Brent Schiestl from Perforce recently spoke about the growing influence of AI on real-time workflows and the tools companies are using to manage digital assets and assess productivity. While businesses are increasingly adopting AI technologies, proving that these tools deliver clear, measurable benefits remains a significant challenge. While tracking costs is relatively straightforward, measuring actual output or productivity gains is more complex. A practice known as "tokenmaxxing," which involves setting arbitrary usage targets for AI, has been shown to be an ineffective way to gauge AI's true impact on productivity. According to Perforce’s 2026 State of Real-Time Workflows Report, real-time 3D engines—often associated with video games—are now being widely adopted across a range of industries, including aerospace, defense, public services, education, and media. These tools have demonstrated measurable improvements in productivity, largely due to their established success in the gaming sector. However, AI has not yet provided consistent, measurable returns on investment for many businesses. Concerns also include fears among employees: 50% believe AI could replace their jobs, 49% worry about the accuracy of AI-generated content, and 48% have ethical or compliance concerns about its use. Schiestl explained that one reason it's difficult to show AI's value is the difference in how people perceive its use in different contexts. For example, consumers are more comfortable with AI helping to fix bugs in source code, whereas AI-generated binary assets—like game assets or 3D models—face more scrutiny due to concerns about data privacy and lack of consent from original creators. To build trust, companies are adopting measures such as public disclosure of AI usage, ensuring human oversight in asset creation, and maintaining detailed records of where AI-generated content comes from. There is a growing shift away from tokenmaxxing, as companies seek more reliable ways to measure AI's impact. Today, metrics such as release frequency, content included in each release, usage telemetry, and industry standards like the DevOps Research and Assessment (DORA) are being used. However, new challenges have emerged, such as engineering teams manipulating metrics for better performance and difficulty in isolating AI's specific contributions from other changes in product releases. Real-time 3D engines are being adopted more widely by non-gaming industries, but their initial use was limited by their association with gaming. Barriers included complex licensing, tools that were difficult to use for non-artists, and challenges in integrating them with enterprise systems. Renaming reports to focus on "real-time workflows" instead of "game technology" helped reduce these barriers. However, industries new to these tools now face increased infrastructure costs, including GPU computing, storage, bandwidth, and licensing fees. Justifying these costs with revenue growth is challenging, especially for organizations that lack the technical capacity to manage such infrastructure. To manage the technical debt associated with AI, companies are using various tools such as AI-aware code reviews, static and dynamic analysis tools for AI-generated content, license and provenance scanners, and version control systems that treat AI-generated commits as standard parts of the development process. AI is also influencing the open source ecosystem by making it easier to develop and share solutions. However, some open source repositories are becoming private due to the high volume of AI-generated code submissions. Traditional software licensing models are being tested as companies may not grow as quickly as before the rise of AI. Whether the decision to build or buy software is changing as a result remains an open question.