Abbey Library of Saint Gall, UNESCO World Heritage Site. Picture Credit: Wikipedia
The inspiration to write this essay goes back to an invitation to be guest-speaker and panel-member at an Alumni-Event of the University of St.Gallen in June 2019. Under the topic ‘Leadership 2025’, several issues will be discussed, last but not least the impact of AI in respect to guiding corporations five years from now. 2025 is close enough to support one’s imagination as to how a leadership scenario could evolve. However, as viewed by the AI community, we are still decades away from reaching the goal of human-like General Artificial Intelligence (AGI). According to a global study conducted by the University of Oxford last year, 352 AI researchers saw a 50% chance of AI outperforming humans in 45 years from now. Asian respondents expected this date much sooner. Their estimates ranged from 15 to 70 years until AGI would be reached. Regardless of who is right, 2025 represents just an intermediate step in the ongoing effort to advance AI to human level intelligence.
Fuelled by the discussion on ethics in AI, new initiatives have been launched to advance AI from an interdisciplinary human-focused point-of-view. Human thought – as applied in philosophy, history, psychology, sociology and art – needs to complement the natural sciences like physics, mathematics, quantum science and biology. MIT’s interdisciplinary ‘Quest for Intelligence (QUEST)’ initiative launched in March last year, followed by Stanford’s ‘Human-Centred-AI (HAI)’ Project announced last month, stipulate that the potential negative aspects of AI, such as brain hacking or human identity manipulation, require a much more human-centric view to foster the potential benefits of AI, both in an individual as well as a social context.
From today to 2025: Digital Transformation
Digital transformation is in full swing and the corporate thought-model to be discussed later assumes that by 2025 digital transformation has been accomplished. This implies that all data produced by the activities of the corporation are labelled and classified with no built-in biases distorting the data. The data analytics procedures and the corresponding algorithms are in place, providing the foundation to model corporate activities and test possible opportunities with a digital ‘sandbox’. In manufacturing and supply logistics, ‘smart’ robots are in place to cut cost and improve the quality of products and services. The timeframe of five years to reach this point is ambitious. While digital transformation relies on artificial neural networks and deep-learning algorithms to detect patterns which can be used to define deviations from a given norm, this mathematics-based approach, although very successful in specific applications, significantly differs from the way the human brain generates intelligence. Consequently, AI mathematicians have started to team-up with neuroscientists to overcome the limitations of what is referred to as ‘Narrow AI’. Major research efforts, studying brain activity as response to human behaviour or modelling brain activity with computer software, are under way to better understand human intelligence. These efforts are designed to augment humans, for example by improving their memory or cognitive capacity.
2025 and Beyond: With Common-Sense Reasoning to Human-AI Transformation
By 2025 corporations can expect to be confronted with a new transformation process titled ‘Human-AI Transformation’. This process includes the application of machine-based common-sense reasoning. Wikipedia defines common-sense as the basic ability to perceive, understand, and judge things that are shared by (‘common to’) nearly all people and can be generally recognized without need for debate. It is common-sense that helps us quickly in answering the question, ‘can an elephant fit through the doorway’? We are usually not conscious of the vast sea of common-sense assumptions that underlie every statement and our every action. This unstated background-knowledge includes: a general understanding of how the physical world works (i.e., intuitive physics); a basic understanding of human motives and behaviours (i.e., intuitive psychology); knowledge of the common facts that an average adult possesses. The way to obtain common-sense knowledge is by learning it or experiencing it. Humans have the advantage that their bodily sensual experience based on sound, smell, vision, touch and taste is processed and stored by the brain’s cognitive capability in real-time without the need for labelling huge datasets. In a corporate context common-sense knowledge represents an accumulation of all experiences and values that are performance-relevant. An intelligent machine which exhibits common-sense reasoning will be capable of predicting the results of people’s behaviour and intentions and their perception of humans’ natural understanding of the physical world with a high degree of accuracy. Nevertheless, sound human judgement and intuition must be kept in the loop to provide a ‘second opinion’ before predictions are used to shape decisions.
One of the economic drivers to advance AI with common-sense reasoning is the fight against fake-information distributed over social-media networks. Currently many thousands of individuals are contracted by Facebook, Google and Amazon to uncover fake information. The success rate of this effort is by no means adequate to prevent the ongoing erosion of trust which – additionally fuelled by cyberattacks – threatens the credibility of today’s global communication infrastructure. Common-sense enabled systems will identify fake information far more efficiently. With huge research efforts on the way, we can expect that by 2025 software tools will be available to build and maintain a corporate common-sense library.
A 2025 thought-model of corporate activity
By 2025 corporate knowledge-silos will no longer exist. Modelling corporate activities with a digital ‘sandbox’ enables rapid prototyping for testing new business models and business opportunities. Artificial neural networks and deep-learning algorithms that generate knowledge from ‘big data’ are the building-blocks of this ‘sandbox’. Adding common-sense reasoning to information that defines a corporation’s identity and values perceived by the inside as well as the outside world, gives corporate leaders the opportunity to reflect on future strategies or invoke decision-making with a much higher level of confidence compared to the tools provided by ‘sandbox’ data analytics alone. To realize this advantage, it will be necessary for leaders to be able to optimally balance the interplay of intelligent machines with human assets to solve specific tasks. Augmented by AI machines, collective intelligence of project teams will fuel creativity and innovation to arrive at the optimum product and service-offering. The corporate structure to support this business culture can be built on three pillars:
A) Human resource management
- Define the human profiles required for hiring team members meeting corporate goals.
- Provide coaching-support to high-performance teams.
- Continuously support individuals to re-educate themselves as job-profiles change.
- Maintain and adjust hiring policies including the contracting of free-lancers. Assure that evaluation procedures are in line with accepted ethics standards.
- Contribute to the expansion and maintenance of the corporate common-sense library.
B) Technology Management
- Define and maintain the infrastructure required to meet corporate goals including the means required to protect the company from cyber-attacks.
- Define and manage the outsourcing to service providers and cloud services.
- Maintain the corporate, deep-learning knowledge data base.
- Manage the introduction and operation of ‘smart’ robots in the manufacturing and logistic process incl. IoT.
- Stay aware of new technologies that might impact current corporate activities.
C) Communication management
- Define the product- and service offering as well as customer segmentation.
- Define and manage the sales-process incl. applicable AI tools.
- Coordinate and manage customer- communication incl. the contact to outside opinion makers to assure a consistent and trusted corporate image policy.
- Coordinate and manage internal communication and assure consistency with corporate brand policies.
- Contribute to the expansion and maintenance of the corporate common-sense library.
With this structure in place, the corporation will resemble a living organism which can master the ‘survival of the fittest’ in a highly competitive environment. However, this goal is only achievable if trust and credibility remain the guiding principles internally as well as externally. Failing this task might end in a reputation disaster with negative, long-term consequences reducing corporate value. Consequently, the engagement of strong ethic guidelines and its adherence to them must be a constant issue in corporate governance.
Leadership in 2025
The 2025 thought-model distributes leadership over three tightly interconnected units: human resource management, technology management and communication management. The joint leadership-team defines the mission and benchmarks to be met including the long- and short-term strategies to follow. Corporate governance, adherence to ethics standards as well as performance reporting are the joint responsibility of the team leaders. The integrated corporate AI system will provide support for the best course of action for resolving unforeseen problems. Nevertheless, human judgement needs to be kept in the loop to spot machine failures.
Future leaders must comprehend the speed and complexity of the spiralling dynamics of scientific progress. Just as emotional intelligence became an important leadership-asset about ten years ago, we now see a rising need for leaders to apply and support collective intelligence with an appreciation of interdisciplinary thinking, encompassing technology, neurology and the humanities. Continuous education and personal development are required to meet such high standards. Possibly the best candidates for leadership positions are those who show a strong learning ability coupled with curiosity and a high level of self-consciousness.
With a corporate knowledge- and common-sense-library in place, one could question the need for hierarchical leadership. Fascination with organizations that eschew the conventional managerial hierarchy and instead radically decentralize authority has been longstanding. Self-organizing concepts such as Holacracy or Heterarchy might gain new momentum as human centred AI will support a new class of networked operations. Whatever leadership and organisational structure are under discussion, there is clearly a need to discuss a 2025 corporate scenario. Disruption by start-up incumbents has never been so intense and it is likely to increase as huge amounts of venture money, both from the US and China, continue to flow into the market.
I Hope that „Corporate Silos“, human thinking and decision Taking will not be replaced by AI, but complemented by it – i.e. AI in the Service of humans, and not the other way round. I Doubt whether it will be possible to Perfect AI without built-in bias – just as every human being has a bias, so collective decision finding Assisted by AI is the Most Likely way to get a good result.
„Outperforming humans“ means what…?? In what Respect??
Fake news Emerge from Fake Sources – the Problem Starts with Fake identities, i.e. Trolls and Bots…..