In what ways do AI techniques propel decision-making amidst volatility? Annotated bibliography perspectives

Abstract

This research presents a systematic review and approximation, from 2018 to 2023, of how Artificial Intelligence can support decision-making processes when business managers have to resolve between multiple alternatives for the development of new businesses using Agile Frameworks; and as well, will be addressed from the genesis concepts such as VUCA, Agile Mindset, Infinite Game Theory, Agile Frameworks, Innovation, Artificial Intelligence among others with the aim of identifying research gaps in the decision-making process under changing environments. Most of the analyzed studies are focused on Infinity Games theory to better understand innovation processes linked from early stages to deployment phases of products. Furthermore, this paper analyzes AI techniques for decision-making under volatile circumstances and raises 5 research questions that from a logical and chronological perspective in development are resolved during the research. It is important to highlight that AI-related solutions are typically used to make informed decisions; however, few studies adopt AI to support the decision-making process in unstable environments.

Introduction

A plethora of authors have made significant contributions to the development of various novel approaches for generating new business models from ideas. Many of these authors converge on several key aspects, with one of the most noteworthy being the recognition that the markets operate within a VUCA (Volatile, Uncertain, Complex, and Ambiguous) environment, making it impractical to determine conditions in the medium or long term due to their high variability (Johnson & Smith, 2018).

Consequently, the discipline of innovation has gained prominence as the pathway for organizations to create new opportunities, products or services and sustain themselves in increasingly aggressive markets, where startups emerge rapidly and abundantly (Li & Brown, 2018). In this context, various researchers have contributed to the development of agile methodologies, which fundamentally embrace the “fail fast and learn quickly” mindset and also involve continuous experimentation with small hypotheses to facilitate rapid learning about the ideas under investigation (Kim & Lee, 2018; Sommer, 2019). By embracing these agile principles, organizations can swiftly adapt to changing market conditions, iterate on their business models, and seize emerging opportunities (Johnson & Smith, 2018). These adaptive and non-predictive approaches allow solution research teams to iteratively learn through experimentation processes, preparing personnel to navigate innovation challenges in uncertain environments.

Moreover, Sinek (2019) introduced a novel concept based on Game Theory, termed as The Infinity Games. This theory distinguishes between two kinds of games: Finite and Infinite Games. Finite games involve known players with fixed rules and agreed-upon objectives, where players compete to win. On the other hand, Infinite Games have an entirely different nature, where the primary objective is to keep the game going and perpetuate the play. This idea emphasizes the importance of adopting an Infinite Game mindset in the context of innovation, as it encourages organizations to focus on continuous improvement and adaptation rather than solely on short-term victories.

Nonetheless, the latter theory presents distinct features, including: (I) a mix of known (finite players) and unknown players (infinite players), (II) changeable rules, and (III) an objective to perpetuate the game. When finite players face each other, the system remains stable as they operate within the same environment. Similarly, when infinite players compete, the system is also stable as they engage in the same ongoing process, with no winners or losers (Sinek, 2019).

However, challenges arise when a finite player encounters an infinite player. The finite player seeks to win, while the infinite player's objective is to continue playing. Consequently, the finite player often finds themselves in a quagmire, struggling to keep up with the perpetual nature of the game.

Relating this concept to the business context, Sinek (2009, 2019) asserts that the game of business is an infinite game. This is because the players, including competitors, may be both known and unknown within the industry. The rules of the business game are changeable and not universally agreed upon, adding to the dynamic nature of the environment. Whereas, in finite games, technically there is not ultimate winning in the game of business; instead, the objective is to keep playing and adapt continuously to the ever-evolving landscape.

As suggested by Reim et al., (2020, p. 180), there is an academic gap related to innovative business models in conjunction with the introduction of Artificial Intelligence which could be translated into “increased risk of project failure and unwanted results”. To bridge this gap, the purpose of this paper is to provide a systematic review from 2018 to 2023 with the aim of investigating how the combination of Agile methodologies and Artificial Intelligence can help organizations to work in a process of Continuous Innovation (Development of Sustainable and Sustainable Businesses).

Also, it is important to highlight that this paper focuses on how the decision-making process is conducted in unstable conditions; in other words, it focuses on how non-binary logic can further support organizations and their resilience in adapting to changing environments and adopting both Agile principles and the Infinity Games theory alike. This research seeks to provide a comprehensive comparison between traditional approaches and AI-led solutions with regard to the decision-making process.

Methodology

The research is centered around presenting a literature review of traditional and ad hoc practices related to the decision-making process through the use of ICT tools within the Project Management domain. To gather the necessary information for this study, a systematic review process requires a sound and clear method, this paper follows the method proposed by Kitchenham and Brereton (2013) which consists of the following phases: (a) elaborating hypotheses or questions; (b) searching for academic sources; (c) inclusion and exclusion criteria; (d) extraction of information; (e) interpretation and presentation of the results; (f) discussion. phases from (a) to (d) are described next in this section, while (e) and (f) in sections Results and Discussion, respectively.

• a. Elaborating research questions

After extensive research into the common attributes of great leaders in inspiring action and commitment towards a shared purpose, Sinek (2019) determined that the key to their success lies in their communication style. He also unveiled the secret to eliciting commitment from everyone, demonstrating how these leaders effectively motivate people by communicating from the inside-out, starting with the "Why" and extending to the "What".

The main question that arises is what it truly means to "begin with the Why". For any organization or individual seeking success in any domain, it is essential to understand the underlying connections and shared purposes that unite them with others. This connection or common purpose serves as the foundation of the message that every leader must construct to communicate effectively with stakeholders. The key distinction of this communication style, the Golden Circle, lies in its emphasis on Emotional Engagement, establishing a strong connection between the customer and the organization's purposes. This emotional engagement serves as the cornerstone of success in any event, business, or entrepreneurial venture (Sinek, 2019).

Under these circumstances, the research team proposed the following research questions:

RQ1: How can the Infinite Games theory, formulated by Sinek (2019), connect to the changing and volatile attributes of organizations operating in unstable conditions?

RQ2: How can Agile methodologies be utilized as effective tools to design and adapt new business models in response to the challenges posed by a VUCA environment?

RQ3: How does the integration of AI support decision-making processes for investments in innovative business model development?

RQ4: What are the critical factors that organizations must consider while transitioning from Finite Games (characterized by known players, fixed rules, and agreed-upon objectives) to Infinite Games, as proposed by Sinek (2019)?

RQ5: How can organizations effectively balance and integrate the concepts of Uncertainty, Agile methodologies, AI, and Innovation to foster a culture of continuous learning, experimentation, and adaptation?

• b. Searching for academic sources

During this stage, bibliographic sources were identified from top-tier databases such as: Scopus, IEEE, Springer, Emerald, ACM and Google Scholar. This bibliographic search was carried out in August 2023, and basically articles selected were from January 2018 to the first quarter of 2023. Also, technical descriptors used as search strings were as follows: [Business Model Development AND (Artificial Intelligence OR Agile Methodologies) AND Uncertainty OR Innovation]; [Business Model Development AND (Artificial Intelligence OR Uncertainty OR Design Thinking OR Scrum OR Lean) AND Agile Methodologies].

• c. Inclusion and exclusion criteria

Considering that the nature of this study is multidisciplinary, the team considers relevant to retrieve scientific papers from influential databases. In addition, this stage follows a hybrid approach adopting some practices such as "study identification, screening, eligibility and inclusion" (Moher et al., 2009) and bibliometric analysis techniques (Echchakoui, 2020) (Table 1).

Table 1 Search results and selection criteria

During this process, it is important to clarify that for the purpose of this paper, articles, book chapters and conference proceedings are included. In addition, articles that do not provide any particular focus related to this research were removed. Only articles that provide relevant data based on these two search strings are considered. Since the search resulted in a considerable number of articles, it was decided to apply exclusion criteria, remaining 381 articles. Of this amount, the selected articles are those that study the application of hybrid project management frameworks in conjunction with the application of various artificial intelligence techniques to elaborate or improve the decision-making process in changing business models. Furthermore, another filtering process was carried out by discarding studies that do not have the following characteristics: range from 2018 to 2023, project management, project governance, decisions, and uncertainty. The result of both filtering processes allowed the identification of 43 articles detailed in Table 2 and Fig. 1.

Table 2 Top results by search string

Fig. 1

PRISM graphic for systematic review

• d. Extraction of information

To answer each of the research questions, Table 4 was developed which contains a summary including the following characteristics:

• Type of application: risk management (R), project governance (G), decision-making process (D), business model (B).

• Research approach: quantitative (QN), qualitative (QL), mixed (MX).

• Applied techniques: fuzzy algorithms (FZ), machine learning (ML), neural networks (NN), stochastic algorithms (ST), decision trees (DT).

• Industry or domain: hardware (HW), software (SW), education (ED), logistics (LG), cybersecurity (CY), commerce (CO), hydrocarbons (HY), construction (CT), services (SE), healthcare (HC).

• Country where the study was developed.

Research question	References	Applied techniques	Industry or domain
RQ1: How can the Infinite Games theory, formulated by Sinek (2019), connect to the changing and volatile attributes of organizations operating in unstable conditions?	Humlung and Haddara (2019), Ingvarsson et al. (2023), Cardoso Castro (2019)	n/a	CY, SE
RQ2: How can Agile methodologies be utilized as effective tools to design and adapt new business models in response to the challenges posed by a VUCA environment?	Mendonça de Sá Araújo et al. (2019), Vasilieva (2021), Veretennikova and Vaskiv (2018), Yadav et al. (2020)	DT, ST	SW, SE, HW, ED, CO
RQ3: How does the integration of AI support decision-making processes for investments in innovative business model development?	Sharma and Kumar (2023), Nortje and Grobbelaar (2020), Kulkov (2023), Poeppelbuss et al. (2022), Ahmed et al. (2022), Akkaya and Ahmed (2022)	ML, ST, FZ, DT	CO, SE, HC, HW, SW
RQ4: What are the critical factors that organizations must consider while transitioning from Finite Games (characterized by known players, fixed rules, and agreed-upon objectives) to Infinite Games, as proposed by Sinek (2019)?	De Ruiz Diego et al. (2023), Horstmeyer (2020), Silva et al. (2021), Yan and Feng (2018), Zhang et al. (2021)	FZ, ML, NN	SE, SW, CO, CY
RQ5: How can organizations effectively balance and integrate the concepts of Uncertainty, Agile methodologies, AI, and Innovation to foster a culture of continuous learning, experimentation, and adaptation?	Buffardi (2018), Gupta et al. (2022), Schön et al. (2020)	DT	SW

As a summary, regarding Fig. 1, the various research methods used within the industries identified between the period 2018–2023 can be visualized. Also, Figure 2 shows what research methods were most used in the management model. And finally, Figure 3 displays what AI-related solution and/or non-technical approach were employed per industry (Fig. 4; Tables 3, 4, 5, 6, 7).

Fig. 2

Research focus based on industry

Fig. 3

Research methods by management model

Fig. 4

Applied techniques by industry

Table 3 List of papers (criteria met)

Table 4 Summary of articles that study fuzzy logic towards agile project management

Table 5 Summary of articles that study AI techniques towards agile project management

Table 6 Summary of articles that incorporate various techniques towards agile project management

Table 7 Summary of articles that follow a chaos-focused mindset towards agile project management

Results

The research questions previously formulated from the findings section are answered below based on the retrieved data.

RQ1: How can the Infinite Games theory, formulated by Sinek (2019), connect to the changing and volatile attributes of organizations operating in unstable conditions?

The rationale behind the Infinite Games theory is to play continuously in unstable conditions focusing on a long-term perspective rather than “winning” in the short term. In other words, organizations embrace resiliency approaches to adapt (or keep playing) and evolve (operating under volatile markets). Additionally, Humlung and Haddara (2019) claim that gamification can be a useful strategy for maintaining higher levels of engagements if appropriately conducted; thus, creating a positive and friendly environment between employees. In order to answer this first hypothesis, it is important to highlight the theory formulated by Sinak and the study proposed by Ingvarsson et al. (2023) offer a new perspective on how organizations can thrive under volatile markets through:

• Continuous engagement can be achieved through Gamification techniques where stakeholders are involved in a positive and motivating environment focusing on long-term healthy relationships among stakeholders aligning with the theory of Infinite Games,

• Sinak considers that Infinite Games are prone to ongoing improvement which translates into adapting, tailoring and evolving stakeholder relationships achieved through agile feedback and iterative engagement,

• Gamification usually offers short-term results; however, Gamification + Infinite Games can be integrated into a broader and thorough engagement strategy to create a sustainable long-term environment.

In other words, gamification and infinite games complement each other by building and maintaining adaptable, loyal and long-lasting relationships aligning with the complex circumstances of organizations.

To further develop this answer, the Infinite Games theory can also be merged with the ideas and concepts proposed by Cardoso Castro (2019):

• The VSM framework aligns with Sinak’s theory by encouraging organization to adopt a holistic perspective (the ecosystem of stakeholders) and adapt accordingly to the changing circumstances at all levels,

• The VSM model introduces the nature of adaptive structures; this means that organizations can place small units to respond effectively to complex situations, a key feature in Infinite Games. Furthermore, these small units support organizations to “fail sooner and fail better” by learning from previous experiences and adjusting strategies depending on internal/external factors,

• Creating resilient feedback loops is an important aspect within the VSM model is imperative to support and stay responsive to the changing nature of the market.

RQ2: How can Agile methodologies be utilized as effective tools to design and adapt new business models in response to the challenges posed by a VUCA environment?

To properly elaborate this research question, three different studies serve as the foundational basis for identifying tools to new business models under VUCA conditions:

• Design Thinking practices such as collaborative ideation (create novel ideas or proposals for exploring new chances), empathy (usually used to understand needs and pain points of stakeholders) and problem framing (conduct root causes analysis prior to potential solutions) can be easily structured and devised in conjunction with any Agile framework to embrace new business opportunities under VUCA-related issues (Mendonça de Sá Araújo et al., 2019; Vasilieva, 2021),

• Design Sprint activities such as divergent thinking (different perspective about solving a specific problem can be highly effective for creative problem-solving), time-boxed nature (activities are to be performed during specific intervals based on an agenda), prototyping can be creatively conducted to foster rapid adaptation and enhance the organization’s focus on VUCA-related problems (Mendonça de Sá Araújo et al., 2019),

• Compiling Lean startup principles with Agile frameworks provide insight about how to navigate volatile conditions for adapting or creating business models through the use of: MVP, iterative development and rapid prototyping can be used to validate assumptions, hypotheses, or any related artifact and take corrective actions if necessary; strategic pivoting, team members need to be aware to business model adaptation if market conditions tend to change; experimentation and human-led Jidoka, applying lessons learned can guide the organization during turbulent times and trying unknown or automation approaches can be useful as well (Veretennikova & Vaskiv, 2018; Yadav et al., 2020).

RQ3: How does the integration of AI support decision-making processes for investments in innovative business model development?

AI technologies facilitate considerably decision-making processes, for example: AI systems can automate routine tasks, perform predictions, identify patterns and analyze large datasets. In addition, AI can be a useful enabler for developing innovative business models through:

• The TISM-MICMAC approach reinforces the identification of consumer preferences and new market trends by implementing sound AI algorithms (machine learning and fuzzy techniques) for predictive analytics and continuous risk management, purging for data quality, and the availability of AI expert contributing to the enhancement of new business models (Sharma & Kumar, 2023),

• AI decision systems support and automate the alignment of decision workflows and IT strategic objectives by using machine learning tools and data governance models for AI-driven personalization and tailoring experiences; thus, developing innovative business models for companies in unstable market conditions (Nortje & Grobbelaar, 2020),

• Due to standardization and compliance requirements, certain industries benefit from the deployment of AI solutions by identifying new market niches, unfulfilled needs, proper resource allocation, new trends and demands, adherence to national/international regulations, enhanced risk management approaches for investing in novel business models (Kulkov, 2023),

• Conducting iterative uncertainty reduction techniques, as recommended by Poeppelbuss et al. (2022), it is feasible to automate repetitive tasks, continuous innovation, personalization of business needs based on customer insights,

• Integrating Agile techniques and VUCA 2.0 strategies can be used to explore new business model investments focused towards Industry 4.0 and Industry 5.0 (Ahmed et al., 2022; Akkaya & Ahmed, 2022).

RQ4: What are the critical factors that organizations must consider while transitioning from Finite Games (characterized by known players, fixed rules, and agreed-upon objectives) to Infinite Games, as proposed by Sinek (2019)?

A hybrid approach can be implemented to answer this question based on:

• Fuzzy logic algorithms can be leveraged to showcase how the transition from Finite Games to Infinite Games should be achieved usually learning from past experiences, continuous refinement of risk management and the measurement of risk analysis based on quantitative techniques. Furthermore, employing fuzzy techniques and other AI algorithms in conjunction facilitate this transition towards data-driven decision-making processes to support technological insights (De Ruiz Diego et al., 2023),

• Curiosity as a catalyst for transitioning from a rigid structure to a goal-focused organization, from a bureaucratic style to a flexible mindset, from a punishment perspective to embracing ongoing learning from successes and failures. Also, curiosity can be used for soft skills development positioning with the concept of Infinite Games (Horstmeyer, 2020),

• Creating a stakeholder-centric environment which is to consider and engage not only traditional stakeholders within the performing organization, but to also include the wider community (partners, customers, employees, affected parties from the community) aiming for a common sense of purpose and long-term win–win outcomes (Silva et al., 2021),

• Adding AI-led recommender systems for ongoing adaptability based on learning and adapting continuously recommending strategies, user feedback and cultural transformation related to how cross-cultural teams can foster innovation and experimentation to improve recommender systems (Yan & Feng, 2018; Zhang et al., 2021).

RQ5: How can organizations effectively balance and integrate the concepts of Uncertainty, Agile methodologies, AI, and innovation to foster a culture of continuous learning, experimentation, and adaptation?

Nowadays, organizations commonly operate on VUCA environments making this a vast opportunity for organizations to experimentation, learning activities (such as hackathons and training programs), optimization per se, application of proven Agile methodologies for small teams and scaled-Agile settings, among others (Buffardi, 2018; Gupta et al., 2022; Schön et al., 2020); it is feasible to draw lessons to foster adaptation, learning and experimentation under unknown circumstances.

Discussion and conclusion

Zinkin (2020) shares the antidote against the VUCA model through a new perspective and evolution about itself—the VUCA 2.0—means the new way to face the challenge of sustainability for any companies in the market. Moreover, the meaning of VUCA is Volatility, Uncertainty, Complexity and Ambiguity; however, this novel VUCA model 2.0 brings another revolutionary focus and concepts with regard to each variable. Zinkin proposes new theories and meanings such as Vision, Understanding, Courage and Commitment and Adaptability.

Every edge component in VUCA model has its reciprocal variable in VUCA model 2.0 such as Volatility is managed through Vision, Uncertainty through Understanding, Complexity through Courage and Commitment and finally Ambiguity through Adaptability. This new focus shows a new challenge for the managers in the companies because they will have to drive with the new strategies oriented to survive and guarantee the sustainability of the organizations in the VUCA market. Lastly, the author stands out the importance of making a Governance Corporate because they will be who take the responsibility of managing, leading and generating the policies which will be applied into the organizations and allow their survival in complex market.

There are many examples of companies that led the markets in their different industries and that were successful management models at the time, such as Blockbuster, Nokia, Kodak, Blackberry, Atari, among others, however, one of the greatest lessons learned in the business world is that every business model has an expiration date. Therefore, those companies that have not started working on innovation processes to renew their business models or design and develop new models are destined eventually to disappear because someone else will develop a new model that will displace the one, they have not renewed, as stated by Simon Sinek [24] in his theory of Infinite Games.

Some Agile methodologies help organizations to manage new opportunities by designing new business models, however, for these models to be scalable, it is necessary to use Artificial Intelligence as a platform that will oversee processing the data extracted during the model incubation process. In such a way that, as a result of the information processing, these provide us with projections on the sustainability of the business model and, in turn, allow us to make decisions on the feasibility of the investment.

To better manage the uncertainty that surrounds any innovation process, two key aspects must be combined, such as: (I) the use of Agile methodologies, and (II) the use of Artificial Intelligence, because the first of them discovers new business opportunities and the second validates whether that opportunity is scalable as a business model. Therefore, there is a symbiosis between both areas of knowledge, and they complement each other to help mitigate uncertainty when developing new Business Models.

Limitations

It is important to explain the limitations and constraints of this study. For instance, a restricted sample was obtained from journal articles from 2018 to the beginning of 2023. Therefore, the findings cannot be applied to other industries and its applicability at different times is unknown. Additionally, this study has selected some restrictive variables such as uncertainty, business models and project management practices. This may translate that other unknown variables, not considered for this paper, might offer in-depth analysis and description of the problem.