You know, we get a lot of questions about AI in our monthly webinars. People often ask, “How is GreenSlate implementing AI into workflows?” and “What roles will have lower demand as AI becomes a bigger part of accounting workflows?” There’s a genuine  mix of concern about job security and a healthy interest in finding more efficient ways to work.

AI is here. You see it everywhere.  AI integration has spread so far and wide in the wider world of business productivity as a “solution” that you see companies scrambling to integrate it just to say they have AI, even if it doesn’t actually end up being truly useful in the end. 

Undoubtedly, AI is going to play a role in production accounting, but more so as a tool for efficiency and not a human replacement.

In the world of production accounting it feels like AI is still a solution in search of a problem.

Certainly there are challenges for production accounting teams on the ground which AI (when done right) would benefit, but the other side of the coin shouldn’t be neglected either. Studios and production companies could benefit greatly with AI integrations. In the short term, it feels like this is the most accessible entry point for AI to produce meaningful results. Analyzing across multiple projects to identify trends and rank performance will be invaluable to manage cashflow and forecast escalations.

Some entertainment payroll companies are eager to jump on the AI bandwagon, creating a lot of noise and promising revolutionary changes to how production accounting is done. While AI certainly has potential, it's crucial to focus on what really benefits production accountants and studios. 

So, let’s chat about this a bit.

The AI hype cycle

Be wary of companies promising groundbreaking changes to production accounting workflows. Why, you ask? Because the AI hype wagon is moving full steam ahead and it might look enticing on the surface level and in glossy marketing, but when you dig into the actual functionality and benefits of some proposed AI solutions, questions will linger. 

There’s three tenents any application of AI in production accounting must be absolutely tied to and provable in order for it to be more than marketing:

• Accuracy: Production accountants demand precise, verifiable numbers.
• Efficiency: Streamlined processes that save time without sacrificing accuracy.
• Reliability: Consistent results that can be trusted for critical financial decisions.

3 Reasons why AI isn't quite a magic solution (yet)

Lack of determinism

While AI, particularly large language models, may shine in creative tasks, they're not always perfect for things that have a 'right' or 'wrong' answer, also known as "deterministic."

They struggle with processes that require exact answers.  While AI might excel at writing detailed and complicated programming code, it works only because that code can be executed and validated as “right” or “wrong.” 

Then there is the issue with using the same prompts multiple times resulting in different outcomes.  AI operates on probabilities, based on the training data used and provides biased results based on that data.

Verification challenges

Any report generated by AI still needs human oversight, which in some circumstances may negate the intended time-saving advantages.

Trust issues

Again, multiple answers to the same prompt doesn’t instill confidence, and “hallucinations” are commonplace. Experienced accountants are unlikely to rely on AI for vital financial reporting without understanding the underlying calculations, especially when you are after hard numbers that must be accurate. 

AI's potential in production accounting

That said, while AI may not replace core accounting functions just yet, it certainly has the potential to enhance some areas, especially for studios and production companies. 

Here at GreenSlate we regularly conduct customer advisory board meetings to discuss innovation and understand what resonates with our users. Together, we can identify use cases where AI can provide advantages and efficiencies. We always want to know what you think.

Here are a few AI integrations we have been exploring:

• Natural Language Interfaces: Making complex queries more accessible for non-technical users.
• Automated Transaction Entry: Loading transactions into the application.
• Automated Categorization: Helping with the initial classification of expenses.
• Pattern Recognition: Spotting anomalies or trends in large datasets.

Some early ideas show promise, while others have shown to miss the mark. For instance, during these sessions we explored "natural language search" and found some neat use cases. However, it didn't quite excite our accountant feedback groups; when asked they often said, “I would so much rather run a report.” Indeed, sometimes adding a complex, but eloquent solution to a simple process is nothing more than window dressing.

Embracing change might appear at odds with the accountant sentiment above, but I don't think so.  We are all chasing efficiencies and there is no doubt that AI is here to stay, but this particular use case seems to be more hype than help.

And that is perfectly fine. Innovation comes from experimentation and exploration of ideas, and then executing on the ones that can be identified as having true, quantifiable impact. 

It is important to be open and curious about how the use of  AI can help make strides towards efficiency, and upskill so we can use the tool to become more efficient.

Our approach: efficiency through automation

Many of the use cases I hear floating around are attempting to solve “problems” we have actually already addressed within our thoughtfully designed system.

Our focus has always been to deliver tangible value through:

• Robust GL Platform: Providing accurate, real-time financial data. With drill-down capabilities to get to the transaction level details are already baked into our platform.
• Automated Calculations: Utilizing rule-based systems for complex pay calculations. A powerful example of this is our Hours to Gross System (HTG) that replicates the judgment a human expert would provide to convert tracked hours into gross pay based on union contracts.
• User-Centric Design: Developing features based on actual user feedback, behavior, and needs.

It’s true that production accounting is in the middle of a significant and long overdue evolution. 

There is absolutely no question that AI will play a vital role in the production accounting and payroll space. 

But the industry still needs solutions that prioritize accuracy, efficiency, and reliability. And while AI shows great promise, it’s not a one-size-fits-all solution for the challenges in production accounting. 

As we continue to innovate, our commitment remains clear: to deliver meaningful benefits to production accountants when they need it and without the marketing window dressing. Let’s keep the conversation going!