Pandemic changes focus of machine learning projects
- Many enterprise IT leaders have said the pandemic has made them realize artificial intelligence (AI) and machine learning (ML) initiatives matter even more than they thought
- Half of those are planning to increase AI/ML spending going forward
Artificial intelligence (AI) and machine learning (ML) do not happen in a vacuum. Building an ML platform from scratch can take years and easily cost more than budgeted when examining the cost of staff, missed opportunities, and mistakes made along the way.
Amid the ongoing pandemic, it can be especially important to ensure existing operational processes are not limiting data science technologies that teams could be taking advantage of. It’s imperative that an organization’s operational approach to machine learning is reducing time to delivery while increasing the capacity of models in production.
According to a new survey by a leading provider of ML Operations & Management solutions, Algorithmia, companies are actually planning to increase their spending on AI/ML because of the ongoing pandemic.
The 2020 Enterprise Trends in the New Normal survey examines the economic impact the global pandemic is having on AI/ML initiatives. The survey was conducted in August and involved over 100 IT directors working in companies with at least US$1 billion in annual sales, and with more than 5,000 employees.
The survey went on to explain that despite the events of the past six months disrupting plans and tightening budgets, IT organizations aren’t halting AI/ML projects and are instead shifting their focus. Since the onset of the pandemic, organizations have refocused AI/ML efforts by up to 20% for projects that supported remote workers and 17% that impacted customer experience.
Projects that produce short-term cost savings or provide top-line insights about customer behavior are a top priority, as organizations seek a competitive advantage in a tough marketplace. Long-term process automation and back-office projects, meanwhile, have decreased in priority.
The IT industry is said to historically respond to economic disruption with both short-term cost-cutting and long-term technology adoption shifts. The current state of disruption is said to be no different, with AI and machine learning expected to grow in importance when compared to where it was in early 2020.
42% of IT leaders responding to Algorithmia’s survey stated that at least half of all their AI/ML projects had been impacted from a priority, staffing, or funding standpoint due to the ongoing pandemic, while 54% said their AI/ML projects were focused on financial analysis and consumer insight before the pandemic.
The recent survey released last week also suggests that while IT may lock-in on cost-cutting when times are tough, priorities over the long-term will lie in the adoption of new technologies that enhance and enable automation. Therefore, coming out of the pandemic, IT leaders expect 59% of projects to focus on cost optimization, and a further 58% to focus on customer experience.
Machine-learning and AI are considered ‘high promise’ technologies that have a lot of potential but take longer to realize. This can be attributed to the fact that urgent and immediate projects often crowd out longer-term efforts during a busy time.
The response to the pandemic by many organizations has highlighted the importance of AI/ML projects. The survey noted that 66% of organizations realized that AI/ML projects matter more than they thought before the pandemic, with nearly 23% believing it should have been their highest priority before and during the pandemic.
Following this, IT leaders now agree that AI/ML projects will be a priority going forward. 65% of survey respondents said that AI/ML projects were at or near the top of their priority list before the pandemic and that 33% of these applications are now higher on their list since the onset of the pandemic.
“When we come through the pandemic, the companies that will emerge the strongest will be those that invested in tools, people, and processes that enable them to scale delivery of AI and ML-based applications to production,” said CEO of Algorithmia, Diego Oppenheimer.
The data from the survey shows a clear projection for more spending by companies on AI/ML going forwards. 91% of survey respondents said they were already spending at least US$1,000,000 annually on AI/ML before the pandemic, and that 50% plan to spend more than that going forward.
A lack of in-house staff with AI and machine-learning skills has been the primary challenge before the pandemic, which has been further exasperated due to the increasing demand for people with AI/ML job skills. The most important AI/ML-related job skills coming out of the pandemic are set to be security, data management, and systems integration.
The well-known talent shortage around data science and machine learning has inevitably slowed many organizations amid COVID-19. AI/ML skillsets are highly sought-after and underscore the competitive search for talent to support complex projects. The risk of fraud in times of uncertainty has also highlighted the importance of security in all machine learning and IT projects. Despite this, a slew of AutoML tools has emerged that have the potential to help companies become more productive, without the need for specialized developers with incredibly rare skill sets.
The latest research to emerge stands to consolidate just how important AI and machine learning is to organizations. Amid the global COVID-19 pandemic and an uncertain future, business involvement in AI and machine learning seems to be even more plausible during a downturn.
As businesses begin to come out of the other side of the pandemic, organizations that emphasize AI and machine-learning stand to gain a clear competitive edge to those that do not.
“These past several months have been difficult for most companies, but we believe investments in AI/ML operations now will pay off for companies sooner than later. Despite the fact that we’re still dealing with the pandemic, CIOs should be encouraged by the results of our survey,” added Oppenheimer.
22 September 2023
22 September 2023
21 September 2023