Bits With Brains
Curated AI News for Decision-Makers
What Every Senior Decision-Maker Needs to Understand About AI and its Impact
MIT Researchers Examine Enterprise Applications of Generative AI
12/23/23
Editorial team at Bits with Brains
As interest in generative AI technologies surges following ChatGPT's 2022 debut, MIT Sloan scholars have closely analyzed opportunities for—and design considerations around—deploying these new tools in business settings.
Their aim? Ensuring organizations maximize benefits while minimizing potential downsides.
Studying call center trials, one team found conversational AI boosted productivity 14% on average for inexperienced agents. Though unintuitive, the largest gains impacted newcomers rather than veterans—helping address longstanding inequalities as fewer skills are gained over time on the job.
Additional research suggests even knowledge experts may see advantages if change is managed properly. By fostering greater peer learning, accountability and flexibility in roles, companies empowered highly-trained staffers with AI assistants—though subtle integration was critical.
In other work, job seekers randomly assisted by AI in resume writing experienced an 8% edge in hiring rates. As mundane but impactful as formatting may be, enhanced quality communicates candidate value—opening doors that would otherwise stay shut.
Early applications include high-level document summarization, personalized online experiments, and automated code snippet generation. Yet challenges also emerge, like how much domain familiarity is needed and whether results are repeatable.
Strikingly, when AI origins were unknown, people intrinsically preferred its writing to that of human counterparts. Though bias reversed knowing the author, mixed-source materials faced no such prejudice—suggesting productive human-machine collaborations may reassure.
Among policy ideas, MIT analysts propose ensuring tax parity between labor and capital, bolstering workplace surveillance safeguards, and establishing a central AI advisory body. Beyond enterprises, legal and technical fields appear ripe for generative help with court document research or linguistic barriers.
Sources:
https://mitsloan.mit.edu/ideas-made-to-matter/generative-ai-research-mit-sloan
Sources