Artificial Intelligence (AI) has stormed to prominence in recent years, with advances in machine learning grabbing headlines and its ethical and societal implications being a hot topic for debate, but the field of AI is nothing new. In fact, it was first founded as an academic discipline back in the 1950s.
My very first tech job was as a product manager at an AI company. In those days, we trained the system to think like an expert by capturing rules that would act when certain conditions were met, mirroring the logic of that expert.
More recently, Machine Learning has become mainstream, where you feed vast amounts of data to a system to train it on patterns to look for and actions to apply. Then you use the models you’ve built to automate key tasks. It’s more efficient, and often the machine finds patterns we didn’t know existed.
In our world of video conferencing, for example, we have data on millions of video meetings across locations around the globe. We can use that data to train our solution on how to handle changing network conditions and predict what locations and connections are likely to have bandwidth issues at certain times of day, based on traffic or other conditions.
The Business Impact
AI is helping to improve productivity across many industries, but is applied differently by industry use case. In manufacturing, you see it used in robotics and self-driving cars. In finance, AI is often implemented for purposed od fraud detection and transaction processing. In life sciences, AI helps researchers sifting through vast amounts of data looking for patterns that may lead to better drugs or outcomes for at-risk populations.
AI has already become quote prominent in video conferencing. At BlueJeans, we work closely with our partner Dolby to apply intelligence to improve meeting quality. Intelligent algorithms are used to filter out background noise and adjust video resolution to meet changing bandwidth conditions. We can now use voice recognition to start meetings and create transcriptions. In the future, we expect to see AI automating processes for situations like following up on meeting action items.
The Next Wave of AI
Over the next few years, experts like Isaac Sacolick, president and CIO of StarCIO and author of Driving Digital, expect AI practices to continue to progress and become increasingly prevalent in business, but not without a hitch:
“Aspects of AI such as supervised learning will be easier and more mainstream as ML platforms mature. More promising areas such as reinforcement learning algorithms will be used by early adopters with deeper data science skills. Proactive data governance, the steps to have data cleansed and ready to be used for ML experiments will remain a challenge for many organizations.” – Isaac Sacolick, President and CIO of StarCIO and Author of Driving Digital
Personally, I have no doubt we will see continued improvement in the accuracy of AI-driven processes and that will drive adoption in more and more use cases.
What we see emerging is a human-machine partnership, where increasingly AI automates routine processes and frees up people to do what we do best—create, inspire, teach, sell, recruit, collaborate, negotiate, and act. AI becomes an enabler of greater levels of human productivity, because we no longer need to manually update the system or send a document and can focus on doing our best work.
We also see it advancing in waves. Today AI is tackling things that it is proven to do well. In the video conferencing space those are things like background noise suppression or using sensors in a conference room to adjust to lighting conditions and intelligently frame people and white boards for a video conference. But as we talk to our enterprise customers, they’re still reticent to apply AI in areas where it could potentially create compliance, security or reputational issues, for example auto-dialing an unauthorized meeting participant into a meeting based on a misinterpreted voice command. As the technology matures and the track-record becomes solid, then a wave of increased adoption will follow.
Preparation is The Key to AI Success
One challenge with machine learning is that machines don’t always make it easy to discern why they are taking a specific action based on a patterns in a large data set. Just like an organization, AI needs governance. Make sure you have carefully thought through where you are comfortable applying AI and how it relates to your overall company policies with respect to privacy and security.
As you evaluate what use cases are ready for AI-based automation, run tests within a controlled environment to ensure it’s ready for prime time before rolling it out more broadly. From there, make sure you test and retest your machine learning models to avoid unforeseen outcomes. Whatever your industry, it’s not too early to start preparing for how the human-machine partnership can unlock better collaboration outcomes across your organization.