Lars Lehne is Global CEO of Incubeta.
Assuming you’ve not spent the past year living off the grid in a lakeside cabin, hunting game, foraging for berries and navigating using only the stars, you have probably noticed some conversation around artificial intelligence. There’s been a lot of it. And yet, despite the fact that it’s all we’ve been talking about for the past year, these conversations are still littered with misconceptions and inaccuracies.
In the marketing industry particularly, it seems that while everyone wants to be heard discussing AI, few have taken the time to understand it. Instead, they’ve rushed to jump on the bandwagon, proudly demonstrating their AI solutions without a clear picture of what the technology really entails.
Frankly, it’s a little exhausting. And while the irony of writing an article about AI that starts by saying there is too much talk around the subject isn’t lost on me, it is about time we move away from the hype and start to truly acknowledge this tool for what it is—looking at its history, capabilities, risks and limitations.
AI Is Not A New Technology
AI has been a silent workhorse in our daily lives for quite a while, long before ChatGPT brought it to the forefront of watercooler conversations. Whether it’s suggesting the next show to binge-watch on Netflix, recommending a product on Amazon, or autocorrecting and predicting text in messaging apps, AI-based algorithms have been quietly making our lives easier for years.
What we have seen, however, is a huge leap forward in the capabilities of generative AI. Suddenly the technology had a name and branding, and it quickly became the next viral sensation. The advancements in large language models (LLMs) have moved us away from a vague, complex technology operating in the background using relatively straightforward statistical models to an AI that is relatable, accessible and able to understand the context and nuances of natural human language.
In essence, we crossed a threshold in human-computer interaction. Looking at the early days of voice-control technology, for example, the interactions were often one-sided. To a degree, they still are. Users need to adapt their language and speech patterns to meet the limitations of the voice assistant, and the novelty of Alexa telling us a “crude” joke about a vacuum cleaner quickly wore off.
LLMs are now paving the way for more natural interactions that are less about issuing commands and more about engaging in conversation. Not only does this continue to expand AI’s potential, but it also has a democratizing effect, making its advanced capabilities more accessible to those who may be less tech-savvy or unfamiliar with specific jargon. It is no wonder then that the demand for generative AI voice applications continues to rise.
AI Won’t Solve All Our Problems (But It Will Help)
Many people’s understanding of what AI is capable of has been heavily influenced by science fiction and popular media, where depictions of AI often veer toward the fantastical or dystopian. In truth, the AI we have today, exemplified by models like ChatGPT and DALL-E, is a form of “narrow AI.” It is highly skilled at specific tasks but lacks the cognitive flexibility and consciousness to tackle unfamiliar challenges. This means that currently, it is only as good as the data it is fed.
In other words, you put garbage in, you get garbage out. Should the data fed to your AI system be skewed or biased, the AI’s outputs will be too. Whether it’s the perpetuation of gender or racial stereotypes or the reliance on incomplete data sets, the technology inherits our flaws, limitations and ethical complexities. So, while AI can no doubt be a powerful business tool—used for everything from automation to fast generative execution—it is bound by the quality of human input.
That said, it is evolving. Even experts and those who work in the field of AI are constantly revising and exploring the potential of its future capabilities. And while concepts like the singularity—a hypothetical point at which technological growth becomes uncontrollable and irreversible—are speculative in nature, they do make for stimulating debate.
This leads nicely to my next point.
AI (Probably) Isn’t Going To Enslave Humanity
As we start to explore the unpredictable future of this advanced technology, there is both excitement and apprehension over where it will take us. Films like I, Robot and The Terminator may have skewed our vision slightly, but fortunately, we are still a very long way off from the days of robots rebelling against us. The current state of AI is more about augmenting human capabilities for specific tasks rather than replacing them.
There are, however, real and immediate concerns about the deployment of AI technologies that we need to address. Issues relating to privacy, technological bias, misinformation and weaponization pose severe risks, and as the technology continues to evolve, comprehensive regulation will become increasingly necessary to mitigate these. Governments, policymakers, ethicists and technologists will need to collaborate to ensure that AI continues to serve as a beneficial tool for humanity rather than a source of harm.
A Sensible Approach To Using AI
In the ever-evolving realm of AI, it’s essential that we move forward with informed footsteps, especially as an AI pioneer’s warning of its risks is eerily reminiscent of the moral dilemma discussed during the building of the first nuclear weapon.
The buzz around AI in sectors like marketing shows its allure, but we must look beyond the hype. AI is neither a panacea nor a looming disaster; it’s a powerful tool limited by its programming and data. To harness its potential responsibly, we need a nuanced understanding and vigilant debates about its risks and benefits.
Forbes Agency Council is an invitation-only community for executives in successful public relations, media strategy, creative and advertising agencies. Do I qualify?
Read the full article here