Co-intelligence; Living and Working With AI

Having recently finished Ethan Mollick’s Co-intelligence; Living and Working With AI, I found his framework for human-AI collaboration to be a reinforcement of several themes we have explored here at Legalisims. His work mirrors our ongoing discussions regarding the "character" of law in an automated age, the essential "Strategic Bridge" of human judgment, and the looming risk of professional atrophy when we offload the cognitive "struggle" of drafting. While Mollick advocates for treating AI as a "person" to navigate its "Black Box" nature, he ultimately brings us back to the core challenge we’ve analyzed: managing the inherent transparency and accuracy limits of these probabilistic systems (p. 65).

For me, the most consequential takeaway from this framework is how AI acts as a 'great leveler,' effectively narrowing the performance delta and raising the floor for all practitioners. Mollick cites a study where law students near the bottom of their class used AI to equalize their performance with those at the top (p. 189). This occurs because many foundational legal tasks like drafting, summarizing, and initial analysis, can now fall within the Jagged Frontier, the uneven boundary where AI capabilities currently exceed human effort (p. 47). By automating the structural heavy lifting, AI provides a powerful cognitive prosthetic that raises the floor of professional output, potentially mitigating traditional inequalities between elite and non-elite practitioners (p. 189).

However, this equalization introduces the risk of "falling asleep at the wheel" (p. 129). When lawyers trust the machine blindly, their performance can actually drop below human-only levels if the task falls just outside the frontier of the AI's competence (p. 130). To maintain a professional edge, lawyers must realize that the value of human expertise has shifted from production to curation (p. 54). The most successful practitioners will be those who know exactly where the machine’s competence ends and their own expert judgment must begin, ensuring they remain the essential "human in the loop" to catch plausible but dangerous hallucinations (p. 52).

By mastering the boundary of the Jagged Frontier, lawyers can leverage AI to boost productivity without sacrificing the accuracy. The profession’s future lies in the hands of those who utilize innovation to broaden their reach, provided they maintain the essential human oversight required for complex legal work.

Co-intelligence; Living and Working With AI Mollick, Ethan,  2024, Book , 234 pages; 9780593716717