I was asked recently whether we make use of AI at TEQ. This is becoming a common talking point with business owners. Sufficiently common that I turn the question to become How do I use AI in my business? The use of AIs, specifically tools like ChatGPT, has become so common that we should be trying to understand – where relevant – how these tools get used. I think we have moved well past the binary ‘yes/no’ question.

And I think this is a fair question, particularly for a firm like TEQ, where our work product is made up of ideas, analysis, and written reports. As a client, what do you get for your money? The business owners I have discussed this matter seem to be motivated by one of the following concerns:

1. Am I paying for GPT-based analysis and reporting dressed up as something more complex and human centric?
2. If not, would GPT-based analysis and reporting provide a better outcome?

As of early March, 2024, our approach to the use of GPTs in our business is as follows:

• All of our client work is original and ‘hand crafted.’ We do use proofing tools like Grammerly and those included in Word, Excel, Outlook, and Powerpoint.
• We do not use any AI platforms to analyse any client data. Your data is not being used to train AI systems.
• We use AI image-generation tools to create ‘artwork’ used in posts on our website.
• GPTs are often used to flesh out and review articles and posts on our website. We then take that output and rewrite in ‘our’ voice.
• We have started using ChatGPT for article ideation. In other words, coming up with topics for posts. This is currently quite rare (but that may change) as we like our website to be authentic. As the AI tools develop, authenticity is likely to be easily replicated.
• We don’t use AI output in correspondence (letters, emails). This may also change over time as GPTs get better integrated into Office apps like Word and Outlook.

So, no, we do not pass off AI output as ‘ours.’ All of our client work is done by humans.

The second concern is whether we could be using AI for better (client) results, but don’t (i.e., maybe because we are Luddites). Our primary concern here is the security of client data. No one wants their data turning up in a GPT result. But we also find drafting our reports to be a critical part of our process. We identify gaps in the data, inconsistencies in assumptions and findings, and other potential shortcomings through the drafting process.

We are seeing a lot of tools becoming available to ‘automate’ the business valuation process. So far, we have found them wanting in quite important ways. The market will shift, undoubtedly. I don’t think these types of AI tools are going away, and they will only become more powerful, more integrated, and more natural (‘humanlike’) in their output. This is a good thing, and we aim to adapt along the way.

Our key commitment is to make it clear to our clients where and how we use AI tools in our work product. We also commit to be transparent about the use of client data with AI tools.

I will endeavour to keep this post updated with any changes to how we use AIs and related tools. If you have any specific queries about how we use AI in our business, leave a comment and we will get back to you.