When the artificial intelligence tool ChatGPT launched in November, it was the first exposure and experience many people had to AI, says Paul Roetzer, founder and CEO of Marketing AI Institute.
“It really raised the awareness about this form of AI, called generative AI, which is the ability for the machine to create text, images, videos, audio and code — the five major things it can generate,” Roetzer says.
While AI isn’t new — having been around since the 1950s — the introduction of ChatGPT awakened a lot of businesses to the technology, he says. “The challenge becomes who in their organization even understands this technology and can figure out its application in business,” Roetzer says. “We have a lot of businesses that have more awareness that [AI] is something they should care about, but they really don’t know what to do about it.”
AI can be a powerful tool for industries that put it to work. But company leaders have a number of things to consider when implementing AI, including knowledgeable staff. And with the rapid proliferation of AI — and its rapidly changing abilities — there are legal and ethical concerns surrounding its use.
In March, the Future of Life Institute issued an open letter signed by AI industry leaders calling for a six-month pause on “the training of AI systems more powerful than GPT4” until a shared set of safety protocols is developed and overseen by independent experts.
The concern is that these AI systems are “now becoming human-competitive at general tasks” and may pose “potentially catastrophic effects on society,” beginning with the impact on creative/knowledge-based jobs and the havoc it can wreak with misinformation.
Roetzer signed the letter, not because he believes the pause will or should occur, but simply to highlight the need for key discussions surrounding AI. “If nothing else, it raises awareness that there should be some conversations around the use of AI in political campaigns, and what’s going to happen in education,” he says. “There are these very real issues right now.”
Secondary and higher education officials, for example, are grappling with whether to allow students to use ChatGPT. “That, to me, is the stuff that matters,” Roetzer says. “The Future of Life Institute letter was a bit dystopian … it at least raises awareness that there are some very real, near-term issues related to society, higher education, government, that we should be talking about in the open.”
The speed of technology
Dave Razor, a Princeton-based motion graphics and animation specialist for EPIC Creative in West Bend, says the proliferation of AI is “extremely dizzying because it is moving so fast, and most people aren’t used to that. I’ve heard people compare what’s happening now as going to have as much of an impact on society as the internet did, but within two years as opposed to 20. This is going to fundamentally change our society, and it is already.”
In advertising and marketing, AI is streamlining processes such as content calendars, which previously took weeks to develop and now are created within minutes with ChatGPT. It also can create graphic designs.
“If you can imagine that impact, even on an employer, they’re looking at employing five people to do this task,” Razor says. “And now it can be done by one person in a quarter the amount of time. It’s frightening because it can get really out of control really quick.”
ChatGPT is popular and took off so rapidly because it’s conversational and doesn’t require technical know-how to use it, Razor says. “You can just talk just like we’re talking right now. And it can translate that into some complex tasks.”
However, Razor estimates a 10% “misalignment” with the information produced by a prompt fed into the GPT systems currently, meaning what it gives you back is not quite accurate or what the user intends.
“And if you keep building off your original prompts, if you don’t go in there and really tweak it, by the time you’re 10 prompts in, it’s starting to loop in on itself. And you can definitely tell it’s a machine.”
That flaw, however, is already being improved.
AI also can be used as a generative tool such as text to image (as with Midjourney or Stable Diffusion) or text to video, he says. That’s causing some ethical issues with art use and permission as some tools will generate art by scraping other work from the internet. The flip side is some companies like Adobe or Shutterstock have developed their own models for image generation that pull from their own artists. That’s offering a new opportunity for artists to be paid for their work and to protect companies from lawsuits against unauthorized use of intellectual property.
A new tool growing in popularity is Auto-GPT, in which the user can create AI agents to complete tasks like coding. AI also can build “digital twins” of manufacturing plants or office spaces to help companies determine space, workflow, environmental impacts or heating and cooling, maximizing efficiencies, Razor says.
To keep up with the addition of these new tools, businesses need to upscale their people to be able to work with this technology, understand how it can help, and what it means for their workforce and clients. “You need people that know the AI world to be able to navigate through it,” Razor says. “Otherwise you just fall into a rabbit hole and get lost.”
Every business wants to personalize experiences with stakeholders, whether that means consumers, partners or vendors, Roetzer says. “AI enables you to continually process information about the people you care about, where you can truly personalize all your communications and experiences with them,” he says. It can automate repetitive tasks, and enhance creativity, innovation and decision-making.
Companies still need the human factor when working with AI. “No matter what AI model that you’re using or building, the ethical issue always falls on the humans,” Razor says. “We can’t expect a machine to be ethical. It’ll always be the human.”
Michelle Schuler, manager of TechSpark Wisconsin for Microsoft, says the adoption of any new technology “should be approached with caution and ensure that they are taking steps to address potential concerns and mitigate risks.” These include areas such as data privacy and security, bias and fairness, ethical considerations, how it will integrate with existing systems, and the human impact.
But AI “can solve some of the most pressing problems of our society today” and New North companies “are likely to be using AI technologies in various ways to drive innovation, increase efficiency and improve customer experience,” she says adding that higher education institutions are exploring how AI tools can transform teaching and how students research and write. “There is a growing recognition that AI has the potential to revolutionize the entire educational experience, and many educators are cautiously eager to learn more,” she says.
Schuler says since Microsoft announced its AI-powered Bing search engine in February, it’s been at the forefront of every community meeting she attends. “Leaders from government, business and nonprofit sectors are eager to learn more about how they can incorporate AI into their organizations. They want to understand how AI can help them stay competitive and innovative, and how they can upskill their employees to take advantage of the latest technological advancements,” she says.
Impact on information
Eric Severstad, creative director for Weidert Group, oversees a group of four other writers who create digital content for business-to-business companies. He’s always looking for ways to help his team become more efficient and started working with some AI tools. At the time, he wasn’t very impressed with the writing that the AI tool offered. But then, ChatGPT appeared.
“It wasn’t until ChatGPT came out at the end of November that it really opened my eyes and a lot of other people’s as well,” Severstad says. The tool became an overnight sensation, both because it produces a large amount of information very quickly and has a conversational tone.
But the thing to be aware of is how confident it sounds. “When you read the results you get from ChatGPT, it sounds like it knows exactly what it’s talking about — and it doesn’t all the time,” Severstad says. “That’s the downfall. In the AI world, they call it ‘hallucinations,’ in which basically the AI tool will make stuff up, it’ll create facts. It does its best to try to give you the information you want.”
It will even invent sources, making it sound even more credible.
“In the business we’re in with very technical writing, we have to be exactly correct with everything we do,” Severstad says.
So AI isn’t replacing human writers — yet.
Since its introduction in November, ChatGPT has already developed into a “bigger, better, cleaner” version with fewer hallucinations. “It’s impressive and it’s terrifying,” Severstad says. “It can do so many things. It’s not just a language model. It writes code, it strategizes, and it does so many things that we wouldn’t expect a machine to be able to do.”
But very quickly, it’s also becoming the norm, Severstad says. “And every business, no matter what you do, is going to be influenced in using AI tools this year. And so much change is going to happen, and it’s happening right now immediately.”
Severstad posed an Insight question to ChatGPT — how are companies currently using AI or could be using AI?
It offered a list including automating tasks, personalizing customer experiences, improving decision making, enhancing efficiency and improving security.
“Overall, AI has the potential to help businesses operate more efficiently, make better decisions, and provide better experiences for their customers. As AI technology continues to evolve, it will become an increasingly important tool for businesses of all sizes and industries.”
It also said that AI can help personalize customer experiences, improve decision-making, enhance efficiency and improve security threats.
“The first thing I always think about, even in the businesses that we’re working with, is efficiency, driving efficiency — if we can get things done quicker, that’s the name of the game, especially now in business more than ever,” Severstad says.
To respond to the rapidly growing use of AI, Weidert Group created a public use statement on AI, outlining “how we will use it, how we won’t use it, what it means to us as an organization,” says Severstad, who leads an AI task force at Weidert.
The policy builds on one offered by Roetzer, who spoke in Green Bay in April for Experience Inbound, hosted by Weidert Group and Stream Creative.
“Once you understand what it is, you have to stand for something as an organization,” Roetzer says. “I think that’s an important first step for organizations, and you’ll probably see more of them throughout this year creating those kinds of things, once they understand why it’s important that they have [a public use statement].”
Down the road, Severstad sees potential in having “your own personal AI platform that knows you, that understands your business, and will keep your information secure and it’s not going to share,” he says. “So I think in the future, we’re all going to have the ability to use the tools safely without worry … that’s down the road, but for now, we really have to be careful about how we’re using the tools and what we’re telling (AI tools), because they remember everything.”
Roetzer advises business leaders to form internal AI councils, like the one he leads at Weidert Group. Companies can use AI in new ways to drive revenue and efficiency and reduce costs, “but it’s going to take all new talent and those people don’t exist,” he says.
The fastest way to build that talent is from within, through reskilling and upskilling.
“I think it’s really important that organizations figure out how they will responsibly use this technology, because it can be used for great good — but it can also be used for bad,” Roetzer says.
There’s a need for society at large to understand AI and what it’s capable of, Roetzer says. That includes knowing that it can generate very realistic-looking videos, including a recent national political ad. “It was the first salvo in a political cycle that is going to be dominated by synthetic media,” Roetzer says. “You’re going to have misinformation, disinformation, propaganda, synthetic content, probably from both sides. That is going to be everywhere. You’re going to have a society who doesn’t even know AI is capable of creating it.”
Companies may be excited about AI and its capabilities at creating efficiencies and reducing the workforce, but they also need to understand its implications.
“At a business level, we really need the leaders of these organizations to understand what AI is capable of today, what it’s going to be capable of in the coming months and years, and what that means to business and to workers.”
Editor’s note: Dave Razor is married to Insight Editor Amelia Compton Wolff.
