The Way We Work: How AI plays out in the workplace

If 2023 was the year artificial intelligence stepped out of the pages of science fiction and became a going concern, then 2024 is the year we turn to the latest bot, engine or app and ask, “What have you done for me lately?” In short: we’re ready to put these AI offerings to work.

One of the most pervasive (if least flashy) ways that AI will work for us is in the form of so-called “answer engines.” Search engines dump a list of links on our screen and tell us to scour them ourselves, but answer engines like San Francisco-based platform Perplexity provide conversational summations—easy shortcuts to the information we want. It’s as though every query leads to an executive summary. The ease that answer engines provide is addictive and, at work, these new tools promise massive productivity gains.

Neel Singh, a Vancouver-based strategic advisor with a specialty in AI, notes that generative AI tools are “great compressors of time” that allow us to output work faster. Fact-finding speeds up, comprehension can be capsulized and, by extension, “activities like writing, trend analysis and competitive research all become easier.” Singh compares this moment to the advent of the calculator: “Doing long division on paper is an ancient and unproductive practice today.” That said: calculators do not hallucinate their answers, whereas AI tools like Perplexity are only on point most of the time. Users must bring a journalistic skepticism to the table, including an awareness of source material (which Perplexity supplies in a list of citations). Instantaneous “answers” might be cribbed from a peer-reviewed academic journal, or they might derive from some guy’s LinkedIn blog.

Very soon, Singh predicts, the ability to intelligently use generative AI tools will be as obvious a prerequisite as the ability to type: “It’s not hard to imagine that, during an interview, candidates will be asked how AI is integrated with their workflow and productivity.”

Mitch Joel, a Montreal-based tech commentator and creator of the Six Pixels of Separation podcast, feels that the distinction between those who embrace such tools and those who don’t “is going to be a greater problem than the original digital divide. This is going to scale to a much more dramatic place.” Getting employees comfortable with tools like answer engines—and fast—will be crucial work.

When an app like Perplexity can provide such immediate and comprehensive content, however, it’s also worth wondering whether answer engines will replace more workers than they aid. The Business Council of B.C. published an assessment of our province’s vulnerability to automation, which found that 42 percent of B.C. jobs could be automated in the near future. Ninety percent of jobs, meanwhile, could be partly automated. And it’s not only  rank-and-file employees who’ll have their work lives reimagined. Managers and execs, says Singh, will need to vet answer engines and the data banks that fuel them. At the same time, he predicts, “companies will have to adopt circumvention measures to protect their own IP and privacy. At some point, there will be several B2B tools to help with this.”

Joel agrees, suggesting that companies decide “effective ways to implement these tools so there’s no IP leak… outside of the corporate structure.” The attraction of using third-party tech must be balanced against the threat of that technology’s ability to harvest data.

A company that creates its own guidelines and then leans into answer engines and other AI tools will have a competitive advantage in the short run, Singh predicts. But, he adds, “With this adoption, they’ll need a strong investment in security and preventative measures.” No free lunch, in other words; those who allow answer engines to supercharge their company’s productivity will have to do the honest work of safely and responsibly wielding that new power. And those planning to punt that responsibility may want to  read up on Air Canada’s recent courtroom loss—the airline was found liable for its  chatbot’s mistakes and was forced to  repay a customer who was denied bereavement fare.

Finding intelligent ways to work with new technologies, and not against them, is usually a better way to bulletproof one’s work. Singh’s advice is to bring AI tools into your wheelhouse by first letting go of static ideas about your specialization. “Leverage generative AI so you have a broader range of problem-solving abilities,” he says. “And make it unique to you and the way you think, work and create.” Answer engines, for example, are only as useful as the humans who prompt them. And there’s an enormous difference between naive users—who don’t know how to think critically about an answer engine’s source material or outputs—and someone able to navigate an app like Perplexity to quickly create strong, far-reaching and actionable material.

Joel calls the move to answer engines an inevitability. “I don’t think it’s an either/or thing; I think it’s a with problem,” he says. And how best to work with an answer engine? “I’m not looking for it to provide me with correct answers. I’m looking for it to help me with my own thinking. I ask very specific things—keep it as tight as possible—because what I really want is my own creative work.” Holding on to one’s own intentions while receiving a flood of prefabricated “answers” may feel like dealing with a blowhard at a cocktail party. But it’s a skill that will define tomorrow’s workplace.

Just Asking Questions

Perplexity invites users to ask it a question, so we did. Here are three questions we tested the platform on

1. How can Vancouver’s housing crisis be solved? Perplexity gave a long answer with four different headings before settling on one conclusion:

In summary, addressing Vancouver’s housing crisis will require a multi-pronged approach that includes increasing the supply of affordable housing, curbing investor demand, and providing support for vulnerable households. Lessons from other cities can also inform effective solutions.

2. Should governments implement Basic Universal Income? The platform outlines the arguments for and  against UBI before summarizing: 

The search results present a mixed picture on the merits  of implementing a UBI. While UBI has the potential to reduce poverty and improve wellbeing, the extremely high costs  and potential negative impacts on work incentives are  significant drawbacks that would need to be carefully  considered. Ultimately, whether governments should  implement UBI is a complex issue without a clear consensus.

3. Who is the most-liked character on the TV show Friends? There is no grey area on this one. Perplexity amalgamates multiple rankings and lists to come up with what it considers to be a slam-dunk answer: 

In conclusion, Chandler Bing is overwhelmingly  considered the most popular and liked character on Friends, thanks to his hilarious personality and strong character development.

RIP, Matthew Perry.