2021 is set to be an exciting year for the world of AI. The market is continuing to mature and 2021 will be no different. Whilst mass-market adoption has not yet been achieved, significant progress has been made over the past twelve months.
The latest insights from Omdia found more than 50% of respondents have deployed or are planning to deploy AI initiatives across 5 of 7 business unit categories within the next 12 months. Additionally, despite the ongoing COVID-19 crisis, confidence in AI stays high, with 71% of respondents indicating they are “confident” or “very confident” that AI will deliver positive results in the next 12–24 months. As the AI market continues to grow throughout 2021, we predict there will be some key trends that will have a significant impact this year.
Organizations will be looking at how they can gain competitive advantages through leveraging quantum computing to move beyond the limitations that traditional architecture presents; and increasing investment in AI-driven automation. Whilst also ensuring they converge Subject Matter Experts with Data Scientists and Engineers throughout various levels of the process.
Meanwhile, to move the market forward successfully, Omdia predicts there will some key areas that will need to be addressed across the next 12 months. These will be, an industry wide standardization of AI measurement and success; implementing standardized regulation and governance; and a greater drive to improve diversity within the sector.
Jenalea Howell, Market Lead - Al & IOT
Navigate directly to any of the trends using the links on this page or using the arrows. on the right side of the page. If you wish to download the report, you can do so via the Contents menu in the top left corner.
AI technology is increasingly being leveraged as a competitive, geopolitical, or economic advantage. As technological innovation swiftly progresses, new business models and innovative disruptive strategies must keep pace.
The rate of AI production and implementation will be accelerated due to COVID-19. Omdia expects the pandemic to boost specific sector revenue growth between 2019-26, with healthcare seeing a 46% increase, advertising a 44% increase, telecom a 40% increase, and education a 40% increase.
The need for ethical policy is vital to ensure a fair playing field. As per a 2020 Omdia survey, 62% of enterprises are either concerned or extremely concerned about the growing accountability gap in AI, which poses the question around who is responsible when something goes wrong.
Competition for talent has never been so strong.
Learn more about our 2021 AI offerings
Although there is a rise in AI adoption across the enterprise and across use cases, the measure of success for AI varies widely among applications and verticals. For AI to realize its full operational potential, corporations and investors will need to pay careful attention to defining the right KPIs and metrics.
Greater scrutiny of AI performance metrics will occur. As per Omdia’s AI Business Performance Metrics Database, values reported vary by more than 90% points in KPIs such as cost reduction, engagement and time reduction.
Enterprises will drive vendors towards standardization of AI metrics. As vendor solutions will be increasingly scrutinized for proven effectiveness, vendor communities will have to lead to codify AI business metrics specific to the function areas, use case or industries they serve.
Measuring AI effectiveness will become part of best practice, standardized around horizontal or industry-specific functions. Lack of universal metrics will slow the ability of enterprises to measure its effectiveness, but since AI will increasingly be broadly applied, it is logical that function, use case or industry-specific metrics will be adapted.
The rise of AI end-to-end solutions is emerging. Having an ecosystem-wide strategy will become increasingly critical as innovation arises throughout the the value chain, from new silicon types and start-up applications, to data center management.
AI chip startups will battle for supremacy. As per Omdia research, more than 100 AI chipset initiatives are on offer from a wide range of companies with different stakes in the market and more than 50% of them being startups.
Software-based acceleration options will increase. Omdia believes that the popularity of AI will lead to SoC accelerators (IP cores) becoming commonplace and being adapted into almost every chipset over the long term. The market for SoC accelerators is expected to increase to $3.3bn by 2025.
AI for data center management will go mainstream and Omdia predicts the AI chipset market for enterprise cloud/datacenter is estimated to grow from $3.8bn in 2019, to $19.4bn by 2025.
AI's manifestations will evolve and intersect with other technology macrotrends, mutually influencing the trajectory and adoption of each other. Both edge and quantum activity are beginning to push AI's boundaries and possibilities.
Hybrid quantum-classical models combine the power of quantum computing and AI-based algorithms. The intersection of AI and quantum is likely to occur in areas like life sciences for drug discovery or in finance for credit risk scoring or trade optimization.
Applications of quantum computing are simulation, optimization and AI/machine learning.
Quantum-based AI is dependent on the algorithms you develop; problems to be solved need to be identified before developing quantum based algorithms.
Today’s ecosystem of AI governance and ethics lacks clear leadership. Voluntary, piecemeal solutions for ethical AI uses are emerging in corporate responsibility models, for topics such as data privacy, usage and anonymization. For AI to truly permeate the enterprise, the global tech industry must set aside individual, parochial views and desires for short-term gains and work with governments to forge workable, consumer-acceptable regulation. As per a 2020 Omdia survey, 60% of enterprises feel data privacy is "slowing" or "significantly slowing" their AI initiatives, with 65% asking for regulation.
Determining culpability when AI goes wrong will lead to greater regulation, with the U.S. - the world's leading litigious society - paving the way.
Tech giants, such as Google and Amazon, will be less inclined to work with smaller industry players to form a unified, cooperative stance with governments on AI regulations, preferring instead to go their own way.
Key 2021 initiatives to watch will be: EU’s framework on ethical rules for AI, Japan’s Social Principles of Human-Centric AI, U.S.’s AI Accountability Act.
As AI investments drive start-ups and innovation across the globe, responsible innovation will be critical to the industry's future. The concept of "AI for Good" and "AI for All" will become a benchmark for how society uses AI during crises. This industry must tackle the startling lack of diversity, and the danger this poses, as AI innovation evolves, including both conscious and unconscious biases in data and algorithms shaping future solutions.
As governments and workers try to operate in the new normal that COVID-19 brings, "AI for Good" is going to play a very important role.
The four principles of responsible AI are inclusivity, transparency, accountability and privacy.
The startling lack of diversity in this industry must be addressed this year.
Experimenting with AI proofs of concept is all very well, but business applications will falter unless AI's power is democratized and accessible to regular business users, rather than to data scientists and engineers alone. The convergence of data scientists, engineers and subject matter experts (SMEs) across the organization will be crucial to realize its operational benefits and to ensure accountability, responsibility, and ethics are baked into the very heart of model build and deployment.
Increased value will be placed on data science and engineering talent who can “switch codes” between their language and that of the boardroom. As per Omdia, building a business case for AI is amongst the top 3 requirements of AI practitioners within the enterprise, suggesting the importance of SMEs and the need for projects to be linked to business outcomes.
Explosion of adoption of AutoML (and similar) technologies to involve more SMEs directly in model build, performance, suitability, and accountability.
The standardization and affordability of the tools required to automate machine learning deployment at scale will increase.
AI is redefining rules in many labor markets, potentially creating a loss of voice for unions and shifting support back to the enterprise, while increasing company resiliency. Although AI technology will create labor shifts, the larger trend promises to identify and upskill talent.
Disruptions throughout 2020 demonstrated the importance in 2021 of digital transformation as a foundation for AI outcomes.
Intelligent automation within RPA will serve as a framework for smartly automating front-office tasks and empowering smarter decisions. As per Omdia ICT Enterprise Insight Survey 2020/21, more than 50% of enterprises are trialing or have fully deployed intelligent automation tools.
Investment in AI-driven automation and augmentation will improve corporate KPIs and employee retention by allowing employees to undertake higher complex tasks and make more impactful business decisions. Additionally, having a human in the loop helps continuously train the software, thus making the system smarter and generating further time savings for the employee.