AI in the Workplace

It’s Not Magic: The Information’s Readers Speak Out About AI in the Workplace

Art: Clark Miller

The topic of artificial intelligence in the workplace touched a nerve with The Information’s readers. Some described AI as significant for the future of our species or as a new industrial revolution, while others said it’s just an app and not a new way to work.

Often they couched their commentary in social terms, reflecting the broad impact of AI outside the company walls. That impact cannot be overstated, considering that a majority agree AI will not be an equalizer for all employees, and companies need to prepare their workforces for the technology.

The 220 readers who participated in the survey (see “Methodology”) addressed the impact of AI in the workplace, how different groups of employees feel about AI and how well prepared organizations are for AI. This article presents statistical data gathered via the survey, augmented by respondents’ comments and interviews with founders of AI-related companies.

What Will Be the Impact of AI on Work and Beyond?

At the most enthusiastic end of the spectrum, some attribute almost supernatural powers to AI. “AI is allowing us to transcend the limits of our biology, creating a future where we can achieve more than ever before. This is a significant development for the future of our species,” writes the founder of a productivity-focused AI startup. Adds another respondent: “AI is the new industrial revolution, here to set the pace for businesses and governments for the next 120, 150 years.”

Others see AI as decidedly evolutionary and not revolutionary. “For now, AI is more like another app, not a new way to work,” says one survey respondent. Says another one: “AI will transform work and the workforce, but it’s not magic. It’s just like any other technology. It must be applied strategically with clear ties to the business outcome. Leaders are either overhyped or underinvested in AI today; the transformation is taking longer than it should.”

While AI may not be magic, Suman Kanuganti, co-founder and CEO of Personal AI, believes that how it transforms work can lead to magical outcomes.

Personal AI is a startup building “digital twins” for subject matter experts, lawyers, scientists, healthcare providers, executives, educators and thought leaders. It aims to equip workers with proprietary, institutional knowledge to help them make better decisions. This type of business acumen typically resides in the heads of corporate high achievers with years of experience. Personal AI’s goal is to make it available to all those who can benefit from its technology.

”The company’s focus is on highly compliant and regulated enterprises, where the high-skill labor is locked into a few professional services. “Anything that requires heavy human capital thinking,” says Kanuganti.

Personal AI doesn’t spew out information like general AI tools, which can dispense wrong information and are thus not considered trustworthy by many users. Personal AI is based on small language models, trained on contained information from within the client corporations.

“It’s less about getting an answer to a question and more about the synthesis,” says Kunaganti.

“The idea of the magical is this core decision-making support that we are able to provide to people to be more thoughtful and insightful.”

The Information’s survey reveals that a vast majority of respondents see AI as very beneficial for elimination of mundane tasks (100%), increasing time available for higher-value work (92%) and offering a more effective way to analyze data and draw insights (92%). However, that does not necessarily translate into improving employees’ work-life balance (39%), or providing a more effective and less stressful way to interact with customers (39%) or a more effective and less stressful way to interact with co-workers (23%).

How Do Employees Feel About Using AI at Work?

A majority of respondents say AI will not be an equalizer. Just 29% of survey respondents agree that AI will be a rising tide that lifts all boats in terms of workers’ careers.

Some respondents put the onus to define how AI will impact careers squarely on the shoulders of individuals. “AI will not empower every employee in the same way,” says one survey respondent. “People with more enthusiasm, more will, more courage, more ambition will benefit more greatly from AI. People focused only on their tasks risk losing jobs.”

Others believe that how AI will work out for employees also depends on how organizations manage its implementation. “Lack of a vision and inability to demystify the exact use cases and applications has created a level of fear within the workforce, which makes it challenging to drive adoption,” writes one survey respondent. “Promoting it as a complementary tool and coaching the teams accordingly will go a long way in driving adoption.”

The Information’s survey shows that employees’ levels of enthusiasm, challenge and concern about AI vary depending on the roles they have in the organizations (see the chart “The Enthusiasts, the Challenged and the Concerned”).

Somewhat surprisingly, considering that top executives are often hesitant about adopting new technologies, the C-suite is the most enthusiastic about AI. Some survey respondents believe the C-suite may be rushing into AI: “It is such a hot and trendy topic that C-suite and other higher positions will implement solutions that are not ready, useful or particularly cost-effective, under the guise of innovation and saving money, and the people at the bottom of the ladder will suffer the most.” Agrees another: “Gap in executive expectations relative to current AI capabilities is a concern.”

But Personal AI’s Kanuganti is not surprised by the C-suite’s excitement about AI. “[With AI] they will be able to scale revenue exponentially. For the first time, after 20 years of the internet, they see a new opportunity for shifting the fundamental balance sheet curves as they relate to human capital, which they have been trying to optimize for years,” he says.

Not surprisingly, senior IT executives, who will be accountable for the success of AI projects, are the most challenged, while they’re also quite enthusiastic about AI. Midlevel and junior IT executives, on the other hand, are among the most concerned.

Creatives—those who generate new content or designs—top the ranking of those most concerned about AI. That’s perhaps because recommendation engines are now able to create content that’s in some cases comparable to what a human might come up with and can do it much faster than a human.  AI may also steal their IP.

Alexandre Robicquet, CEO and co-founder of Crossing Minds, which provides AI-driven real-time personalization and recommendations, takes a philosophical approach to human versus machine creativity. “Recommendation engines, by definition, do not create any data,” he says. “Are we, as humans, ever creating something really from scratch? Or is creativity about how uniquely we assemble the pieces of information that we gather to create something new?”

“A creative person should focus on the question of what makes my thought process, my writing, my curiosity, unique,” says Robicquet. “And try to really push the large language models to do the bidding for what makes any creator unique.”

There is good news in the age of AI for those with high emotional intelligence levels. Seventy-three percent of survey respondents believe the soft skills of humans will become significantly more important. “It will take some time before AI can effectively ‘replace’ human workload outside of pure functionality,” says one respondent. “Audiences and buyers are wanting a ‘true’ connection and experience.”

Are Organizations Ready for AI in the Workplace?

For those who cannot agree whether a top-down vision or a bottom-up initiative is more critical when implementing AI in the workplace, our survey has a clear answer: Both are almost equally important to get right. Bottom up edges out top down by a mere 5% (70% versus 65%; see table). This agreement on the two seemingly opposite approaches underscores the intricacy of implementing AI in the workplace.

For one of the survey respondents, implementing AI in the workplace is like squaring a circle: “AI is too hard for companies to adopt widely [presumably via a top-down approach], and too difficult to control if adopted individually [presumably when adopted via bottom-up initiatives].”

“I’m extremely enthusiastic about the potential for transformation through AI but am decidedly skeptical that companies are presently ready to move quickly in the right direction. I think a lot of prep work remains to be started,” another respondent says.

The Information’s survey confirms that companies are not yet ready for AI implementation, with less than a quarter prepared for any area related to AI (from 23% who have invested in collaborative AI tools for employees to 12% who have redesigned relevant business processes to involve collaborative AI).

Crossing Minds’ Robicquet believes “companies  are much more ready than they think [for AI solutions].” However, he notes that much prep work needs to be done. That’s because, he explains, companies do not have a unified view of consumer behaviors. The information is siloed among different departments, like marketing or IT, with many companies lacking a global way to manage content.

“My company’s job to resurface the right information becomes tricky because the information itself is not properly listed, maintained and shared,” adds Robicquet. At the beginning of an engagement Crossing Minds helps clients agglomerate content and enrich it with tags or SEO information to enable retrieval.

For companies that are struggling with AI implementation, Ravin Thambapillai, co-founder and CEO of Credal, which builds secure AI assistants for enterprise operations, offers practical advice about what works. Based on Credal’s analysis of its clients’ AI adoption patterns, he notes which groups within a company are most successful at AI implementations, whether it’s better to have a preset idea of use cases or rely on experimentation, and how important it is to have AI-savvy employees.

Typically, AI strategy and adoption fall under the purview of an AI task force or an existing AI platform team. “What we saw is that customers that had the data platform team in charge of their AI did better than the companies that had a task force stood up specifically to drive AI adoption,”says Thambapillai.

Some customers come with a preset definition of high-value use cases and focus on getting those into production. Others come with a slightly more experimental approach and an open mind about what might turn out to be valuable. “It turns out that having an overly strong opinion about what use cases are going to be valuable early on tended to be, in the long term, a detriment,” says Thambapillai. “People on the ground who are experimenting are better at discovering use case value than those who are ideating in a vacuum.”

Credal’s analysis of how well companies are driving AI adoption found that those with a higher concentration of AI-savvy employees were four times faster at adoption.

CONCLUSION

The ultimate role of AI in corporate work life will be more up to the individuals and organizations than the technology itself. While there are still differing opinions about how transformative AI will be for the workplace, there are some guiding principles to follow to make it a success:

Design a strategy that combines a bottom-up and top-down approaches to rollout and implementation of AI tools. That can be achieved by encouraging experimentation with AI tools  while setting guidelines and boundaries for users.

Encourage employees to adopt AI tools. There are several avenues for this. Firstly, train employees in how to use AI tools. Secondly, since it is important for them to understand how you intend AI to augment their jobs, make sure the expectations are clear. Change management is key to success here. Thirdly, rethink employee incentives to motivate staff to adopt AI.

Create AI tools that are user friendly and help with the employee work-life balance rather than adding to the workload. Make sure the data to train language models lead to factual outcomes that will engender trust.

Methodology: This report is based on a survey of 220 of The Information’s readers, conducted in August 2024.

Company size: 47% of respondents came from companies with less than $10 million in annual revenues, 20% from companies with revenues from $10 million to $100 million, 11% with revenues from $100 million to $500 million, and 21% with revenues of $1 billion or more.

Industry: Respondents came from across all major industries, with the biggest groups representing technology, media and telecommunications (46%), professional services (15%) and the financial sector and capital markets (10%).

Functional area: Respondents came from across all functional areas, with the biggest groups from general management (32%), research and development (14%), marketing and communications (12%) and information technology (11%).

Title: Respondents’ titles ranged from CEO to employee, with the biggest groups being directors (18%), CEOs (17%), senior vice presidents and vice presidents(14%) and CEO owners (13%).

SOURCE：The Information Partnerships