The prevailing wisdom among the presenters and panelists at the recent World Summit AI conference in Montreal can be reduced to these two statements:
In short, both organizations and their people are at risk of being outcompeted. In some industries this will come rapidly, and in others it will be more of a long-term adaptation.
But there are many challenges to be overcome by all organizations that want to emerge from this wave of change at the top of the food chain. Here are some of our key takeaways from the event:
Great tech fails if people won’t use it.
The word adoption probably came up more than any other over the course of the event. Of course this is nothing new - at Differly we help organizations across industries with technology selection and implementation, and adoption is baked into our process from start to finish for good reason.
But AI does have some novel features that make it particularly tricky: prompting, open-ended outputs, different models for different use cases, not confusing it when you provide reference documents or include past chat histories, and of course there are concerns about information and job security. Here's what we heard about adoption:
Employees are busy, with lengthy to-do lists and meetings to attend. For most people, if an AI tool doesn't provide an immediately useful solution to a problem, they'll continue doing things the way they always have.
The most common first step that organizations take when rolling out AI is to give a subset of employees Copilot licenses. This often hinders AI adoption. Why? Because Copilot uses less powerful models that tend to produce generic outputs that aren't all that useful.
Absolutely give your most innovative problem solvers access to licenses (with an appropriate usage policy) so that they can discover new solutions to high priority problems. But for most employees, their AI journey needs to be thought through more carefully:
AI technologies come with inherent information security risks. Sensitive or private information may be exposed to the technology for use in its training, meaning your data becomes part of the system's 'knowledge' and could appear as outputs of the system to users outside of your organization. Incorrect outputs may be relied upon as being true, with potentially severe consequences.
Leadership teams need to understand the risks and make informed decisions about what level of risk they are comfortable with. Without a clear AI usage policy, organizations face two likely outcomes:
While we were dazzled with presentations detailing how enterprises like Spotify and Wayfair are using AI, there is a lot of low hanging fruit for SMEs that won't require major IT investments and platoons of data scientists:
And of course, these lists of practical use cases tend to forget about what the research is showing is the most valuable benefit of AI for knowledge workers today: having a thinking partner to plan, ideate and solve problems with.
Get in touch with us today to learn more about our AI Executive Briefing. We will get you started on your journey of adopting AI-fueled advancements towards your strategic goals. We take the technical complexity out of the conversation, focusing on priority business needs and the people who ultimately make change happen.
Sign-up here and we'll send you our latest news, insights and blogs as they are published.