If AI is already on a path to replace specific skills performed by humans, why are we still pricing AI software using size-based or usage-based pricing? When AI skills[1] replace human skills, should we not compare and price AI software to wages? If AI replaces tasks, skills, and eventually jobs, then the prevailing pricing model of software should allow for a direct comparison of the AI software costs to the wage for a person in a specific job. And the time the AI skill is used during the day. No different than the sum of skills, making up a job for which an employee is paid a salary. Either hourly, weekly, or monthly. While complex, such a shift to compensation-based pricing could benefit software providers since there would be a similar activity, human work, currently done at a price. It would be easier for organizations to compare software features and benefits with the current approach. And perhaps most interesting, it would, from a societal and economic perspective, allow for a more manageable cost and-benefit comparison of AI software’s impact on the overall economy.
As outlined in our recent article on AI skills, the move to a skills-based view of jobs, the organization's capabilities, and the rapid emergence of AI skills are underway. This may change how AI software is priced, particularly AI software that mimics a specific human skill. We predict that organizations will compare the cost of AI solutions that address one or several critical skills with an employee's salary or a portion of their compensation. Why is this change possible now? The reason is that AI-based software solutions, because they address decisions humans make and allow for the automation of tasks, will be evaluated differently than past process-software solutions (such as ERP implementations) or office productivity software. If or when this was to happen, comparing the cost and value of AI skills solutions become far more straightforward. However, organizations must still consider multiple factors and adopt a TCO perspective of AI skills software solutions. These factors include, among others:
Human employment challenges: For work done by humans, we always need to consider that work is affected by attrition, scarcity of skills, need for development and learning, challenges of job satisfaction, or location of talent, to mention a few. For skills handled by AI, these challenges are minimized or eliminated. As leaders look to where they deploy AI skills, the human employment challenges will inevitably need to be considered, even if their impact will vary from one organization to another. Comparing the cost of human versus AI skills will require understanding the cost impact of these factors. It may lead to AI skills solutions being more expensive than comparable human skills.
Productivity differences: In a world of competent AI skills, a particular AI skill will isolated, be more productive than a human skill. The challenge is that most skills don't exist in isolation, they are a component of what makes a job, and jobs are an element of the work done by a team or organization. So, while the productivity of a specific AI skill compared with that of a human may be more productive, that may not lead to increased job productivity or organizational productivity. As organizations deploy more AI skills and reorganize how humans work, executives and managers need to understand better the inter-relationship between skills and how productivity improvements in one through AI skills will affect other skills. And in turn, consider how work gets organized as skills evolve (both AI and human skills). This will result in TCO analysis of AI skills versus human skills remaining complex to ascertain.
Evolution of wages: In a world where AI replaces entire jobs, the tipping point becomes when the total compensation cost of AI is lower than the human compensation cost. Essentially when the wage (total compensation) of AI is lower than that of a human. We are still some years from this being the case, but already today, we can talk about the cost of an AI skill versus that of a human. It’s clear that for this comparison to become viable, technology companies offering AI skill solutions must make them comparable to the wage of their target market. Conversely, HR organizations must break down jobs and total compensation to the skill level as part of their focus on skills. This would enable an apples-to-apples comparison of AI and human skills. However, it will require a new pricing approach for many AI-focused software companies. When that happens, how we look at wage and productivity growth from an economic perspective will change. No longer will productivity be only human productivity but also the productivity of AI skills used in organizations. And the cost of labor will again be both human and AI, with both likely, evolving at different speeds.
Cost of AI Skills: Not all AI solutions can be considered AI skills. Therefore, not all AI software should be priced to allow comparison to human wages. But where possible, it would entail a more straightforward comparison and enable HR organizations to evaluate AI skills solutions, providing insight into the solution's functionality and attractiveness from a pricing perspective. AI software providers and organizations (led by the IT function and HR) should seek greater clarity around AI software pricing. Both sides, buyer and seller, would benefit.
Whether there will be full-fledged AI skills solutions comparable to human skills and pricing so that you can compare it with what the human is paid is likely. More likely to happen short-term than complete digital human AI-software solutions that eliminate entire jobs. The AI industry and buyers of AI solutions would benefit from greater transparency in pricing, especially for AI skills solutions that compete with human skills.
[1] “AI skills” describe software solutions involving one or more algorithms or ML environments that, combined, mimic a specific skill that makes up part of a job performed by a human.
Kommentare