A novel experiment at Procter & Gamble reveals that synthetic intelligence isn’t only a software—it’s turning into a real teammate that may match human collaboration.
For many years, the holy grail of enterprise efficiency has been efficient teamwork. We’ve organized whole corporations across the premise that collaboration beats particular person effort—that two heads are higher than one. However what occurs when a kind of “heads” is synthetic intelligence?
A exceptional discipline experiment involving 776 professionals at Procter & Gamble—and led by Harvard’s D^3 Institute, the place I’m an govt fellow—has essentially challenged our assumptions about teamwork, experience, and the way forward for collaborative work. The outcomes recommend we’re witnessing the emergence of what researchers name the “cybernetic teammate”: AI that doesn’t simply help however actively participates within the collaborative course of.
When teaming with AI equals teaming with people
The P&G experiment was elegantly easy but profound. Cross-functional groups of business and R&D professionals had been randomly assigned to work on actual product innovation challenges in 4 totally different configurations: people working alone; conventional two-person groups; people with AI; and groups augmented with AI.
The headline discovering is hanging: People working with AI delivered measurable efficiency enhancements—almost 40% features —that elevated them to the identical degree as conventional human groups. In different phrases, AI seems able to replicating the elemental advantages we’ve lengthy attributed to people when it comes to collaboration, together with the progressive energy of a number of human views.
Take into consideration what this implies. For generations, we’ve structured organizations round groups as a result of collaboration permits us to pool totally different experience, catch blind spots, and generate higher options than people working alone. The P&G examine exhibits that AI can present these identical collaborative advantages to a single particular person.
The cross-functional AI benefit
However right here’s the place the story will get much more attention-grabbing. Whereas AI-enabled people might match conventional groups, AI-augmented cross-functional groups delivered outcomes that had been in a completely totally different league. When researchers examined the highest 10% of options—the breakthrough concepts that might drive actual aggressive benefit—AI-enhanced cross-functional groups had been thrice extra more likely to produce them.
This wasn’t a marginal enchancment. The mix of numerous human experience working along with AI created a multiplicative impact that neither human-only groups nor AI-enabled people might obtain. It means that the way forward for high-stakes innovation lies not in changing human collaboration with AI, however in supercharging cross-functional groups with synthetic intelligence. The P&G experiment successfully exhibits that simply as AI enhances the efficiency of particular person employees, it could possibly to the identical with whole groups with unmatched high quality outcomes.
Giving entry to lacking experience and breaking down silos
One of many vital benefits of working in groups is the entry to complementary experience offered by numerous workforce members. The experiment found that AI might fulfill this position fairly effectively.
People, both technical or industrial, when working with out AI predictably created options that favored their very own experience area. Nevertheless, these people working with AI produced options that had been as balanced as cross-functional groups working with or with out AI. R&D specialists started proposing extra commercially viable options whereas industrial professionals developed technically sounder approaches. AI acted as a bridge, serving to workforce members entry and combine views outdoors their area experience.
Few enterprise commonplaces are as oft repeated as the necessity to “break down silos.” But doing so stays troublesome in apply for many organizations. Now, with the assistance of AI, companies will probably be higher in a position to essentially change how cross-functional groups function. As an alternative of gathering specialists who advocate for his or her purposeful perspective after which compromise, AI allows groups the place every member can assume holistically throughout features.
Within the case of the P&G experiment, the consequence was extra technically possible and commercially engaging options from the bottom up, moderately than negotiated settlements between competing viewpoints—simply the tangible “breaking down of silos” that leaders so typically search and may solely poorly approximate by means of reshuffling of reporting traces.
For organizations combating the basic challenges of cross-functional collaboration—territorial disputes, communication gaps, and suboptimal compromises—AI affords a pathway to real integration.
The emotional dimension: AI as a motivational associate
This experiment additionally challenges every thing we thought we knew about expertise’s affect on office satisfaction. Removed from creating a chilly, mechanical work expertise, AI collaboration enhanced constructive feelings in ways in which rival human teamwork itself.
The info is hanging. People working with AI confirmed a 46% enhance in constructive feelings—pleasure, power, and enthusiasm—in comparison with these working alone. However AI-augmented groups skilled an much more dramatic 64% enhance in constructive feelings. Concurrently, AI decreased adverse feelings like nervousness and frustration by roughly 23% throughout each particular person and workforce settings.
What makes this discovering attention-grabbing is that it reveals AI filling a job we by no means anticipated: that of the emotionally supportive and motivational associate. Historically, one of many key justifications for teamwork has been its psychological advantages—the power that comes from collaboration, the decreased stress of shared duty, the joy of constructing on one another’s concepts. The P&G experiment exhibits that AI can replicate many of those emotional advantages.
Maybe most tellingly, members who skilled these constructive emotional responses additionally reported considerably increased expectations for future AI use. This creates a virtuous cycle: Constructive AI experiences drive larger adoption, which results in extra refined AI interplay expertise, which in flip generates even higher outcomes and extra constructive experiences.
This emotional dimension helps clarify why AI adoption typically exceeds preliminary expectations as soon as folks really use it. It’s not nearly effectivity features—it’s about AI making work extra pleasant and fascinating.
Strategic implications for leaders
These findings reveal particular organizational design ideas that forward-thinking corporations ought to implement instantly:
Reimagine your innovation structure. The normal mannequin of assembling giant, numerous groups for each innovation problem is now out of date. As an alternative, deploy AI-augmented cross-functional groups as your major innovation unit. These groups mix the boundary-busting energy of AI with numerous human experience to persistently produce breakthrough options. This strategy can cut back the overhead coordination of conventional innovation processes whereas maximizing the likelihood of producing top-tier concepts from the beginning.
Remodel cross-functional workforce dynamics. Cease tolerating the dysfunction that plagues most cross-functional groups—the territorial battles, the compromised options, the countless coordination conferences. AI can get rid of these friction factors by enabling every workforce member to assume and contribute throughout purposeful boundaries. This isn’t nearly including AI to current groups; it’s about redesigning how cross-functional groups function from the bottom up. AI as a cybernetic teammate can bridge silos and herald lacking experience, enabling groups to work sooner and higher.
Practice ‘T-shaped’ AI collaborators. The P&G findings reveal that efficient AI collaboration requires a brand new kind {of professional}—one who combines deep area experience with refined AI interplay expertise. Workers can now not depend on broad, shallow information; they should deepen their purposeful experience whereas creating the flexibility to “dance” with AI throughout domains. The vertical stroke of the “T” turns into much more essential because it gives the substantive information wanted to information, problem, and refine AI outputs. In the meantime, the horizontal stroke expands dramatically as AI allows professionals to contribute meaningfully past their core experience. This twin improvement—deeper specialization paired with AI-enabled boundary spanning—creates professionals who can extract most worth from human-AI collaboration.
Corporations that construct these competencies first may have sustainable benefits in innovation pace and high quality.
Restructure challenge economics and timelines. If one AI-enabled particular person can match a conventional workforce’s output whereas working 16% sooner, your useful resource allocation must be adjusted accordingly. Initiatives that beforehand required months of coordination between features can now be accomplished by smaller, AI-augmented groups in weeks. This isn’t about cost-cutting—it’s about dramatically accelerating time-to-market whereas sustaining or enhancing output high quality.
Design for the emotional multiplier impact. The emotional advantages of AI collaboration create a compounding benefit. Workers who’ve constructive AI experiences develop into AI advocates, driving natural adoption all through the group. Conversely, poor preliminary AI experiences create resistance that’s troublesome to beat. Make investments closely within the first 90 days of AI implementation—coaching, assist, and thoroughly curated use circumstances that guarantee constructive emotional responses from the beginning. Within the coaching, have staff instantly expertise how AI allows them to carry out higher and develop new expertise as an alternative of the fears of deskilling and substitute.
The way forward for collaborative work
We’re witnessing the early phases of a basic shift in how information work will get finished. AI isn’t simply automating duties accomplished by particular person employees; it’s additionally now able to collaborating within the collaborative processes that drive innovation.
This doesn’t imply human collaboration turns into out of date. Somewhat, it suggests we’re coming into an period the place essentially the most highly effective problem-solving models will probably be human-AI ensembles that mix the very best of each worlds: human instinct, creativity, and area experience with AI’s huge information base, sample recognition, and skill to quickly discover resolution areas.
The businesses that determine easy methods to orchestrate these cybernetic groups—the place AI actually features as a teammate moderately than only a software—may have a big aggressive benefit within the innovation financial system. The long run belongs to those that can grasp the artwork of human-machine collaboration.
François Candelon is a associate at non-public fairness agency Seven2 and govt fellow at D^3 Institute at Harvard.
Learn different Fortune columns by François Candelon.
