We have covered the cloud and its constituent trends in the previous article and let us examine the trends of Generative Artificial Intelligence in 2024 here.
Generative AI
As per this McKinsey research, it is estimated that generative AI could add the equivalent of $2.6 trillion to $4.4 trillion of value annually. According to this Precedence Research paper, the worldwide market for generative AI will reach USD 118.06 billion by 2032 from the current USD 10.79 billion in 2022, evincing a CAGR of over 27% during this forecast period.
The last few years have seen immense advances in transformers, GANs (generative adversarial networks), VAEs (variational AutoEncoders), diffusional networks and neural radiance fields (NeRFs). 2023 also saw developments and the emergence of multimodal generative AI models. Generative AI is expected to be a major disrupter in 2024 from multiple perspectives – use cases and models, overall ecosystem developments and regulations. Let us examine each in some detail:
Models and use cases
Beyond search engines, image creation and editing, social media and “consumer areas”, 2022 and 2023 have witnessed the growth of transformers, diffusional networks and VAEs across enterprise use cases.
Spurred by the availability of large datasets, transformers and LLMs shall continue to enhance the digital customer experience and marketing functions. Automotive, BFSI, manufacturing, media, telecom, healthcare, utilities and several other industries shall vastly improve their initiatives around conversational AI, intelligent chatbots, virtual assistants, digital marketing campaigns and targeted personalised content creation. As per this Gartner research, around 30% of outbound marketing messages from large enterprises shall be fueled by generative AI by 2025.
Generative AI is augmenting and value-adding to existing use cases and business processes that are leveraging Intelligent Automation and AI powered data management. On the other hand, focus on green computing, responsible AI and ESG goals is necessitating the justification of these huge computational power requirements vis-à-vis envisaged business benefits.
In 2024, automotive, manufacturing, healthcare, facility management and other verticals that use commercial, enterprise and consumer IoT, shall leverage VAEs to crystallise statistical time series and insights from their vast data sets for better decision making and operational excellence. BFSI organisations shall also use VAEs to draw out useful insights and predictions from macro and micro economic, funds, stock indices and other datasets, thus improving process excellence, product development and decisioning across investment banking, wealth management and other functions.
In 2024 and beyond, companies shall also be combining the power of LLMs/ transformers along with VAE generated time series to improve customer life cycle management based on Internet of Behaviour (IoB) and automate timely delivery of personalised marketing content.
Manufacturing, automotive, aerospace and healthcare industries shall use diffusional network models to enhance imaging and analysis, protein engineering, 3D CAD design, 3D printing, prototyping and materials management functions. Diffusional models are also being harnessed along with LLMs for better accuracy and output quality.
Tools such as GitHub Copilot, Amazon CodeWhisperer and ChatGPT are enhancing productivity and efficiency for software code generation, testing and reviews.
2024 is expected to witness the proliferation of the industrial metaverse especially across healthcare, smart manufacturing, smart cities, geospatial and other segments. As per this Deloitte research, by 2030, revenue driven by the industrial metaverse shall touch US $100 billion, which is significantly higher than the consumer and enterprise metaverse. Besides AR/ VR and digital twins, industrial metaverse leaders are also considering application of the powerful image and rendering capabilities of NeRFs.
GANs are being harnessed for detection of intrusions and anomalies, especially in financial transactions, cybersecurity and medical imaging. It is also envisaged that usage of GANs shall expand in electronic media, movies and video entertainment, art, fashion and retail in a bigger way.
From a horizontal perspective, generative AI enabled synthetic data eliminates the compliance, privacy and data protection risks associated with real life sensitive personal data, especially benefitting design, R&D, quality control, testing and other functions across healthcare, FSI, automotive, fashion and apparel and other industries.
As mentioned before, generative AI models have synergies with and popularise other technologies such as 3D/ 4D printing, medical/ electronic/ mechanical design, protein generation, genomic sequencing, circular green manufacturing, digital twins, data management and the metaverse.
From an IT perspective, generative AI is also expected to bolster application modernisation, user experience, data management and testing processes as well.
Throughout the pandemic, AI has permeated the entire gamut of cyber security technologies right from ransomware and malware protection, intrusion detection, extended detection and response (XDR), insider risk management (IRM), cyber data lakes, cloud security access broker (CASB), secure access service edge (SASE), security information and event management (SIEM), security orchestration, automation and response (SOAR), VAPT observability, backup and recovery and others. 2024 shall witness further value-addition by generative AI in the cyber-physical, IoT and metaverse world through its ability to interpret gestures, expressions and other nuances and aggregate corresponding risks and threats.
A big challenge yet opportunity for Industry leaders is to identify solid business use cases for generative AI, and to accordingly restructure their technology, data and human capital footprints, stacks, workflows and architecture.
This EY study mentions that almost two-thirds of tech CEOs acknowledge that their companies must action Generative AI projects immediately to ensure their competitors do not get a strategic advantage.
Overall generative AI ecosystem developments: Companies are harnessing the power of generative AI through the cloud and open-source models. As mentioned in the cloud computing section, hyperscalers are incorporating the necessary infrastructure, architecture, and other pre-requisites to offer their customers the tools and necessary computational power to harness the power of generative AI for better business outcomes. Public cloud providers have been enhancing their large LLMs, AI chips, and other capabilities to build platforms, general and industry specific offerings for their customers.
A leading hyperscaler has recently enabled an Asian travel company to localise its partner and customer experience management covering communication, intelligent chatbot, and content functions in Southeast Asian languages. In 2024, the cloud provider’s generative AI ecosystem will also extend to partner onboarding and customer service functions.
This cloud hyperscaler has also recently launched healthcare and life sciences industry specific generative AI services.
Another public cloud provider has harnessed its generative AI capabilities to enable a leading sports media house to enhance its customer experience by enabling footage/ commentary-based archived automated and intelligent video search.
Several Big Tech companies have embedded generative AI features into their end user computing applications and tools as well. Besides rapid developments in the industrial Metaverse ecosystem, semiconductor and hardware companies are also aggressively working on AI PCs and pivoting AI processing and inferencing to the edge, thus immensely saving on computational power, time and costs. 2024 is expected to see significant developments in this area, and could be a compelling alternative to running large scale open market LLMs on costly back-end servers in the cloud.
Trust, Security and Regulations
2023 has seen a clear shift from AI to be an enabler of organisations that not only meet business KPIs, but also embracing sustainability and ESG goals, improving societal outcomes, enhancing responsible citizen indices, and lifting investor sentiment. This has resulted in an internal and external focus of enhancing responsibility, accountability, ethics, trust-worthiness and security quotients of AI.
The latter half of 2023, witnessed temporary bans, widespread discussions amongst the public, companies and governments around incorrect outputs, data and AI algorithmic biases, rising incidents of deep fakes, infringement violations and unscrupulous use of generative AI. Thus, necessitating the global need for incorporating AI aspects in regulatory and governance frameworks. Some global activists have demanded dedicated overseeing agencies and implementing ethical and responsible AI frameworks, and others have even lobbied for a halt in AI adoption, till these aspects are not addressed. There is also an increasing public demand to view and comprehend the effect of AI on green computing, emissions and other sustainability indices.
With 2023 witnessing the US Executive Order on regulating AI and crystallisation of the EU AI regulations, and considering elections are taking place in major countries throughout 2024, it is anticipated that 2024 shall see an accelerated thrust in the aspects of Trust, Risk and Security Management (TRiSM). In 2024, G7, the UN, India, Brazil and other countries shall be accelerating their processes of building, reviewing and seeking inputs on AI legislations, acts and regulations. There are accelerated global efforts elsewhere to also crystallise common frameworks, foundational models and policies.
Gartner has ranked AI TRiSM as the top strategic technology trend of 2024 in this recent report. And there is proactive thought to incorporate and integrate these capabilities in the heart of AI programmes and ethos. This is especially true considering easy accessibility to open-source generative AI capabilities, general public’s AI unfamiliarity and low literacy, increasing proportion of partner and third party AI ecosystems, misuse of and malicious attacks on AI systems, and rapid evolution of organisational, country and wider regulatory controls for AI ecosystems.
This detailed Gartner article mentions that by 2026, AI models from organisations that incorporate TRiSM solutions and capabilities, will result in a 50% increase in acceptance, adoption and achievement of KPIs and goals.
This recent World Economic Forum report highlights the 7 core tenets of ethical AI viz. Having a foundation of positive human impact, embedding inclusivity, ensuring accountability, improving accessibility, eliminating harmful and irresponsible usage, and fostering equitable outcomes.
Besides Gartner, AI TRiSM principles have also been incorporated in other regulatory frameworks such as National Institute of Standards and Technology (NIST), World Economic Forum (WEF)’s guidelines for responsible AI, and at Big Tech companies such as Microsoft, Google, Tesla and Meta.
As per this Allied Market Research report, the global AI TRiSM market is expected to rise to USD 7.4 billion by 2032 from a value of USD 1.7 billion in 2022, thus witnessing a CAGR of 16.20% during this forecast period.
Companies shall expect to score high AI TRiSM scores by:
- Embedding these principles at all stages in the AI lifecycle
- Possessing a robust Data Ethics culture
- Incorporating AI model interpretability and explainability
- Ensuring frequent Data and AI audits
- Having strong Cyber Resilience posture especially covering areas such as proactive detection, remediation and recovery from adversarial AI system attacks, backdoor AI model attacks, data poisoning, AI model thefts, and LLM prompt injections
- Complying to necessary privacy, data protection, and AI regulations
- Embedding risk taxonomy and management
- Having a culture of responsible and ethical AI usage with AI governance council across the extended enterprise
While there are efforts to democratise generative AI across the wider enterprise, it should be done within the ambit of TRiSM, responsible AI and strong governance.
2024 will continue to see rapid developments of AI regulations and legislations across the globe, as well as demand justification of AI computational power requirements amidst the focus on green computing and ESG goals.
To fully harness the power of generative AI, companies that imbibe the tools, processes and culture of TRiSM principles can improve quality, acceptance, compliance, cost effectiveness, reputation and resilience from their AI initiatives.
We will cover sustainability and green IT, cybersecurity, supply chain, risk and miscellaneous trends in the subsequent articles.
Wanda Parisien is a computing expert who navigates the vast landscape of hardware and software. With a focus on computer technology, software development, and industry trends, Wanda delivers informative content, tutorials, and analyses to keep readers updated on the latest in the world of computing.
