It’s not time to freak out about AI…yet

Artificial intelligence became the most ubiquitous topic on the minds of business and government leaders in 2023, inspiring most to quickly form task forces to understand the technology’s opportunities and, more importantly, the way it would likely force their organizations to transform (yet again) in the years ahead. At this point, nobody can confidently tell you exactly how AI will shake things up 3…5…10 years from now. What many comms pros can tell you – including the more skeptical of the two authors of this newsletter – is that figuring out where to start to wrap your head around AI’s potential can be dizzying.

Is experimenting with an AI tool “cheating” in my role? If I use generative AI (GenAI), is the data I input into it secure? Will my role be obsolete in just a few years?

Luckily for us, very smart people have been thinking about the potential, the implications, the methods of safe use and more for quite some time. We recently sat down with David Sánchez Carmona, creative director-turned-AI strategist and gifted interactive designer at a multinational communications agency to dig into an array of questions.

David recognized that AI would change almost everything about brand and communications years ago and decided he’d take it seriously way before most people even let the word “generative” roll off their tongues. He’s since earned certifications from MIT and Microsoft and contributed to the successful launch of an adaptive intelligence model that supports the delivery of commercial insights.

The TL;DR of our conversation: AI shouldn’t be scary to communications pros, but you should be terrified if you’re not taking steps to understand how the technology can support your professional growth. Read on for the gems.

22,000 years

David described the large language models (LLMs) that serve as the foundation for most AI bots as a high-dimensional space. “If you and I were to read everything that GPT has been trained on, it would take 22,000 years.”

That level of knowledge is insurmountable for any human, not to mention the processing speed to pick apart all of that content and make any sense of it. But LLMs are really probability engines that determine what would be the next likely element of a sequence. You give it a prompt or serve it a question, and it spits out a string of characters it’s mathematically determined to make sense. 

So while a GenAI platform might be a great number cruncher, it lacks critical thinking abilities. That’s your strength as the human mind in the mix – and a capability that AI won’t soon be replacing. 

The “cost of intelligence”

You can’t really talk about AI these days without discussing Sam Altman, the recently embattled but restored CEO of OpenAI. He’s been a big proponent that the “cost of intelligence,” i.e., the investment needed to rationally process vast amounts of data, is diminishing through GenAI, as well as being democratized. This alone will transform how entire professions operate. But given that AI won’t be your critical thinking partner anytime soon, what benefit can it bring to brand strategists?

A couple logical areas come to mind:
1. Accelerate idea generation: With 22,000 years-worth of concepts at its disposal, GPT can be an ideal tool to sift through data to produce early-stage ideas that fit a desired context. Herein lies an area that comms pros shouldn’t shy away from: learn how to develop effective prompts (or queries) that can direct GenAI to deliver useful results to you. There’s actually an art to it, and this is a baseline skill most of us can easily build, in service of not falling behind.

2. Pattern and trend-spotting: Because it’s not as critical as we (ahem) are, AI’s not good at recognizing relevance, but it’s damn good at recognizing resonance. With the right prompt, a good chatbot can likely help you spot a particular trend that’s gaining traction within a community well before anecdotal feedback could make it evident. Patterns that your naked eye might miss would be much more recognizable to an AI tool.

Hubris check!

Don’t go getting too confident that GenAI will make you appear to be an instant expert on topics where your actual knowledge is light. Many a lazy attorney is finding out the hard way that prompting GenAI to draft court filings is a quick way to racking up legal fails. In addition to pulling irrelevant results into a document, GenAI can also provide false or inaccurate information (typically referred to as “hallucinations”). So David posed the simple notion to us: “If you’re working in an industry that is highly sensitive to mistakes, be mindful that one thoughtless generative AI communication can destroy your market share, just like one bad tweet could a few years ago.”

David advises that it’s more important than ever to keep your domain experts close; we remain far off from the point where GenAI can be considered “expert’ on anything. “If you're not a healthcare expert, do not fucking write about healthcare with AI. You're going to look like a prick when some medical professional with actual knowledge destroys you.”

Get a grip

If you work in comms, brand, reputation or marketing, there are pretty simple things you should consider when exploring ways to remain up on the powerful and evolving capabilities of GenAI. David recommends:

• Pick an AI tool that won’t disclose trade secrets. For instance, data you enter into ChatGPT is not confidential. But an upgraded version called ChatGPT Teams has enhanced data protections and may serve as a better option for those who want to experiment with a generative tool while not allowing third parties access to your prompts or other data.
• Value pricing has officially entered the chat for consultants. Theoretically, since the processing power of AI should make some really complex tasks much faster while maintaining – or enhancing – quality, there’s a value to this speedier output. For instance, if you have to create personas for target audiences, there’s typically an astronomical amount of data to comb through. With AI, you could drop that data into a model and have personas quickly produced. Your human intelligence and years of experience then equip you to double-check the persona outputs. Remember: in PR, you can typically create an output with two of the following three attributes: fast, cheap or high-quality. If someone tells you they can do all three…well, that’s just…not smart. Price your work appropriately, people.

If you’ve felt left behind by the neverending stream of mentions of AI in the news – and likely in the majority of business meetings you sat in last year – don’t freak out quite yet. But also, don’t get left behind.

For what it’s worth, AI will be responsible for the creation of entire career tracks we have yet to even dream of. David closed out our interview by advising us: “The more experienced you are with AI, the more you’re going to be operating at a higher level.” 

Skepticism aside, he’s definitely made us further believers.