Ownership of Digital Assets Generated by Human Intelligence Remains Unclear

The emergence of artificial intelligence has brought about significant changes in various industries, but one crucial aspect remains unclear: who owns the valuable assets generated by human intelligence? The concept of "traces refers to the digital footprints left behind by individuals on systems they interact with, including corrections, judgments, and creative work. These traces are considered a company's most valuable assets.

The question of ownership is becoming increasingly urgent as companies build AI models using employees' work without clear guidelines. If not addressed soon, powerful tech firms may set the terms for accessing human intelligence, forcing others to pay a rent for using their own data. This could lead to an uneven distribution of benefits and create significant economic disparities.

The rise of radio broadcasting in the 1920s serves as a cautionary tale about the importance of establishing clear rules and regulations. American broadcasters quickly staked out radio frequencies, leading Congress to pass the Radio Act of 1927 and later the Communications Act of 1934 to ratify their control. Similarly, the airwaves were treated as a shared resource, but now human intelligence is the valuable asset at stake.

The stakes are high because companies building AI models on employees' work are staking claims about who owns the system's learning outcomes. If left unregulated, powerful tech firms may dominate the market and dictate terms for accessing human intelligence, further exacerbating economic disparities. The need for a social contract that addresses this issue is pressing.

A proposed social contract has been put forth, comprising three layers: platform, organization, and individual. This framework aims to establish clear guidelines for who owns what in the intelligence economy, ensuring that individuals and organizations have control over their own data and creations. However, much work remains to be done to shape these rules and ensure a fair distribution of benefits in this emerging field.

The way organizations interact with AI platforms is crucial to preventing a single model from dominating the intelligence economy. This means providing choice at the platform level, so companies can switch between models without being locked into one particular infrastructure.

However, this freedom of choice is often restricted by providers who make switching costly and difficult. As a result, many organizations are forced to stick with the first AI model they adopted, rather than exploring alternative options that may better suit their needs.

At the organization level, there is an issue of sovereignty over the learning loops, workflows, and domain expertise built on top of AI systems. This expertise is often created through human effort and should remain under organizational control.

One area where this is particularly relevant is in the use of public AI tools by employees at work. When staff use these tools to complete tasks, they are inadvertently sharing valuable information about their organization with external agents, including how it operates and makes decisions.

When individuals leave a job, their judgment should be valued more than when they first started. This idea is based on an investor's perspective, where a person's worth increases over time. However, this concept raises questions about whether humans can improve their skills like AI models do.

The over-reliance on tools like ChatGPT can hinder learning and retention. In one study, students who used the tool as a study aid scored lower on a test 45 days later compared to those who relied on traditional methods. This suggests that while the tool provided accurate answers, it didn't promote long-term understanding or recall.

The design of AI tools can significantly impact their effectiveness. Researchers at Harvard conducted an experiment where they tested an AI tutor that required students to reason and find answers on their own before providing assistance. The results showed that these students achieved better learning gains immediately after the lessons and spent less time in the classroom compared to peers using active-learning methods.

The goal of technology should be to enhance human capabilities, not replace them. This is in line with our 2017 Coaching Networks thesis, which focused on sharing top performers' skills with others. However, the next generation of AI tools must go beyond mere knowledge transfer and enable people to develop new skills and strategies.

Friction is essential for workers to stay adaptable as AI continues to advance. Without it, individuals may become complacent and lose their ability to innovate and improve over time. This highlights the need for AI tools that promote critical thinking and problem-solving skills, rather than simply providing answers.

As professionals transition out of their roles, they often leave behind a wealth of knowledge and expertise that is valuable to their former employers. However, this can also lead to a loss of personal ownership and control over one's own skills and judgment.

The concept of a cognitive passport has been proposed as a solution to this problem. This would involve creating a portable, anonymized record of an individual's approach to problems, excluding sensitive information such as confidential processes and deals. By doing so, individuals can retain ownership of their expertise while still allowing companies to benefit from the knowledge they've gained.

For instance, consider an enterprise sales representative who has spent four years with a company. During this time, its AI systems have watched thousands of her calls and learned valuable insights into her negotiation strategies, pricing models, and customer interactions. This includes details such as how she opens discussions with buyers, handles CFOs, and adapts to technical buyers.

When the sales representative leaves the company, all these valuable assets - including call records, lead lists, customers, and deal details - remain with the business. What's more, the AI system retains its coaching model, which is informed by everything it learned from her experience.

This highlights the importance of developing AI tools that prioritize knowledge retention over data ownership. By doing so, companies can foster a culture of continuous learning and growth, while also empowering employees to take their expertise with them as they move on to new challenges.

The data generated by employees is not just a collection of metrics and statistics, but also contains unique patterns and habits that are shaped by their work. For instance, an employee's tendency to deviate from scripted responses or override systems when necessary can be tracked and recorded in their passport". This digital document not only reflects the employee's skills and performance but also includes information about company policies, such as discount floors.

While AI has the potential to automate certain tasks, its primary goal should be to augment human thinking and make it more powerful, durable, and portable. By doing so, organizations can reap the benefits of increased productivity and efficiency without sacrificing the value of human expertise. However, there is a risk that the intelligence generated by employees will not benefit them directly but rather flow to platforms they do not control.

The use of AI in business functions has become widespread, with 89% of organizations employing it in at least one area. Nevertheless, only 44% have managed to scale its implementation effectively. This disparity highlights a critical juncture where companies can still make informed decisions about how to integrate AI and maximize its value.

A crucial factor in determining the success of AI integration is the level of trust between employees and the system they use. When people feel confident that their thinking and expertise are valued, they are more likely to contribute meaningfully to the organization's growth and development. Conversely, if employees perceive the system as controlling or exploitative, they may resist its implementation or even leave the company.

The window for making a genuine choice about AI adoption is still open, and companies must seize this opportunity to create systems that empower employees rather than simply automating tasks. By prioritizing transparency, trust, and employee engagement, organizations can unlock the full potential of AI and drive lasting growth and innovation.

As this framework is proposed, it's essential to subject it to rigorous stress-testing by a diverse group of stakeholders, including founders, researchers, legal scholars, and labor voices. The author acknowledges that they may be mistaken about certain aspects of the framework.

The unique ability of humans to adapt and innovate in response to changing circumstances makes them the only true mutation engine in an era dominated by artificial intelligence. In contrast to AI's capabilities, which can excel within predetermined rules, people possess a vital quality - their judgment - that should be respected as co-authorship rather than mere raw material used to train AI systems.