Charting AI for AEC Marketers: Optimizing Utilization with Trends, Tools, and Tips

Several months ago, Sam Altman, CEO of OpenAI, predicted in a book called Our AI Journey by Adam Brotman and Andy Sack, that AI will handle 95% of marketing work. For marketers, this is sobering. Jobs will change, that is certain. But the biggest threat is digging your head in the sand and not learning about it or how it can make you more effective at your job.

We view ourselves as an AI-forward organization, and since our inception, we've embraced a mindset of education, communication, and experimentation. Education is the salve for fear, so we encourage employees to learn and practice as much as possible in their areas of expertise while keeping tool limitations in mind. In addition, we have created space in our organization to share learnings as well as mistakes with each other. We have multiple team members taking on and experimenting with a number of different tools, channels, and technologies to see what works and is the most relevant to our business.

Be cognizant of the potentially private or confidential data you input into models. The LLMs use prompts and feedback as reinforcement learning in fine-tuning models. For many of the LLMs, you have to pay to get better security. A rule of thumb is that if it is public information, i.e. front-facing on a website, it's probably fine to enter into the model. Also, read the fine print regarding meeting transcriptors. You might want to disable them. It's tempting to get a nice summary and action items after client meetings, but you might have discussed sensitive NDA-protected information that your client would not want used to train a large language model.

ChatGPT is good for researching, writing/rewriting, and even strategizing, but it is also a good idea to use and test the other major frontier models, such as Google's Gemini and Anthropic's Claude. We like Perplexity.ai because it sources without prompting. There was an AI tool explosion in 2023. Chat GPT is getting so advanced and multi-modal, with image, video, and analysis capabilities coming soon, that it can do what many of these tools offer. You just need to know how to prompt accurately. For ad versioning, we use Creatopy. For social graphics, we use AI capabilities built into Canva. For voiceovers, we've had luck with Murf.ai. These tools have resulted in welcomed production efficiencies.

Seek out lifelong learners and candidates with a willingness to use AI and learn new skills. Also helpful is flexibility and an open mind that a job today might look different next year. Ask candidates how they currently use AI and have them provide you with specific examples of tools, successes, and even learnings.

We educate them constantly on how AI is changing the landscape and how it is making our work better for them. We also disclose its use in our yearly planning agreements.

We are excited to experiment with GPTs, creating custom GPTs for various tasks and digital twins and custom GPTs by role (such as a marketing specialist) that mirror and support IRL teams. We are curious to see if the AI assistants will uncover creative insights we have not considered. We're also looking forward to testing text-to-video tools as well as tools for editing longer-form videos into clips. For 2025 planning, we look forward to gaining insights into target audiences using synthetic focus groups. We're hearing that the latest model of ChatGPT will offer more capabilities and could come as soon as this fall. Then there is always interesting conversation speculating when the models will reach AGI (artificial general intelligence), when they equal or surpass human intelligence, can operate across domains, and can learn, improve, and solve problems independently.