Artificial Intelligence (AI): 7 trends to watch for in 2022 – The Enterprisers Project
Of the many technologies with the potential to deliver significant value in the near future, Artificial Intelligence (AI) seems firmly planted atop the list for CIOs. Indeed, nearly all (95 percent) of the CIOs, CTOs, and technology leaders surveyed by IEEE agreed that AI will drive the majority of innovation across almost every industry sector in the next one to five years.
2022 will be a year in which AI will mature from experimental to essential. “The focus will shift more toward AI-enabled transformation that solves more significant business problems with business-focused solutions,” says Jerry Kurtz, executive VP, Insights & Data, at Capgemini Americas. “AI is an enabler and powerful capability, but the time for proofs of concept and science projects is quickly coming to an end. In 2022, expect AI engagements to become larger, more strategically significant, and more mission-critical – with a focus on long-term scalability.”
That will challenge most CIOs. Outside of the technology powerhouses, many enterprise IT organizations are relative newbies when it comes to AI. “AI adoption continues to be gain momentum but is still at infancy,” says Yugal Joshi, a partner at Everest Group. “One of the key challenges before CIOs is to ensure they are investing in the right use cases that give the maximum ROI, especially because AI applicability is quite broad.”
[ Check out our primer on 10 key artificial intelligence terms for IT and business leaders: Cheat sheet: AI glossary. ]
AI trends to watch
Against that backdrop, there are a number of more developing trends in AI that IT leaders will be wise to keep an eye on this year.
1. Data wrangling tops the agenda
Most enterprises are relatively early on in their AI journeys. Unlike the Googles or Facebooks of the world, they spend the majority of their time and resources wrangling data. They need to build modern data pipelines.
[ How can public data sets help your AI work? Read also: 6 misconceptions about AIOps, explained. ]
“Most AI models are hungry for massive amounts of data, and organizations need to build flexible data pipelines that can evolve to support thousands of sources, incorporate structured and unstructured data, and provide it to data scientists in a meaningful and reliable way,” says Erik Brown, a senior partner in technology at West Monroe. “Traditional ETL (extract, transform, and load) and relational stores must be complemented for more scalable data lakes, and in many cases data streams must be provided to be processed in real-time.”
2. Automated process discovery boosts RPA efforts
The future will be streamlined. Enterprise leaders can visualize their organization’s automation potential using new process discovery technologies. “Though not purely focused on automation opportunities alone, these technologies will provide process related insight not gained by any other means,” says Wayne Butterfield, director at ISG. Process mining, task mining, and the up-and-coming conversation mining enable “lean on steroids,” Butterfield says, giving the enterprise more autonomous means of developing a robotic process automation (RPA) pipeline. “These technologies will really come to the fore in 2022 and amplify the usage of Intelligent automation in the process.”
[ Read also: 4 Robotic Process Automation (RPA) trends to watch in 2022. ]
3. AI enables effective supply chains
Intelligent supply chain applications should become the rule rather than the exception going forward. “From supply and demand planning to digital manufacturing and logistics, supply chains in 2022 will need to be continuously transformed, AI-enabled, and most importantly given all the recent disruption, future-proofed,” says Kurtz of Capgemini Americas.
4. Customer-facing AI plows ahead
“The pandemic saw AI adoption in customer-facing roles such as virtual agents increase,” says Joshi of Everest Group. “This will continue but with more maturity and complex use cases.”
5. Natural Language Generation (NLG) goes mainstream
OpenAI recently made its GPT-3 large language model, already being used by hundreds of apps, available by API. The most public example of the power of NLG, GPT-3 can be used in applications that require a deep understanding of language, from converting natural language into software code to generating answers to questions.
“NLG, which historically focused on turning numerical data into text based insight, is now generating text from text-based data points and starting to change the game in creative writing,” says Butterfield of ISG. “The possibilities are endless, with GPT 3 also being used to create unique training data sets for NLP, real-time generation of unique responses in conversational AI platforms, and the capability is even being used to generate software code. 2022 should see even more uses of NLG, and really catapult it to the masses.”
6. Talent shortages threaten progress
Given how rapidly the AI landscape advances, effective talent management has become a strategic differentiator for enterprise IT organizations. “This will need to consist of world-class recruiting and retention-related initiatives that promote inclusivity and a lifelong learning culture,” says Kurtz of Capgemini Americas. “The market has never been more competitive for people with AI skills, and this trend is likely to continue for years to come. As such, strategic partnerships will also be key across organizations and industries.”
7. AI transforms IT productivity
The increasingly complex and powerful IT environment of the future is too much for human tech pros to manage alone. “Another area of increased AI adoption will be managing modern systems that CIOs are building,” says Joshi of Everest Group. “These systems cannot be just managed by humans. The observability, intervention, and deep analysis needed for these systems will be AI-enabled.” Joshi is looking for real-time, actionable interventions.
Expect help on the app dev side of the house too, given the growing excitement around generative AI. “In 2022, CIOs will also start to evaluate AI applicability into the engineering organization for fundamentally transforming developer’s productivity,” Joshi says. This area has been researched for a long time, he notes, but meaningful progress has been made recently.
[ Where is your team’s digital transformation work stalling? Get the eBook: What’s slowing down your Digital Transformation? 8 questions to ask. ]