"Think about the tropes of culture around office work, the TV show 'The Office,' the movie 'Office Space,' and 'Dilbert' the cartoon ... they involve long meetings, too much email, those have been a constant for decades," he said today at VivaTech, the giant tech conference in Paris.
"People tend to not see that as part of their work," he said, referring to the way that meetings get in the way of the actual tasks people's jobs consists of.
"Over the last 20 or 30 years in most occupations people have become more productive," he said.
As an example he talked about recruiters. Today, they have LinkedIn, email, "tools for auto-sorting and scoring resumes as they come in." But "in 1991, all they had was the White Pages."
"The area that hasn't improved so much is communication," he said.
Anyone who has found themselves involuntarily contributing to a companywide reply-all chain can sympathise. We are 30 years into the digital age. Why is it that mere communication still seems like a chore?
The problem is so bad that Stewart believes most "work" done in an office environment is "wasted."
Think about "the theoretical maximum productivity of our company," he said, which has about 1,000 employees. "We're probably at about 70% wasted output. And we're better than average. It might be 90% in many organisations. I'm entirely making that up - 70% is how I feel rather than some data that I have," he joked.
That's why something that can automate rote communication tasks, like Slack, or Gmail, make such huge differences, he said. Small steps make a big difference. "Things as simple as Gmail's predictive suggestions for how to respond to email takes a little bit of the friction out of it."
Butterfield says he is hopeful that AI will free us from the drudgery of "Office Space," the Mike Judge comedy about a group of office workers whose lives are a living hell because their jobs require them to do only trivial tasks.
"Think about people with knowledge-worker positions, there will be more of a reliance on human intelligence [in the future], automating away things humans do poorly — like remembering things, doing arithmetic [or] comparing hundreds of millions of items with each other and detecting patterns," he said.
We can give the chores to the machines. The future will belong to humans who know how to operate software, who know what it is capable of, who know how to handle analytics and data science around organisations.
That is "going to pay huge dividends," he said.