Less Is More

Highlights

Top Insights

AI advantage is shifting away from access to technology and toward organizational capability. The tools are increasingly available to everyone; the differentiator is whether a company can repeatedly turn them into business outcomes at scale. McKinsey’s study of 20 companies found average EBITDA improvement of about 20%, with roughly $3 of incremental EBITDA for every $1 of one-time investment.

The strongest performers did not spread investment across dozens of disconnected use cases. Two-thirds concentrated on three business domains or fewer, choosing places where small operational improvements have large economic consequences.

Source: The new management playbook for AI: How to move faster and create more value (McKinsey Quarterly)

Top News

1. Z.ai released MIT-licensed GLM-5.3-Flash
2. Google released Gemini Omni 1.1 Flash with scene extension up to 40 seconds.
3. OpenAI added shareable scheduled and webhook-triggered tasks to eligible ChatGPT Enterprise, Edu, and Healthcare workspaces.
4. Salesforce expanded Headless 360 with MCP servers, Data 360, Slack integrations.
5. Apple introduced an M5 Max and M5 Ultra Mac Studio for large on-device models.

Additional Insights

1. The decision dividend: How AI creates economic value (McKinsey)
McKinsey argues that AI’s biggest economic payoff is not primarily labor savings, but a “decision dividend”: dramatically lowering the cost and time required to make high-quality business decisions. Because companies can access powerful AI through shared cloud infrastructure, AI-assisted decisions can be hundreds or thousands of times cheaper than human-mediated processes and can compress decision cycles from weeks to seconds. Yet simply adding copilots or chatbots to existing workflows tends to produce modest gains; the strongest performers redesign end-to-end processes around AI, with McKinsey citing roughly 20% EBITDA improvement among leading adopters. The value comes mainly from three sources—acting faster, extracting more output from existing assets, and capturing opportunities that slower organizations miss—and these benefits can compound without necessarily reducing headcount. Accordingly, McKinsey recommends measuring decision throughput, the share of important decisions informed, accelerated, or automated by AI, rather than tracking licenses, models, or pilots. Leaders should focus AI first on high-value decisions, prove impact in a few areas before scaling, establish clear human accountability and governance, measure actual financial outcomes, and treat implementation delays as a real economic cost. The ultimate competitive advantage, the article concludes, will belong not to companies that merely possess AI, but to those that redesign their organizations to consistently turn faster, better decisions into better business results.

2. ‘No One Works Here’ urges firms to replace human bottlenecks with AI efficiency (Ideas Made to Matter)
MIT Sloan senior lecturer Paul Cheek argues in No One Works Here that many established companies are slowed less by inadequate technology than by “structural debt”: layers of hierarchy, meetings, approvals, and information routing designed around human limitations. In an AI-driven enterprise, he says, organizations should deliberately remove people from routine decisions—such as pricing changes, supply-chain routing, and customer-support tasks—when machines can act faster and reliably, while preserving human judgment for high-stakes situations where its distinctive capabilities matter. Cheek acknowledges a “reliability tax”: early AI-agent deployments can initially create more work because humans must audit outputs, producing a temporary slowdown before systems earn enough trust to operate autonomously. His larger warning is that companies clinging to human-in-the-loop processes everywhere may face an existential competitive disadvantage as AI dramatically compresses the time required to make decisions, build products, and respond to markets.

3. Lessons from Chinese business leaders about integrating AI in the workplace (WEF)
China’s experience with workplace AI suggests that sustainable productivity gains will come less from simply automating jobs and more from redesigning how work gets done. The article describes a three-stage transition—disruption, augmentation and eventually job creation—with companies already automating defined tasks such as coding and customer service, while discovering that human context, judgement and coordination remain essential. Leaders should therefore fix inefficient processes before adding AI, identify where accountability and human judgement belong, and value employees who combine domain expertise, learning agility and AI fluency rather than focusing only on technical specialists. A major risk is that automating junior-level tasks could eliminate the experiences through which employees traditionally develop judgement and prepare for leadership, so organizations need new career-development pathways in an increasingly “mapless” career environment. China is also treating workforce policy as part of AI infrastructure, linking labour-market signals, continuous training, internships and human-AI collaboration through its 2026–2030 planning. The broader lesson is that successful AI adoption requires process redesign, continuous capability building and career mobility to be built into AI strategy from the outset—not added after the technology is deployed.

4. Why ‘Identity Permission’ Is the New Competitive Advantage (Knowledge at Wharton)
Wharton professor Americus Reed II argues that the real competitive scarcity is shifting from attention to “identity permission”—a consumer’s willingness to let a brand, product, institution, or technology become part of how they see and express themselves. AI and algorithmic media are making attention easier to manufacture through endless personalized content, but they cannot manufacture the deeper psychological response of “this is me.” Brands such as Stanley, luxury labels, universities, and potentially AI platforms create durable value when people use them not merely for function but for self-definition, community, status, or aspiration. For marketers, this means moving beyond impressions, clicks, and engagement toward signs of identity attachment—such as uncompensated advocacy, displaying brand symbols, developing rituals or communities, defending the brand, or tolerating inconvenience to stay associated with it. Crucially, Reed says this permission is borrowed, not owned: when an organization changes its values or cultural meaning in ways that conflict with customers’ identities, the reaction can feel like personal betrayal. In an AI-saturated marketplace where competent content becomes ubiquitous, enduring differentiation will increasingly come from earning a meaningful place in people’s sense of who they are.

5. When AI Thinks with You: Governing the Strategic Reasoning of the Firm (CMR Insight)
The article argues that as generative AI shifts from helping managers produce content to actively helping them think through strategy, pricing, investment, innovation, negotiation, and risk, companies must govern not only the data employees enter but also the strategic reasoning revealed through those conversations. Even sanitized prompts can expose a firm’s assumptions, tradeoffs, decision criteria, competitive logic, and early ideas—effectively revealing “how the company thinks.” A survey of 130 senior executives found widespread AI use in decision-making but weak organizational visibility and limited understanding of how platforms retain, review, or reuse interactions. The author therefore proposes a platform-sensitive governance model: classify the sensitivity of the thinking itself, assess the governance characteristics of the platform rather than focusing only on model capability, and match sensitive reasoning to environments with appropriate retention, deletion, access, review, and reuse controls. Leaders should map where AI-assisted thinking already occurs, distinguish routine content work from judgment work, create protected environments for strategic reasoning, educate managers about “reasoning exposure,” and make verifiable deletion and non-retention part of governance. The central message is that competitive advantage increasingly depends on governing where and under what rules AI participates in managerial judgment, not simply on having access to the most capable models.