30% AI models will incorporate multiple modalities by 2026

IDC predicts that, By 2026, 30% of AI models will incorporate multiple modalities of data to improve learning effectiveness and address the current everyday knowledge shortcomings in single-modality AI solutions. This is just one of the predictions unveiled in the report titled, IDC FutureScape : Worldwide Artificial Intelligence and Automation 2023 Predictions — Asia/Pacific (Excluding Japan) Implications.

Since the onset of the pandemic, the changing market dynamics are pushing APAC organizations to take a disciplined approach to scale AI initiatives and generate improved business impact. For AI to reach its full potential throughout the region, sharper business use cases and a more robust data ecosystem are required. Multimodal AI, a new AI paradigm, bridges the gap between humans and computer intelligence by combining various data types such as images, texts, speech, and numerical data with multiple intelligence processing algorithms and achieves higher performances.

“With AI systems now starting to leverage multiple data sources, multimodal AI outperforms single-modal AI to deliver more accurate results across a wide range of applications. Organizations are looking to maximize returns on existing technology investments, as opposed to investing in large transformation programs. AI will be leveraged for increased employee productivity, optimized supply chains, and enhanced customer experience,” according to Deepika Giri, Associate Vice President, Big Data & AI, IDC Asia/Pacific including Japan (APJ) Research.

This study presents the top 10 predictions for AI and automation initiatives through 2028. Each prediction is assessed based on its impact (a mix of cost and complexity to address) and time frame to the expected stated adoption level. This study also highlights IT impact and guidance for technology buyers for each prediction statement.

The following 10 predictions represent the expected trends with potential impact on AI and automation initiatives:

#1: Talent Gap: Persistent talent gaps will drive 55% of IT organizations to invest in AI skills by 2023, both to run automated IT operations and to support business end users adopting AI and automation solutions.

#2: Embedded AI: By 2026, AI-driven features will be embedded across business technology categories, and 65% of organizations will actively use such features to drive better outcomes without relying on technical AI talent.

#3: AI for Risk Management: As AI models become strategic, model risk and governance will consume more business attention until 2026, then 60% of Asia-based 2000 organizations (A2000) CFOs will incorporate AI risks as part of their enterprise risk programs.

#4: Low-Code/No-Code Adoption Accelerates: By 2023, most organizations will leverage codeless development tools for at least 30% of their AI and automation initiatives, helping to scale digital transformation (DX) and democratize AI.

#5: Foundation Models: By 2026, massive (>1 trillion parameter) foundation models (for natural language processing [NLP], AI-generated images, etc.) will become standard industry utilities provided only by the largest vendors.

#6: Multimodal AI: By 2026, 30% of AI models will incorporate multiple modalities of data to improve learning effectiveness and address the current everyday knowledge shortcomings in single-modality AI solutions.

#7: AI/ML as the Big Flip: Driven by the componentization of applications and by advancements in AI/machine learning (ML), 25% of A2000 companies will have AI-assembled applications by 2027, disrupting traditional developer roles.

#8: Multimodal Collaborative Automation Platforms: To speed automation development and maximize benefit, by 2025, 30% of organizations will adopt intelligent business execution strategies, turbocharging all aspects of the automation life cycle with AI.

#9: AI-infused Operations: By 2026, 75% of large Asia/Pacific enterprises will rely on AI-infused processes to enhance asset efficiency, streamline supply chains, and improve product quality across diverse and distributed environments.

#10: Sustainable AI: In response to sustainability and economic uncertainty concerns, by 2024, 25% of the A2000 organizations will adopt tooling to quantify, predict, and optimize the cost-benefit of their AI life cycle.

CT Bureau