. As a result, 40% of the professionals surveyed said their companies have scaled back their use of open source software over the past year due to security concerns. Additionally, 31% of professionals said “vulnerabilities” are the number one challenge in the open source community today.
While most organizations use open source software, 54% of the 8% of respondents whose organizations don’t say the top reason is fear of potential vulnerabilities, compromises, or risks. This is a 13% increase from the 2021 report and confirms the escalated security awareness across the industry in 2022.
A lack of talent is a problem for organizations
Organizations trying to scale their data science efforts and accelerate technology advancement and adoption have had to weather the impact of the talent shortage. 90% of surveyed professionals said their organizations are concerned about the potential impact of a talent shortage, with 64% saying they are most concerned about their organization’s ability to attract and retain technical talent. 56% believe that insufficient data science talent or workforce is one of the biggest barriers to successful data science adoption in organizations.
“With data scientists consistently being cited as one of the best careers in America, the talent pool is sure to catch up with demand,” said Jessica Reeves, SVP of Operations at Anaconda. “Solutions that have proven successful in bridging this gap include upskilling an existing workforce and enabling stronger remote work options. Organizations should strengthen the tools and resources available for continuous learning, and academic institutions should close the skills gaps for students and turn them into strengths as they prepare to enter the workforce.”
Ethics, bias and regulation require more attention – especially in education
Ethical challenges in the fields of AI, ML and data collection have never been more in the public eye, but progress is still needed in this area. 75% of professionals surveyed think government should play a bigger role in boosting technology innovation and manufacturing, with 70% saying they would support more funding for STEM and technology-based schooling.
In contrast, only 19% of surveyed students are currently learning ethics in AI/ML/Data Science classes, and 32% of students have rarely or never been taught AI/ML/Data Science classes with bias. These findings underscore the need for educational institutions to adapt learning paths to reflect and prepare those entering the workforce and shaping the future of data science.
“It’s incredible what the community has achieved in just the last ten years. Many companies would not exist without the open source foundations they are built on today,” said Peter Wang, CEO of Anaconda. “But to successfully address these challenges and continue to innovate the future enterprise, we must continue to invest in the open source community and its infrastructure. I am confident about where the industry priorities lie and the next generation of talent is entering the workforce.”
