Large language models (LLMs) have permeated every corner of the digital landscape, handling a myriad of tasks. While they instill unease in some, the future holds a transformation where they will evolve into cloud-based “Generative-as-a-Service” (GaaS) offerings, following the path of other “as-a-service” products and solutions.
Prominent cloud providers, including Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform (GCP), Alibaba Cloud, Oracle Cloud, IBM Cloud (Kyndryl), Tencent Cloud, OVHcloud, DigitalOcean, and Linode (owned by Akamai), will either build, collaborate, or acquire generative AI capabilities to incorporate into their service portfolios.
These tools will give rise to thriving ecosystems, akin to those surrounding the essential enterprise infrastructure and applications that drive companies worldwide. Notably, Google, AWS, IBM, and Microsoft are fiercely competing in the generative AI (GAI) race, with Microsoft currently leading the charge.”
Let’s approach Large Language Models (LLMs) with the same perspective we apply to ERP, CRM, or DBMS (an acronym that may have fallen out of common use). We’ll explore how companies decide which tool to employ, how to utilize them effectively, and how to apply them to address real-world challenges.
Have We Arrived?
No, we are not currently there, but we will undoubtedly reach that point in the near future. The process of productizing Large Language Models (LLMs) and generative AI (GAI) is already in full swing. Initially, granting access to premium or business accounts marks the first phase. As the initial wave of LLMs settles between 2022 and 2023, a competitive race will emerge, driven by both capabilities and cost-effectiveness.
Companies will base their decisions on documented use cases supported by ROI (Return on Investment), OKR (Objectives and Key Results), KPI (Key Performance Indicators), and CMM (Capability Maturity Model). These well-documented use cases will extend across various crucial functions and industries.
To understand the potential of GAI and its exploitation, companies eager to adopt it will refer to these metrics and use cases to conduct internal due diligence. Once this step is completed and promising prospects emerge, the next actionable steps will be taken.
Getting Ready for What’s Bound to Happen
Once Large Language Models (LLMs) are fully productized and eventually commoditized, the next crucial step is to ascertain and outline the specific business functions and processes that could derive substantial benefits from Generative AI (GAI). However, this is not a straightforward task, mainly because many companies lack comprehensive process inventories, and they haven’t thoroughly examined their processes to identify the aspects most compatible with GAI implementation.
Additionally, the complexity arises from the fact that entire business models, consisting of numerous processes, could be potential targets for GAI integration. For instance, let’s consider marketing:
“Marketing professionals within corporate departments aim to capture the attention of key target audiences through strategic advertising. Promotional efforts are tailored to specific demographics and may involve celebrity endorsements, compelling phrases or slogans, unforgettable packaging or graphic designs, and extensive media exposure.”
Marketing encompasses all the actions undertaken by a company to promote and sell its products or services to consumers. These efforts make use of the well-known ‘marketing mix,’ often referred to as the four Ps – product, price, place, and promotion.
Initially, marketing primarily revolved around traditional techniques such as television, radio, mail, and word-of-mouth strategies. However, with the advent of digital advancements, the landscape has evolved, and companies now employ digital marketing approaches, including newsletters, social media, affiliate programs, and content marketing.
At its essence, marketing aims to take a product or service, identify its ideal target audience, and capture the attention of these customers, making them aware of the offerings available.
Among these processes, which ones can profit from Generative AI (GAI)? Can GAI effectively create marketing campaigns, draft press releases, and target customers? The answer is affirmative, but the critical question remains: can it deliver the same level of quality as human marketing professionals?
What processes does HR entail?
Human Resources (HR) serves as the business segment tasked with locating, hiring, evaluating, and providing training to potential job candidates. Additionally, HR departments oversee matters relating to employee compensation, benefits, and terminations.
Human Resource Management (HRM) strategies center on proactively enhancing and refining an organization’s workforce, with the ultimate objective of elevating the overall performance of the organization.
To fulfill their responsibilities effectively, HR departments must remain well-informed about the latest laws and regulations that may impact both the company and its employees.
In recent times, many companies have opted to outsource traditional HR administrative duties, such as payroll and benefits, to external vendors.
How will Generative AI (GAI) impact and revolutionize HR processes? Over time, this transformation is likely to occur relatively seamlessly. (And let’s not overlook the fact that traditional machine learning can also enhance, replace, automate, or reimagine numerous processes.)
This evolution will unfold on a global scale. The process of aligning GAI with existing technologies will occupy the attention of CIOs, CTOs, CEOs, CMOs, CFOs, and other key executives indefinitely, as it will also involve integration with other emerging technologies.
Additionally, companies must closely monitor GAI’s progress, including the initiatives their competitors are undertaking with this technology. As GAI’s capabilities surpass the bounds of Moore’s Law, businesses may contemplate establishing Task Forces or even Centers of Excellence to track and harness its potential applications. The potential need for Chief AI Officers might arise, provided there is room for yet another chief-level position in the organizational structure.
Cloud Services and Ecosystems
All major cloud providers will facilitate the aforementioned advancements. Moreover, there will be a well-defined Generative AI (GAI) ecosystem, characterized as:
“…a network of interconnected digital technologies, platforms, and services that collaborate to deliver value to both businesses and consumers.”
These ecosystems encompass a diverse range of vendors continuously enhancing their products and services, while newcomers strive to disrupt established players.
Consultants will emerge to provide guidance on multi-LLM management, LLM security, efficient engineering practices, and methods to prevent or identify LMM “hallucination.” Undoubtedly, consultancies will also introduce various other GAI services for commercial purposes.
As with any burgeoning technology ecosystem, talent shortages will pose challenges. Universities will attempt to address this by developing new courses and degree programs in GAI, but they might struggle to keep pace with the rapidly evolving field. Curriculum committees may find it difficult to move swiftly enough to meet the demands of this dynamic landscape.
Is it business as usual?
Irrespective of any temporary halts, the progress of Generative AI (GAI) will undoubtedly continue. It will commence as a novel technology, expand its presence in the cloud (with the usual prominent cloud providers), and eventually take its rightful place alongside other major technologies that have significantly contributed to the digital landscape. Much like ERP, CRM, and DBMS applications, GAI will establish its own ecosystem.
As the number of use cases for GAI multiplies, the pace of venture investments will eventually slow down. However, GAI sets itself apart from other disruptive technological leaps, such as cars, planes, submarines, and Uber. Its distinctiveness stems from the vast scope of potential applications and its ability to evolve organically without scheduled maintenance requirements. Furthermore, GAI offers unique opportunities (or threats, depending on one’s perspective), primarily within knowledge-based domains.
Considering that a substantial percentage of the workforce comprises fully remote and hybrid knowledge workers, the relevance of GAI becomes even more significant. As platforms like LLaMa, ChatGPT, Bard, Co-Pilot, and others continue to evolve, large language models will transform into monumental cloud services featuring extensive ecosystems. It is crucial to identify processes where GAI can contribute, as its capabilities extend far beyond mere assistance.
Conclusion
Large language models (LLMs) have become ubiquitous, undertaking diverse tasks, and will evolve into cloud-based Generative-as-a-Service (GaaS) offerings alongside major technologies. Cloud providers will play a significant role, fostering an ecosystem of interconnected technologies. Despite uncertainties, GAI’s potential is vast, disrupting traditional approaches. Businesses must closely track GAI’s progress and explore its applications beyond assistance.
