The trust gap widens as artificial intelligence (AI) goes mainstream, according to research from Salesforce. Brands are turning to generative AI to boost efficiency while improving customer engagement. And customers -- wary of the technology risks -- demand a thoughtful approach built on trust. Eighty percent of customers say it's important for humans to validate AI's outputs.
A deeper look into the widening trust gap as AI goes mainstream reveals that customers remain cautious. As companies focus on efficiency, generative AI promises time and resource savings by scaling content creation. Six in 10 desk workers use or plan to use generative AI. Customers also stress the importance of human touch in the AI era. A mere 37% of customers trust AI's outputs to be as accurate as those of an employee. Accordingly, 81% want a human to be in the loop, reviewing and validating those outputs.
Also: How to use ChatGPT to write code
Nearly nine out of 10 IT leaders believe generative AI will have a prominent role in their organizations in the near future, according to the State of IT report from Salesforce. According to McKinsey, 50% of organizations used AI in 2022. IDC is forecasting global AI spend to increase a staggering 26.9% in 2023 alone.
AI and automation is a focus in the pursuit of efficiency. Eighty-seven percent of IT leaders expect more investment in automation at their organizations during the next 18 months. In a world that is powered by AI, where most processes will be subject to automation and autonomy, a company's ability to sense stakeholder needs in real time will be a must-have capability requirement in the near future. The impact of AI on businesses is boundless. Therefore, companies need to become boundless in their ability to use technologies such as AI to reach their full potential.
Boundless companiestranscend the limits of traditional organizations. They are designed to achieve shared success, generating value for their customers, business partners, and communities as well as for themselves and their employees. This success is realized by resources that are individually empowered to be autonomous, connected, and mobile, and that are collectively organized to be integrated, distributed, and continuous.
Also: AI is more likely to cause world doom than climate change, according to an AI expert
It's natural to ask where to start on the road to becoming boundless. Here are some points we can offer to get you started on a successful path:
The traditional operating model is based on the idea that a company is a thing, an entity, a structure. But boundless organizations are differentiated in their responsiveness to both current and future customer needs and market conditions, and in their associated decision-making and action-taking.
Also: How to use the new Bing (and how it's different from ChatGPT
We introduce the Boundless Operating Model, our update to other contemporary sense-and-respond or situational awareness models, as a guide to help organizations design and develop the necessary processes and capabilities for amplifying and accelerating their responsiveness.
How do companies make themselves aware of what's going on around them, and how do they respond to those inputs? As we started to think about how companies can develop these capabilities, we looked at other situational awareness models to see what we could learn from them.
One of the models that we have used is known as theObserve-Orient-Decide-Actloop, aka the OODA loop, that was designed by military strategist John Boyd. His goal was to explain how fighter pilots can excel in a combat setting. His insight was that the ability to sense and respond to rapidly changing conditions more quickly than an adversary gives the pilot a competitive advantage.
The OODA loop describes the processes required to sense and respond, and enables the military to design the most streamlined ways of doing so. This model has been broadly adopted by the military and by businesses who can use it as a conceptual framework for designing and developing business management processes. The goal, as in the military setting, is to enable faster strategic decision-making processes and innovations through effective process management.
Also: How does ChatGPT work?
As we did our research, we realized that the OODA loop is very similar to theSense-Perceive-Decide-Actuatemodel that is used to describe how autonomous vehicles work. In this model, the vehicle performs the following functions in real time and continuously:
Sense:The vehicle uses its sensors to gather real-time data about its immediate environment or situation -- primarily the road and nearby vehicles. "Sense" is a broader word than "observe," and reflects (1) the fact that not all sensors are based on sight, and (2) that they're not assessing the meaning of the data they're gathering and sharing.
Perceive:The vehicle interprets the data that the sensors are sharing and makes sense of them, using machine learning and AI to do so. Tesla has developed its own supercomputer, Dojo, to train the on-vehicle intelligence with data from across its fleet to improve the accuracy of its perception and decision-making.
Decide: The vehicle then selects from among the options available to it and decides upon the safest course of action, again using its AI to do so.
Actuate:The vehicle initiates the actions that it has decided to take. The three primary controls that an autonomous car needs to actuate are the throttle for acceleration, the steering for direction and the brake for stopping. Of course there are many other actuators in a modern vehicle, but these are the three primary ones that help the vehicle achieve its mission and reach its destination safely.
In the case of the vehicle, the Sense-Perceive-Decide-Actuate model describes how a single autonomous system can operate effectively in and with the world around it.
The BOUNDLESS Business Operating Model includes 4 functions and 2 domains (Connection and Autonomy):
- Vala Afshar (@ValaAfshar) September 28, 2023
? Sense
? Understand
? Deicide
? Act
SUDA is an evolution of the situational awareness and sense and respond models, deliberately designed for dynamically changing... pic.twitter.com/3SKP5zHn9n
How, then, do these models of autonomous systems apply to boundless businesses and how might we adapt them to create a Boundless Operating Model? The short answer is they work very well, but we think the following tweaks are worth making to make them most useful.
First, boundless companies live as part of markets and ecosystems and communities, not apart from them. The idea of a company going it alone is no longer tenable. Identity is not defined in terms of isolation and exclusion, but in terms of connectedness and inclusion. From an operating model perspective, we want to be able to show that a boundless company is connected to a larger ecosystem and that everything it does happens within the context of that ecosystem.
Also: How to save a ChatGPT conversation to revisit later
Second, the models shown above tend to focus on sensing immediate and/or local conditions, what is sometimes described as situational awareness. This is critical to effective decision-making, but nowadays situational awareness by itself is no longer enough. Today's company needs to be horizonally aware as well. Horizonal awareness means being connected to the larger world beyond the immediate here and now. Companies need to be able to see "further down the road" in exactly the same way that an autonomous car can be aware of conditions anywhere along its journey, and can take active steps to anticipate and avoid problems, all because of its global as well as local connectedness.
Third, as may have become clear from the discussion of the same model applying to one autonomous system as well as to a "family" of them, boundless systems are self-similar at various levels (also known as fractal). Individual resources within a boundless company are themselves boundless, and have the same responsibility as the company to be responsive to the immediate needs of their customers and to current market conditions, and to be attuned to and prepared for future ones. And just like the company as a whole, when individual employees and teams take action, they do so in the world, not in a vacuum.
Our model must place action and reaction in its context, local or global, and it must be able to work at both the individual, the team, and the company levels. The outcome of these considerations is theBoundless Operating Model.
The Boundless Operating Model separates the four functions discussed above into two domains: Connection and Autonomy.
On the left side of the model, the domain ofCONNECTION, we have the Sense and Act functions that happen in, and in connection with, the larger world of which the system is a part. The point here is that these systems don't act in a vacuum. They act in and on the world and the world acts on them. There are two loops on this side. The smaller loop (solid line) describes the local and immediate world in which the system acts. This is the same as the "situation" in situational awareness. The fighter pilot, the autonomous car, the company, and the team, sales rep, or project team all have to be enabled and equipped to sense and respond to the situation they find themselves in, to react to a car slowing down in front of them, to have the information available to them to respond effectively to a customer call, and so on. The larger loop (dotted line) describes the global and near future or emerging world in which the system also has to act.
Also: How (and why) to subscribe to ChatGPT Plus
The autonomous car is not only connected to the other cars around it and knows through its sensors its real-time position relative to them and the other entities on the road, it is also connected to the larger world and knows the emerging conditions further on down the road. As a result it can predict the likely impact on its own "mission" and take pre-emptive actions to avoid it. All companies now have access to predictive and prescriptive data analytics, and the successful ones will be those that use these tools to make informed decisions and take anticipatory action, either to avoid knowable challenges or to seize knowable opportunities or both. We call the dotted loop "horizonal" to express this broader context in both time and space that all companies must be connected to in order to succeed.
On the right side of the model, the domain ofAUTONOMY, we have the Understand and Decide functions. Of immediate importance here is that in a post-pandemic, do-anything-from-anywhere world, a company's teams and employees are all increasingly distributed and remote from the center and must therefore be given the autonomy to make timely and accurate decisions about how to fulfill their missions. The company will need to revamp its management and supervision processes in the form of resource orchestration in order to deal with this decentralized and autonomous workforce. Meanwhile, the company itself must also be autonomous in the sense of having a unique identity and mission, despite its connection and its boundless nature.
As previously mentioned, the Boundless Operating Model is designed not only to support the requirement for all companies to have situational and horizonal awareness, but also to work at both individual and collective levels. Companies can use the model to design and develop processes for individual employees, teams of employees and organization-wide business management processes.
As a final point, there's broad consensus about the names of each function in the models we've shown, with only a few minor differences. We believe Sense is more appropriate than Observe given that this model can apply as well to technology systems like autonomous vehicles as it can to human organizations. We prefer Understand to Perceive or Orient because it has more general usage. All models agree on Decide. We prefer Act to Actuate again because it has more general usage, although we also recognize that this function may involve both actuation (putting into action) by the orchestrator component of the overall system and action by the executor component, where those components are decoupled from one another. This leaves us with the acronymSUDA (Sense, Understand, Decide, Act)for the Boundless Operating Model.
Also: ChatGPT is the most sought out tech skill in the workforce, says learning platform
In short, the Boundless Operating Model, orSUDA, is an evolution of other situational awareness or sense-and-respond models that are designed to reflect dynamically changing conditions, unlike for instance the PDCA or Deming Cycle which, was designed for continuous improvement in stable conditions or relatively controlled environments.
We did not invent the SUDA model. We have extended the 'sense' and 'act' portion of the 'connection' logic for both internal and external factors. It places an autonomous entity, be that an employee, a team or a company, in the context of a connected world both locally and globally, and it can be used by companies as a conceptual model to identify, define and develop processes that will enable it to act decisively, learn and improve continuously, and respond at the speed of need in that world.
This article was co-authored by Henry King, business innovation and transformation strategy leader at Salesforce, and co-author of Boundless. "Boundless: A New Mindset for Unlimited Business Success " by Henry King and Vala Afshar (published by Wiley).