No matter what field you work in, there’s little escaping the conversation surrounding artificial intelligence. With the potential for increased productivity and sharper, timely insights, it’s easy to see why leaders are eager to adopt AI tools in their businesses.
According to The 2023 State of Social Media report, 94% of business leaders feel very confident about integrating AI into their workflows. Yet despite such assuredness, 98% of leaders acknowledge they need to better understand the long-term potential of AI.
This discrepancy is born from the ambiguity we all struggle with, between hyped potential and reality. Leaders are coming to ask important, unanswered questions about AI, from salient concerns like reliability and brand safety to fundamental ones, namely, which use cases will be right for their teams and when.
We’re in the early innings of this transition, with most of the value glimmering well ahead. Our tools of today are just beginning to be reimagined with AI. Business leaders should begin to develop their point of view on how, when and under what conditions AI products will be ready for their organization—and what change needs to occur internally before seriously pursuing adoption. Powerful AI will eventually become a common thread in our business tools, and now is the time for executives to build and prepare for that future together.
The gap between expectation and reality
The past year saw the introduction of AI in a number of productivity tools, from copywriting to graphic design to social media management. At Sprout, we leverage AI and automation to democratize tools like social listening, giving everyone easier access to social data. Tools like Grammarly and Notion use AI to expedite manual tasks like copyediting and project management.
While these are exciting examples, I would argue we’re just getting started. We are seeing incredibly advanced technology—“intelligence” even—that is equally a jack or joker of many tasks, and a consistent ace of few (although quite good at standardized tests).
Given that, let’s consider how AI might be used today for business and how not. To do so, let’s go to the source, and consider the opportunities and risks of using ChatGPT directly. As an open-ended tool, it’s easy to imagine the potential uses across many job functions. Many already find it useful for rough drafts of emails, or copy explorations for a social ad. But note these uses are fairly low-stakes, and depend on the human to correct for the AI’s flaws.
For generative AI to be used at scale by many job functions, much more refinement, controls and human preparation are needed. It is not ready for situations where accuracy is critical, unless a competent human is ready to diligently babysit the AI. Case in point, a federal judge recently issued a requirement for lawyers to certify they didn’t use AI to draft their filings without a human checking their accuracy, after a cavalier lawyer presented ChatGPT’s confident fictions in a court hearing.
Were executives to rush into AI without thoughtful consideration, like that negligent lawyer, it could manifest in real business consequences. Consider that an eating disorder hotline had to shut down its AI-powered chatbot because it was giving bad and even harmful advice. Similarly, robots that are trained using AI have been found to be racist and sexist, raising many questions around the ethics of AI programming. For all that AI has proved to be capable of, we have a ways to go before it can be treated as more than a virtual assistant. And even the “assistant” workflow assumes that the human team is trained to stay in charge (and that their software helps keep them accountable).
What we’re seeing today among AI tools and workflows is the first generation. In other words, the current state of AI is much like where the iPhone was when it first released in 2007. It was groundbreaking at the time, but we didn’t really understand what the iPhone was fully capable of until five, 10 years later, after the core technology advanced and a surrounding ecosystem was built and matured. Remember that the iPhone launched with no App Store.
The same could be said of generative AI. The surrounding ecosystem of business tools, from the application layer down to the infrastructure, has a ways to catch up. We vendors have been given a very curious gift, and we’ve yet to make the most of it. Forcing an immature AI tool to perform sophisticated activities or not rethinking your team’s training or workflows—particularly where sensitive discretion or accuracy matters—could stupendously backfire.
AI-informed leadership requires internal change and external collaboration
When the first office computer burst onto the scene, business leaders didn’t wake up one day and decide every desk would have a Xerox Alto. Going from an analog way of working to a digital one required technical implementation and even greater change management to build machines to match the work, and to adapt the work to fit the machines—that took time, education and internal buy-in.
Similarly, as AI tools evolve and become more intuitive, business leaders need to identify how their workforce and existing systems need to adapt for AI to be successfully onboarded. This time things will move at a much faster pace, but we can’t be hasty.
There are the obvious educational pieces that need addressing, with 39% of business leaders saying a blocker to AI implementation is insufficient AI training and development. Coupled with the 37% of executives who say there’s limited organizational experience working with AI and ML, it’s clear that the current skill sets of most workplaces aren’t adequately prepared for an AI-powered one. We’re all in this situation.
Take my realm of software development, as an example. You’d be excused for thinking that engineers are the best prepared. Nope.
As AI becomes part of a developer’s toolkit and shoulders some core responsibilities like writing and deploying code, what new roles do human developers play? Their jobs probably don’t go away, but their responsibilities certainly change. Job functions will shift to be more akin to a supervisor than a coder, forcing developers to develop new muscles. Are they ready? And, are they willing to accept the change? Could it inspire backlash from those who fear AI is replacing them or eliminating the craft and creativity of their work? Beyond ramming change for the sake of productivity, savvy leaders will find framings that motivate. I see metaphors like power tools or exoskeletons or assistants as realistic comparisons, and helpful mental models.
There’s also the technical work business leaders need to consider when investing in AI and, given its complexities, organizations are leaning on vendors to assist with its execution. AI isn’t one big destination feature; very few organizations will directly integrate with models on their own or have their teams chat directly with them. AI is a substrate that will become embedded throughout your stack, from chips, to databases, to application software.
Think of it not as a new type of tool, but a stronger building material for your existing tools. For vendors, there’s a responsibility to deploy AI solutions as an add-on to existing workflows, minimizing friction and prioritizing intuitive design. And for functional leaders like CMOs and CTOs, there’s a responsibility to observe how their teams leverage AI and share that feedback with their vendor partners for future iteration. We’re building the future of work together.
Slow and steady wins the AI race
The emergence of AI has already impacted the way some organizations work and how leaders are thinking of their future technology investments. From increasing productivity to simplifying data analysis, AI has shown early proof points of its potential.
But there are untapped opportunities we’ve yet to realize because AI, and the tooling that embeds it, needs time to mature. We still have to answer questions around safety and ethics, and to establish rules of engagement for how AI should be leveraged and where. There’s also the internal change management that needs to occur before executives even consider AI implementation. All of this is dynamic, and will evolve over time.
As business leaders increasingly consider AI for their tech stack, now is the time to do the foundational work required to prepare. Familiarize yourself with what AI can and can’t do, and where it can fit within your business workflows, building a point of view both on today and the future. Start vetting vendors so when it is time for implementation, you’re partnering with someone who will set your organization up for long-term success. Being early in the cycle, this is as much about technology vision as it is philosophy and collaboration. Is your partner seeing the future as you do, and are they interested in building it together? With a clear understanding of AI’s capabilities and commitment to true internal change management, business leaders will set their organizations up for effective AI adoption today and in the future.
For more insights on where executives see AI supporting their business goals, as well as the challenges they face when it comes to implementation, download The 2023 State of Social Media Report today.
How to supercharge your marketing strategy with AI automationPublished on February 19, 2024 Reading time 9 minutes
Top 15 sentiment analysis tools to consider in 2024Published on January 16, 2024 Reading time 6 minutes
How to use AI analytics for targeted business decisionsPublished on January 9, 2024 Reading time 11 minutes