Commentary

Here's Some Human Intelligence On The Artificial Kind

As we know from watching lightning, it’s easy to ignore threats when they seem distant. When they move nearer, we pay close attention.

This has been the case with the marketing and media industry following the release of ChatGPT-3. Technology that can write efficiently and with enough credibility to pass for a human author hits the industry too close to home for comfort.

Trade media articles, blogs and social media have been filled with executives’ sharing their experiments with ChatGPT-3 or expressing their fear of its impact. For example:

·      Look at how ChatGPT-3 did better than expected in writing _________ (fill in your favorite form of marketing or job-seeking content)

·      Will brand companies no longer need agencies?

advertisement

advertisement

·      Whose jobs will be affected (especially mine)?

An upcoming book, The AI Conundrum: Why Artificial Intelligence’s Strengths Are Also AI’s Weaknesses, authored by noted marketing analytics leader Rex Briggs and his son Caleb, provides some solace. The book, which I had the privilege to help edit, points out that as powerful AI-powered applications can be, they will not fully replace humans, because AI is only as strong as how it is programmed:

·      Even the most well-intentioned AI programmers inadvertently insert biases into their work. For example, facial-recognition software has had difficulty identifying people of color, because the training set used was not sufficiently diverse.

·      AI relies on learning from the past. It cannot incorporate what it has not been programmed to do or anticipate the unknown.

As a result, there are some areas in which humans must remain involved, even as AI become more proficient:

·      Humans will need to oversee AI outputs to ensure that they are relevant, comprehensive, accurate and unbiased, given limitations with AI programming.

·      While AI can quickly and brilliantly capitalize on what it knows, it doesn’t have a creative spark to generate captivating new ideas. 

Steve Wozniak reinforced this at the NBCU's recent One23 conference.

As the initial shock and awe of ChatGPT-3 wears off, the industry must tackle a thorny question: “How should marketing and media organizations evolve?” 

Based on learning from The AI Conundrum: Why Artificial Intelligence’s Strengths Are Also AI’s Weaknesses, here are some areas in which they should begin to focus:

·      The importance of creativity is unlikely to change. If AI language-generating apps spit out the same language to everyone, creativity will still be required to differentiate communications. However, processes will need to evolve to incorporate AI’s strengths in improving efficiency and bringing in new information that will underpin creativity.

·      While attention to date has focused on headcount, the nature of jobs will need to change. Job descriptions and criteria for hiring and evaluations need to be rethought. For example, who will provide the oversight to avoid incorporating inaccurate or biased AI output?

·      Even if AI handles rote tasks more efficiently than humans, some of that work provided valuable training. Training programs therefore will need to be revamped, including how ChatCPT-3 and its competitors affect education for younger employees.

·      The impact on big tech -- and beyond -- needs to be carefully monitored. The press has reported on ChatGPT-3's impact on Apple, Google, Meta, and Microsoft. The ramifications of language-generating AI will extend far beyond these giants.

·      Changes cannot be made in a vacuum. Companies should encourage employees to experiment immediately with ChatGPT-3 and its soon-to-released successor, ChatGPT-4. Unlike with many prior tech programs, ChatGPT does not require a big outlay of money (at least for now). Still, companies will need to set aside employees’ time and encourage reporting on their experiences to assess the potential impact and implications.

The saying goes that lightning does not strike the same place twice. Marketing organizations successfully adapted when computers were able to assume many of the mathematical responsibilities that had previously fallen to humans. 

Lessons learned from that experience can be useful, although technology that can take over writing tasks, of course, presents new and different challenges. 

The good news is that human creativity will be needed to figure out solutions, and that is what great marketing organizations do best.

Next story loading loading..