Short notes, pointers and references.
-
The authors of this article introduce the notion of Human-AI Team Chemistry.
Many organizations expect AI to automatically improve teamwork, but research shows the opposite can occur. Without intentional integration, AI can reduce engagement, narrow participation in meetings, and shift ownership away from the team. A five-month study of managers identifies a critical new capability: Human-AI Team Chemistry. This does not emerge naturally and must be deliberately developed through three practices. First, teams should engage with AI as a team, ensuring it responds to the group’s full context rather than a single user. Second, they should leverage AI’s role fluidity, assigning it multiple personas to deepen discussion. Third, teams must maintain collective ownership, jointly shaping prompts, debating outputs, and critically evaluating AI’s contributions. When applied, these practices significantly improve outcomes. Teams report higher engagement, stronger alignment, and better decisions, while also reducing risks like over-reliance on AI.
It’s Hard to Use AI as a Team. These 3 Practices Can Help., by Gabriele Rosani, Elisa Farri, Daniel Trabucchi and Tommaso Buganza.Importantly, a feature of teams where this is working well is that AI is integrated and utilized as a thought partner to help improve the perspectives and thinking of the humans on the team. They do not just hand off that thinking.
-
The emergence of increasingly capable artificial intelligence has been causing a fair amount of uncertainty in the mathematics community.
Jessica Randall, a South African mathematician for Google Developer Groups, says she sensed a collective existential dread rising among the young mathematicians [at the 12th Heidelberg Laureate Forum].
[…]
AI has the potential to transform their discipline. But there’s far less consensus on what that transformation will mean in practice.
What It Means to Be a Mathematician When AI Does the Math, by Benjamin Skuse.There is an important struggle ahead and I think we will be seeing a lot of it in the coming years.
-
Having recently read Nudge and Sludge, this New Yorker article on self-storage was particularly interesting.
Annual revenues are estimated to be more than forty billion dollars.
[…]
Last year, two professors from Stanford and one from Texas A. & M. published a paper in the American Economic Review in which they conclude, based on transaction data from a large payment-card network, that “cancellation frictions roughly double seller revenues on average.”
[…]
Cancellation frictions are a boon to the self-storage industry. The annoying, time-consuming, and often expensive chore of emptying a junk-filled storage unit is easy to postpone for a month, and then for one more, especially if an unnerving paper bill never arrives.
The American Religion of Self-Storage Facilities, by David OwenIt is a fascinating scenario, where a lot of the sludge is self-imposed, i.e. the inertia is powerful – whereas the vendors make it very easy to, well, just leave it be, not change anything.
-
With so much movement in the market for AI development tooling, the challenge to properly evaluate choices continues.
AI-powered coding assistants have significantly outperformed conventional evaluation methods, creating gaps between vendor claims and actual performance. This study presents Quality Assessment for AI Development tools, a comprehensive, vendor-neutral framework evaluating coding assistants across six dimensions.
Quality Assessment for AI Development Tools: A Comprehensive Framework for Evaluating AI Coding Assistants Beyond Vendor Claims, by Buse Erol Esirik and Ebru Gökalp.Some practical takeaways from the quite academic presentation: Evaluate in six different dimensions (linguistic capability, operational quality, generation ability, interaction quality, trustworthiness and sustainability). Model performance likely varies across dimensions. Take stakeholders into account – different stakeholders will weigh dimensions differently.
-
As AI continues to become adopted and change job functions, the authors of this article propose the new role of the agent manager.
As autonomous AI agents move from experimentation to execution, companies are discovering they need a new kind of leader to manage them. Drawing on examples from Salesforce and other large organizations, this article introduces the role of the agent manager—leaders responsible for orchestrating how AI agents learn, collaborate, perform, and work safely alongside humans. Just as product managers emerged during the software revolution, agent managers are becoming essential to translating strategic intent into reliable outcomes in an AI-powered, hybrid workforce.
To Thrive in the AI Era, Companies Need Agent Managers, by Suraj Srinivasan and Vivienne WeiI would imaging most anyone engaged in knowledge work will find themselves managing AI agents to some extent, regardless of their specific job title.
Perhaps Agent Manager will become a standard job title, though it seems (to me) early to make that prediction or to have real confidence about how how long that title would last.
-
In an uncertain environment, standard advice is to be adaptable.
Jobs in the technology, media, and telecom sectors are experiencing the fastest pace of skill turnover, according to a June 2026 report on the effects of AI from the professional services network PwC. And the World Economic Forum’s most recent Future of Jobs Report, published in 2025, found that employers across all sectors expect 39 percent of workers’ core skills to change by 2030.
What It Takes to Be an Adaptable Engineer, by Gwendolyn Rak.It is a great time to invest in learning skills that have enduring (and ideally transferable) value and to develop strong awareness of your specific industry’s (and organization’s) environment.
-
Camille Carlton (of Center for Humane Technology) speaking on The Future of Human Meaning.
Choice quote from the presentation:
This is what is happening in AI right now. We are in what I call ‘the messy middle.’ It is a time when a powerful technology has arrived, but the rules and protections for what it will affect have not.
The future of human meaning, by Camille CarltonThis is descriptive and points well to the core of the situation: Rolling out a technology faster than understanding its consequences, in an environment that does not yet offer appropriate protections.
-
MIT Technology Review recently presented coverage on how kids feel about AI.
What surprised us most was how much young people might be able to teach adults about AI, and how clearly the kids who use it could name what they will and won’t hand over. They’re not as worried that it will take their jobs as they are that it might harm society. And with increasing access to tools that could in theory do their thinking, their talking, or even their friend-making for them, it sounds as if most want to keep their hands on the wheel.
How kids feel about AI, in their own words, by Jen Swetzoff and Keeley McNamaraThe stories are heartening. Anecdotally, as a theme it matches what I have been hearing from our own daughter (13).
-
In the department of how AI might be involved in bringing about the end of the world:
The report, according to one of the sources, was “entirely false.” But it also “almost started a war,” the source said. Any US operation against a Chinese vessel could have risked spiraling into an armed conflict between the two nations.
[…]
Analysts have long feared that AI could lead to a catastrophic miscalculation if nation states are relying on poor or corrupted data — the kind of miscalculation that might lead the United States to fire on a Chinese ship based on inaccurate information.
Exclusive: US military had close call after using AI for false intelligence report, sources say, by Katie Bo Lillis, Zachary Cohen.This is terrifying in its implications.
-
The authors of the following paper examine community response (fracture?) related to Stack Overflow’s decision to utilize user-generated content in their effort to commercialize AI product offerings.
We examine how Stack Overflow’s AI initiatives impact its open source knowledge community. Motivated by visible community fracture, such as a moderator strike and content vandalism, we use netnography with trace data to follow contributor reactions to AI venture announcements.
Productivity Gain, Community Strain: Stack Overflow’s Community Response to its AI Initiatives, by Dewan Scholtz, Anastasia Griva, Efpraxia D. Zamani, Kieran ConboyThis is highly relevant and telling as a cautionary example, particularly because many online communities are probably working on figuring out their positioning with respect to AI.